Search

Perspective on This Article from Lung-Cancer Chemoprevention by Induction of Synthetic Lethality in Mutant KRAS Premalignant Cells In Vitro and In Vivo

Background The mechanism of recurrence and metastasis of hepatocellular carcinoma (HCC) is complex and challenging. Methyl-CpG binding domain protein 3 (MBD3) is a key epigenetic regulator involved in the progression and metastasis of several cancers, but its role in HCC remains unknown. Methods MBD3 expression in HCC was detected by immunohistochemistry and its association with clinicopathological features and patient’s survival was analysed. The effects of MBD3 on hepatoma cells growth and metastasis were investigated, and the mechanism was explored. Results MBD3 is significantly highly expressed in HCC, associated with the advanced tumour stage and poor prognosis in HCC patients. MBD3 promotes the growth, angiogenesis and metastasis of HCC cells by inhibiting the tumour suppressor tissue factor pathway inhibitor 2 (TFPI2). Mechanistically, MBD3 can inhibit the TFPI2 transcription via the Nucleosome Remodeling and Deacetylase (NuRD) complex-mediated deacetylation, thus reactivating the activity of matrix metalloproteinases (MMPs) and PI3K/AKT signaling pathway, leading to the progression and metastasis of HCC Conclusions Our results unravel the novel regulatory function of MBD3 in the progression and metastasis of HCC and identify MBD3 as an independent unfavourable prognostic factor for HCC patients, suggesting its potential as a promising therapeutic target as well.

BACKGROUND Depression escalating public health concern and the modest efficacy of currently available treatments have prompted efforts to identify modifiable risk factors associated with depression symptoms. Physical inactivity, poor nutrition, or other lifestyle behaviors are among the potentially modifiable risk factors most consistently linked with depression. Past evidence regarding the single effect of physical activity (PA) or dietary quality (DQ) on reducing the risk of depression symptoms has been well-documented. However, the association of the joint effect of PA and DQ on depression symptoms has never been investigated in a representative sample of adults. OBJECTIVE This study investigates the association between PA and depression symptoms and between DQ and depression symptoms, and their combined effects on US adults. METHODS Data were obtained from the National Health and Nutrition Examination Survey (NHANES) 2007 to 2018 cycles. The primary exposures were DQ and PA, measured using the Healthy Eating Index (HEI)-2015 and the metabolic equivalent (MET) minutes per week reported in questionnaires, respectively. Depression symptoms were defined as a 9-item Patient Health Questionnaire (PHQ-9) score of ≥10. We created 4 lifestyle categories: healthy diet and active individuals, unhealthy diet but active individuals, healthy diet but inactive individuals, and unhealthy diet and inactive individuals. Participants were considered to have a healthy diet if they fell within the 60th percentile of the HEI-2015 or to be active if they met the current guidelines for PA. A survey-multivariable logistic regression approach was used to model adjust the variables relevant to the associations, and an age-adjusted prevalence for depression symptoms was calculated following the NHANES guidelines. RESULTS In total, 19,295 participants represented a weighted number of 932.5 million adults aged 20 to 80 years in the noninstitutionalized US population. The total age-adjusted prevalence of depression symptoms among all respondents was 7.08% (1507/19,295). Of the respondents, 81.97% (15,816/19,295) met the PA recommendation and 26.79% (5170/19,295) scored at or above the 60th percentile on the HEI-2015. Depression symptoms were inversely associated with a higher level of PA (adjusted odds ratio [AOR] 0.819, 95% CI 0.716-0.938) and healthy DQ (AOR 0.809, 95% CI 0.701-0.931), respectively. A healthy diet combined with recommended PA was associated with a significantly lower risk of depression symptoms (AOR 0.658, 95% CI 0.538-0.803) than those who consumed an unhealthy diet but were physically active (AOR 0.890, 95% CI 0.765-1.038) or consumed a healthy diet but were physically inactive (AOR 1.077, 95% CI 0.817-1.406). CONCLUSIONS Our findings indicate that people with a healthy diet and recommended PA have a lower risk of depression symptoms than those with an unhealthy diet and a low level of PA. A healthy dietary habit and regular PA are potential precautions against depression.

This study compared whole blood dilution versus density gradient centrifugation for pre-processing blood samples prior to circulating tumor cell (CTC) capture on the efficiency of CTC separation by size-based isolation.Whole blood from a healthy volunteer spiked with SKBR3 cells was used to optimize the whole blood dilution protocol for sample volume, dilution ratio, and paraformaldehyde (PFA) concentration. Whole blood from healthy volunteers spiked with SKBR3, A549, or PC3 cells, and whole blood from patients with advanced gastric, esophageal, or liver cancer, was used to compare pre-processing by the optimal whole blood dilution protocol with density-gradient centrifugation. All statistical evaluations were performed using Student t test of the Statistical Package for Social Sciences (SPSS version 17.0).In blood samples from healthy volunteers, spiked SKBR3 cell recovery rates were highest in 5 ml of whole blood, diluted with 2.5 ml buffer, and fixed with 0.2% PFA, and spiked SKBR3, A549, and PC3 cell recovery rates from 5 ml whole blood were significantly greater when using the optimized whole blood dilution protocol (87.67% ± 1.76%, 79.50% ± 0.50% and 71.83% ± 1.04%, respectively) compared to density-gradient centrifugation (46.83 ± 1.76%, 37.00 ± 1.50% and 41.00 ± 1.50%, respectively).

Transformer-based deep learning models have become a ubiquitous vehicle to drive a variety of Natural Language Processing (NLP) related tasks beyond their accuracy ceiling. However, these models also suffer from two pronounced challenges, that is, gigantic model size and prolonged turnaround time. To this end, we introduce ET. that rE-thinks self-attention computation for Transformer models on GPUs with the following contributions: First, we introduce a novel self-attention architecture, which encompasses two tailored self-attention operators with corresponding sequence length-aware optimizations, and operation reordering optimizations. Second, we present an attention-aware pruning design which judiciously uses various pruning algorithms to reduce more computations hence achieves significantly shorter turnaround time. For the pruning algorithms, we not only revamp the existing pruning algorithms, but also tailor new ones for transformer models. Taken together, we evaluate E.T. across a variety of benchmarks for Transformer, BERTBASE and DistilBERT, where E.T. presents superior performance over the mainstream projects, including the popular Nvidia Enterprise solutions, i.e., TensorRT and FasterTransformer.

Rationale αv integrins, key regulators of transforming growth factor-β activation and fibrogenesis in in vivo models of pulmonary fibrosis, are expressed on abnormal epithelial cells (αvβ6) and fibroblasts (αvβ1) in fibrotic lungs. Objectives We evaluated multiple αv integrin inhibition strategies to assess which most effectively reduced fibrogenesis in explanted lung tissue from patients with idiopathic pulmonary fibrosis. Methods Selective αvβ6 and αvβ1, dual αvβ6/αvβ1, and multi-αv integrin inhibitors were characterized for potency, selectivity, and functional activity by ligand binding, cell adhesion, and transforming growth factor-β cell activation assays. Precision-cut lung slices generated from lung explants from patients with idiopathic pulmonary fibrosis or bleomycin-challenged mouse lungs were treated with integrin inhibitors or standard-of-care drugs (nintedanib or pirfenidone) and analyzed for changes in fibrotic gene expression or TGF-β signaling. Bleomycin-challenged mice treated with dual αvβ6/αvβ1 integrin inhibitor, PLN-74809, were assessed for changes in pulmonary collagen deposition and Smad3 phosphorylation. Measurements and main results Inhibition of integrins αvβ6 and αvβ1 was additive in reducing type I collagen gene expression in explanted lung tissue slices from patients with idiopathic pulmonary fibrosis. These data were replicated in fibrotic mouse lung tissue, with no added benefit observed from inhibition of additional αv integrins. Antifibrotic efficacy of dual αvβ6/αvβ1 integrin inhibitor PLN-74809 was confirmed in vivo, where dose-dependent inhibition of pulmonary Smad3 phosphorylation and collagen deposition was observed. PLN-74809 also, more potently, reduced collagen gene expression in fibrotic human and mouse lung slices than clinically relevant concentrations of nintedanib or pirfenidone. Conclusions In the fibrotic lung, dual inhibition of integrins αvβ6 and αvβ1 offers the optimal approach for blocking fibrogenesis resulting from integrin-mediated activation of transforming growth factor-β.

Cerebral ischemic stroke (IS) is still a difficult problem to be solved; energy metabolism failure is one of the main factors causing mitochondrion dysfunction and oxidation stress damage within the pathogenesis of cerebral ischemia, which produces considerable reactive oxygen species (ROS) and opens the blood-brain barrier. Dichloroacetic acid (DCA) can inhibit pyruvate dehydrogenase kinase (PDK). Moreover, DCA has been indicated with the capability of increasing mitochondrial pyruvate uptake and promoting oxidation of glucose in the course of glycolysis, thereby improving the activity of pyruvate dehydrogenase (PDH). As a result, pyruvate flow is promoted into the tricarboxylic acid cycle to expedite ATP production. DCA has a protective effect on IS and brain ischemia/reperfusion (I/R) injury, but the specific mechanism remains unclear. This study adopted a transient middle cerebral artery occlusion (MCAO) mouse model for simulating IS and I/R injury in mice. We investigated the mechanism by which DCA regulates glycolysis and protects the oxidative damage induced by I/R injury through the PDK2-PDH-Nrf2 axis. As indicated from the results of this study, DCA may improve glycolysis, reduce oxidative stress and neuronal death, damage the blood-brain barrier, and promote the recovery of oxidative metabolism through inhibiting PDK2 and activating PDH. Additionally, DCA noticeably elevated the neurological score and reduced the infarct volume, brain water content, and necrotic neurons. Moreover, as suggested from the results, DCA elevated the content of Nrf2 as well as HO-1, i.e., the downstream antioxidant proteins pertaining to Nrf2, while decreasing the damage of BBB and the degradation of tight junction proteins. To simulate the condition of hypoxia and ischemia in vitro, HBMEC cells received exposure to transient oxygen and glucose deprivation (OGD). The DCA treatment is capable of reducing the oxidative stress and blood-brain barrier of HBMEC cells after in vitro hypoxia and reperfusion (H/R). Furthermore, this study evidenced that HBMEC cells could exhibit higher susceptibility to H/R-induced oxidative stress after ML385 application, the specific inhibitor of Nrf2. Besides, the protection mediated by DCA disappeared after ML385 application. To sum up, as revealed from the mentioned results, DCA could exert the neuroprotective effect on oxidative stress and blood-brain barrier after brain I/R injury via PDK2-PDH-Nrf2 pathway activation. Accordingly, the PDK2-PDH-Nrf2 pathway may play a key role and provide a new pharmacology target in cerebral IS and I/R protection by DCA.

Although Transformer-based deep learning models have been widely used in many natural language processing (NLP) tasks as well as computer vision, they suffer from gigantic model size and long latency. Network pruning can reduce the computational cost and model size. However, existing works mainly focus on irregular(sparse) pruning, which often causes irregular computations and extra indices per remained weight. In this work, we propose a Tensor-core inspired hierarchical model compression method to push the performance limit on modern GPUs. We present two modes of the two-step process. In the first mode, we use the Tensor-core aware block-based weight pruning method to exploit model sparsity in a coarse-grained manner and then use low-rank [33] decomposition to further reduce the weight storage in a fine-grained manner.In the second mode, we first use irregular pruning to achieve a highly sparse model and then apply the Tensor-core aware weight constraint on the sparse model to decompose the sparse matrix to several smaller but Tensor-core friendly sub-matrices. Experiments on Transformer, BERTBASE models show the proposed method outperforms the state-of-the-art.

Recently, Graph Neural Networks (GNNs) have exhibited high efficiency in several graph-based machine learning tasks. Compared with the neural networks for computer vision or speech tasks (e.g., Convolutional Neural Networks), GNNs have much higher requirements on communication due to the complicated graph structures; however, when applying GNNs for real-world applications, say in recommender systems (e.g. Uber Eats), it commonly has the real-time requirements. To deal with the tradeoff between the complicated architecture and the high-demand timing performance, both GNN architecture and hardware accelerator need to be optimized. Network-on-Chip (NoC), derived for efficiently managing the high-volume of communications, naturally becomes one of the top candidates to accelerate GNNs. However, there is a missing link between the optimize of GNN architecture and the NoC design. In this work, we present an AutoML-based framework GN-NAS, aiming at searching for the optimum GNN architecture, which can be suitable for the NoC accelerator. We devise a robust reinforcement learning based controller to validate the retained best GNN architectures, coupled with a parameter sharing approach, namely ParamShare, to improve search efficiency. Experimental results on four graph-based benchmark datasets, Cora, Citeseer, Pubmed and Protein-Protein Interaction show that the GNN architectures obtained by our framework outperform that of the state-of-the-art and baseline models, whilst reducing model size which makes them easy to deploy onto the NoC platform.

Although Transformer-based language representations achieve state-of-the-art accuracy on various natural language processing (NLP) tasks, the large model size has been challenging the resource constrained computing platforms. Weight pruning, as a popular and effective technique in reducing the number of weight parameters and accelerating the Transformer, has been investigated on GPUs. However, the Transformer acceleration using weight pruning on field-programmable gate array (FPGAs) remains unexplored. This paper investigates the column balanced block-wise pruning on Transformer and designs an FPGA acceleration engine to customize the balanced blockwise matrix multiplication. We implement the Transformer model with proper hardware scheduling, and the experiments show that the Transformer inference on FPGA achieves 10.35 ms latency with the batch size of 32, which is $10.96 \times$ speed up comparing to CPU platform and $2.08 \times$ speed up comparing to GPU platform.

Tumor metastasis is the main cause of death in advanced colorectal cancer. Our previous research showed that upregulation of KIAA1199 predicted poorer outcomes, and promoted cell motility and tumor metastasis in colorectal cancer, with the mechanisms not being fully elucidated. Here, we demonstrate that silencing of KIAA1199 results in reduced tumor metastasis in the orthotopic transplantation tumor model of colorectal cancer. Importantly, we find that KIAA1199 interacts with protein phosphatase 2A (PP2A) through the C-terminal domain and increases phosphatase activity of PP2A, which is essential for KIAA1199-mediated cell motility. Moreover, we identify stathmin, a microtubule-destabilizing protein, as a downstream of KIAA1199-PP2A complex. KIAA1199-induced dephosphorylation of stathmin results in microtubule destabilization and leads to enhanced cell motility. Furthermore, a microtubule-stabilizing drug paclitaxel could prevent KIAA1199-induced microtubule destabilization, and inhibit cell migration and invasion in vitro and tumor metastasis in vivo in colorectal cancer. Collectively, our study reveals that KIAA1199 promotes metastasis of colorectal cancer cells via microtubule destabilization regulated by a PP2A/stathmin pathway, and suggests that KIAA1199 may be a promising target for preventing metastasis in colorectal cancer.

Colorectal cancer is one of the major causes of death from cancer. Metastasis is the leading cause of treatment failure, in which cancer stem cells and circulating tumor cells play crucial roles. Identifying the involved metastatic biomarkers and clarifying the regulation mechanisms are of great importance for targeting tumor metastasis. In the current research, we discovered that KIAA1199, a cell-migration inducing protein, showed higher expression in CD44+ cancer cells from metastatic compared with the paired primary tissues, and was upregulated in colorectal cancer and positively correlated with numbers and mesenchymal phenotype of circulating tumor cells, and predicted shorter progress-free survival. Moreover, we indicated that down-regulation of KIAA1199 suppressed migration and invasion of colorectal cancer cells in vitro, and inhibited metastasis in vivo. Furthermore, we demonstrated that KIAA1199 was one of the direct and functional targets of miR-216a, and miR-216a overexpression led to decreased migration and invasion of colorectal cancer cells in vitro, and inhibited metastasis in vivo. Collectively, KIAA1199 plays a critical role in maintaining an aggressive phenotype of tumor cells, and suppression of KIAA1199-related motilities of tumor cells contributes to reduced tumor metastasis in colorectal cancer.

The research purpose was the characterization of indigenous heavy metal-resistant plant growth-promoting bacteria (PGPB) from the farmlands located on the Le'an River basin contaminated by acid mine drainage and their effects on plant growth, nutrient uptake, antioxidant enzyme activities and metal accumulation. The plant growth-promoting (PGP) traits, including 1-aminocyclopropane-1-carboxylic acid deaminase, indoleacetic acid, siderophore, ammonia production and phosphate solubilization, as well as antibiotics, acid/alkali and salt resistance were determined. Ten isolates with relatively high PGP activities were identified to belong to the genera Burkholderia, Paraburkholderia, Cupriavidus, Pseudomonas and Ralstonia. The numerical classification based on bacterial resistant characteristics was mostly consistent with their phylogenetic positions. Burkholderia sp. strain S6-1 and Pseudomonas sp. strain S2-3 possessed both greater PGP activities and resistant characteristics in overall comparison. Compared with non-inoculated plants, strains S6-1 and S2-3 significantly increased the height, dry weight and N uptake of sorghum (Sorghum bicolor L.). The presence of S6-1 significantly increased Pb accumulation and enhanced the translocation of Zn from root to shoot in sorghum. Strain S2-3 helped sorghum to uptake Cu and Zn and improved the remediation effect of sorghum on Cu and Zn. Overall, indigenous PGPB did not show better advantages in improving plant growth than non-indigenous P. putida UW4. Nevertheless, indigenous PGPB can be used as better candidates in heavy metal phytoremediation to minimize the potential risks of introducing invasive microbial species into an agricultural ecosystem.

Gastric cancer studies indicated a potential correlation between circulating tumor cells (CTCs) in peripheral blood and tumor relapse/metastasis. The prevalence and significance of circulating tumor microemboli (CTM) in gastric cancer remain unknown. We investigated the prevalence and prognostic value of CTCs and CTM for progression-free survival (PFS) and overall survival (OS) in gastric cancer patients.Eighty-one gastric cancer patients consented to provide 5ml of peripheral blood before systematic therapy. CTCs and CTM were isolated using isolation by size of epithelial tumor cells and characterized by cytopathologists. For 41 stage IV gastric cancer patients, CTM was investigated as a potential biomarker to predict prognosis.CTCs were detected in 51 patients; the average count was 1.81. In clinical stage I, II, III, and IV patients, the average CTC counts were 1.40, 0.67, 1.24, and 2.71, respectively. CTM were detected in 3 of 33 clinical stage I to IIIb patients, at an average of 0.12 (0-2). CTM were detected in 13 of 53 clinical stage IIIc to IV patients, at an average of 1.26 (0-22). In stage IV patients, CTM positivity correlated with the CA125 level. PFS and OS in CTM-positive patients were significantly lower than in CTM-negative patients (P.001). CTM positivity was an independent factor for determining the PFS (P=.016) and OS (P=.003) of stage IV patients in multivariate analysis. Using markers of the epithelial-mesenchymal transition, single CTCs were divided into three phenotypes including epithelial CTCs, biphenotypic epithelial/mesenchymal CTCs, and mesenchymal CTCs. For CTM, CK-/Vimentin+/CD45- and CK+/Vimentin+/CD45- phenotypes were observed, but the CK+/Vimentin-/CD45- CTM phenotype was not. CA125 was detected in gastric cancer cell lines BGC823 and MGC803.In stage IV patients, CTM positivity was correlated with serum CA125 level. CTM were an independent predictor of shorter PFS and OS in stage IV patients. Thus, CTM detection may be a useful tool to predict prognosis in stage IV patients.

Obstructive sleep apnea-hypopnea syndrome (OSAHS) is a commonly-diagnosed chronic sleep disorder. It is considered to be an important independent risk factor in the development of insulin resistance (IR). Patients with OSAHS exhibit a variety of metabolic disorders, including obesity and metabolic syndrome. Visceral adipose tissue-derived serpin (vaspin) is an adipokine that is considered to be a link between obesity and IR. The present study aimed to evaluate the levels of plasma vaspin in patients with OSAHS and examine their potential correlation with sleep characteristics. A total of 20 healthy male subjects and 42 male patients with OSAHS were selected, and patients were divided into mild (n=22) and severe (n=20) OSAHS groups. The 20 patients in the severe OSAHS group received nasal continuous positive airway pressure (nCPAP) treatment for 2 months. Venous blood samples were drawn from all patients in a fasting state prior to and subsequent to nCPAP treatment, which were used to measure the levels of biochemical indicators. The sleep parameters and serologic index changes were compared prior to and following treatment. The values of contractive pressure (SBP), neck circumference (NC), waist circumference (WC), waist-to-hip ratio (WHR), body mass index (BMI) and hip circumference (HC) in the two OSAHS groups were significantly increased compared with those in the control group. In addition, the levels of vaspin in OSAHS patients were markedly increased and vaspin was revealed to be positively associated with fasting blood sugar, fasting insulin, triglycerides, homeostasis model assessment-IR, apnea-hypopnea index (AHI), NC, WC, BMI and WHR (P

Colorectal cancer is one of the major causes of death from cancer. Metastasis is the leading cause of treatment failure, in which cancer stem cells and circulating tumor cells play crucial roles. Identifying the involved metastatic biomarkers and clarifying the regulation mechanisms are of great importance for targeting tumor metastasis. In the current research, we discovered that KIAA1199, a cell-migration inducing protein, showed higher expression in CD44+ cancer cells from metastatic compared with the paired primary tissues, and was upregulated in colorectal cancer and positively correlated with numbers and mesenchymal phenotype of circulating tumor cells, and predicted shorter progress-free survival. Moreover, we indicated that down-regulation of KIAA1199 suppressed migration and invasion of colorectal cancer cells in vitro, and inhibited metastasis in vivo. Furthermore, we demonstrated that KIAA1199 was one of the direct and functional targets of miR-216a, and miR-216a overexpression led to decreased migration and invasion of colorectal cancer cells in vitro, and inhibited metastasis in vivo. Collectively, KIAA1199 plays a critical role in maintaining an aggressive phenotype of tumor cells, and suppression of KIAA1199-related motilities of tumor cells contributes to reduced tumor metastasis in colorectal cancer.

Lung cancer is the leading cause of cancer death in both men and women in the United States, with a low 5-year survival rate despite improved treatment strategies. These data underscore the great need for effective chemoprevention of this cancer. Mutations and activation of KRAS occur frequently in, and are thought to be a primary driver of the development of, non–small cell lung cancers (NSCLC) of the adenocarcinoma subtype. In this study, we developed a new approach for the chemoprevention of NSCLC involving specific targeting of apoptosis in mutant KRAS cells. This approach is based on a synthetic lethal interaction among TNF-related apoptosis-inducing ligand (TRAIL), the second mitochondria-derived activator of caspase Smac/DIABLO (Smac), and KRAS. Mutational activation of KRAS modulated the expression of TRAIL receptors by upregulating death receptors and downregulating decoy receptors. Furthermore, oncogenic KRAS repressed cellular FADD-like interleukin 1β–converting enzyme (FLICE)-like inhibitory protein (c-FLIP) expression through activation of Erk/mitogen-activated protein kinase (MAPK)-mediated activation of c-Myc. Smac overcame KRAS-induced cell-survival signaling by antagonizing X-linked inhibitor of apoptosis protein (XIAP). Therefore, the combination of TRAIL and a small molecule mimic of Smac induced apoptosis specifically in mutant KRAS cells without harming normal cells. We further showed that short-term, intermittent in vivo treatment with TRAIL and Smac mimic induced apoptosis in tumor cells and reduced tumor burden in a murine model of KRAS-induced lung cancer. These results reflect the potential benefit of a selective therapeutic approach for the chemoprevention of NSCLC. Cancer Prev Res; 4(5); 666–73. ©2011 AACR.

The effect of glutathione (GSH) on the ultraviolet (UV) induction of lambda prophage was investigated in lysogenic Escherichia coli. The data showed that extracellular GSH could inhibit the UV induction of lambda prophage. The inhibitory rates were concentration dependent, and the maximal rate obtained was 94% with 3.0 M GSH. The effect was also measured in three different lambda lysogens: a wild-type strain (wt), an isogenic GSH-deficient strain, and an isogenic strain producing increased amounts of GSH. The result showed that when subjected to UV irradiation (254 nm, 60 J m(-2)), GSH-deficient strain was approximately fivefold more sensitive to be lysed than wt, whereas the strain with higher intracellular GSH levels was only 28% susceptible to be lysed. With electron spin resonance and spin trapping techniques, we observed that free radical signals occurred in the suspensions of UV irradiated lysogenic cells and the intensity of signals was influenced by GSH levels. These results indicate that GSH can significantly inhibit the UV induction of lambda prophage, and that this effect is correlated to its capacity to scavenge free radicals generated after UV irradiation.

Current approaches for colorectal cancer (CRC) chemoprevention, including the use of NSAID, have tended predominantly toward modulating tumorigenic pathways rather than eliminating large number of premalignant or subclinical malignant cells. This tendency reflects the effort to keep toxicity to a minimum. The effectiveness of such modulative agents, however, is generally low to moderate, and long-term treatment is often required to produce any significant preventive benefit. In turn, long-term therapy frequently produces unanticipated adverse effects even with agents generally thought to have low toxicity. To overcome the challenges in current chemoprevention, we propose to perform chemoprevention in a therapy-like fashion, i.e., through short-term and intermittent treatments. The concept underlying this approach is based on elimination of large number of premalignant cells (to a greater degree than sought generally in chemoprevention) via short-tem treatments by specifically targeting genetic or epigenetic changes in premalignant cells. Periodically reducing or eliminating the number of premalignant cells through a series of noncontinuous/intermittent short-term therapies potentially will significantly reduce the chance of accumulation multiple mutations in same cells, which is required to generate malignant clones. This will delay or even completely prevent cancer in long-term. In this way, the potential side effects and costs often associated with long-term treatment could be minimized and controlled. In a recent paper (Nature. 2010 Apr 15; 464(7291): 1058-61. Epub 2010 Mar 28), we reported the development of a novel therapy-like chemoprevention procedure for CRC. This procedure is based on our finding that a deficiency in the adenomatous polyposis coli (APC) gene and subsequent activation of β-catenin lead to the repression of cellular caspase-8 inhibitor cFLIP expression through activation of c-Myc, and that all-transretinyl acetate (RAc) independently upregulates tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) death receptors and represses decoy receptors. Thus, the combination of TRAIL and RAc induces apoptosis in APC-deficient premalignant cells without affecting normal cells in vitro. In addition, we demonstrate that short-term and noncontinuous TRAIL and RAc treatment induces apoptosis specifically in intestinal polyps, strongly inhibits tumor growth, and promotes survival in ApcMin mice. We further show that TRAIL and RAc induce significant cell death in human colon polyps in organ culture in vitro, providing a potentially selective therapeutic approach for CRC chemoprevention by targeting APC-deficient cells for apoptosis. This study demonstrated the feasibility, efficacy, and potential tolerability of therapy-like chemoprevention in an in vivo preclinical setting. Progress toward a clinical trial will be discussed. Citation Information: Cancer Prev Res 2010;3(12 Suppl):CN13-05.

ARHI (aplasia Ras homologue member I; also known as DIRAS3) is an imprinted tumour suppressor gene, the expression of which is lost in the majority of breast and ovarian cancers. Unlike its homologues Ras and Rap, ARHI functions as a tumour suppressor. Our previous study showed that ARHI can interact with the transcriptional activator STAT3 (signal transducer and activator of transcription 3) and inhibit its nuclear translocation in human breast- and ovarian-cancer cells. To identify proteins that interact with ARHI in nuclear translocation, in the present study, we performed proteomic analysis and identified several importins that can associate with ARHI. To further explore this novel finding, we purified 10 GST (glutathione transferase)-importin fusion proteins (importins 7, 8, 13, beta1, alpha1, alpha3, alpha5, alpha6, alpha7 and mutant alpha1). Using a GST-pulldown assay, we found that ARHI can bind strongly to most importins; however, its binding is markedly reduced with an importin alpha1 mutant that contains an altered NLS (nuclear-localization signal) domain. In addition, an ARHI N-terminal deletion mutant exhibits greatly reduced binding to all importins compared with wild-type ARHI. In nuclear-import assays, the addition of ARHI blocked nuclear localization of phosphorylated STAT3. ARHI also inhibits the interaction of Ran-importin complexes with GFP (green fluorescent protein) fusion proteins that contain an NLS domain and a beta-like import receptor-binding domain, thereby blocking their nuclear localization. By conducting GST-pulldown assays, we found that ARHI could compete for Ran-importin binding. Thus ARHI-induced disruption of importin-binding to cargo proteins, including STAT3, could serve as an important regulatory mechanism that contributes to the tumour-suppressor function of ARHI.

ARHI (DIRAS3) is an imprinted tumor suppressor gene whose expression is lost in the majority of breast and ovarian cancers. Unlike its homologs Ras and Rap, ARHI functions as a tumor suppressor. Our previous study showed that ARHI can interact with transcription activator Stat3 and inhibit its nuclear translocation in human breast and ovarian cancer cells. To identify proteins that interact with ARHI in nuclear translocation, we have performed proteomic analysis and identified several importins that can associate with ARHI. To further explore this novel finding, we have purified 10 GST-importin fusion proteins (importin 7, 8, 13, b1, a1, a3, a5, a6, a7 as well as mutant a1). Using a GST-pull down assay, we found that ARHI can bind strongly to most importins; however, its binding is significantly reduced with an importin a1 mutant which contains an altered nuclear localization signal (NLS) domain. In addition, an ARHI N-terminal deletion mutant (NTD) exhibits much less binding to all importins than does wild type ARHI ARHI and NTD proteins were purified and tested for their ability to inhibit nuclear importation of proteins in HeLa cells. ARHI protein inhibits interaction of Ran-importin complexes with GFP fusion proteins that contain an NLS domain and a beta-like import receptor binding domain, blocking their nuclear localization. Addition of ARHI also blocked nuclear localization of phosphorylated Stat3&x03B2;. By GST-pull down assays, we found that ARHI could compete for Ran-importins binding. Thus, ARHI-induced disruption of importin binding to cargo proteins including Stat3 could serve as an important regulatory mechanism that contributes to the tumor suppressor function of ARHI.