Oxidative stress promotes liver fibrosis by modulating the microRNA-144 and SIN3A-p38 pathways in hepatic stellate cells.

Background & Aims: Reactive oxygen species (ROS) act as modulators triggering cellular dysfunctions and organ damage including liver fibrosis in which hepatic stellate cell (HSC) activation plays a key role. Previous studies suggest that microRNA-144 (miR-144) acts as a pro-oxidant molecule; however, whether and how miR-144 affects HSC activation and liver fibrosis remain unknown. Methods: Carbon tetrachloride (CCl4) and bile duct ligation (BDL)-induced experimental liver fibrosis models were used. Hepatic miR-144 expression was analyzed by miRNA in situ hybridization with RNAscope probe. The in vivo effects of silencing or overexpressing miR-144 were examined with an adeno-associated virus 6 (AAV6) carrying miR-144 inhibitor or mimics in fibrotic mouse experimental models. Results: In this study, we demonstrated that ROS treatment significantly upregulated miR-144 in HSCs, which further promoted HSC activation in vitro. Interestingly, miR-144 was preferentially elevated in HSCs of experimental liver fibrosis in mice and in human liver fibrotic tissues. Furthermore, in vivo loss or gain-of-function experiments via AAV6 carrying miR-144 antagomir or agomir revealed that blockade of miR-144 in HSCs mitigated, while overexpression of miR-144 in HSCs accelerated the development of experimental liver fibrosis. Mechanistically, SIN3 transcription regulator family member A (SIN3A), a transcriptional repressor, was identified to be the target of miR-144 in HSCs. MiR-144 downregulated Sin3A, and in line with this result, specific knockdown of Sin3a in HSCs remarkedly activated p38 MAPK signaling pathway to promote HSC activation, eventually exacerbating liver fibrosis. Conclusions: Oxidative stress-driven miR-144 fuels HSC activation and liver fibrogenesis by limiting the SIN3A-p38 axis. Thus, a specific inhibition of miR-144 in HSCs could be a novel therapeutic strategy for the treatment of liver fibrosis.

Development of a novel fluorescent protein-based probe for efficient detection of Pb2+ in serum inspired by the metalloregulatory protein PbrR691.

BACKGROUND: The accurate and rapid detection of blood lead concentration is of paramount importance for assessing human lead exposure levels. Fluorescent protein-based probes, known for their high detection capabilities and low toxicity, are extensively used in analytical sciences. However, there is currently a shortage of such probes designed for ultrasensitive detection of Pb2+, and no reported probes exist for the quantitative detection of Pb2+ in blood samples. This study aims to fill this critical void by developing and evaluating a novel fluorescent protein-based probe that promises accurate and rapid lead quantification in blood. RESULTS: A simple and small-molecule fluorescent protein-based probe was successfully constructed herein using a peptide PbrBD designed for Pb2+ recognition coupled to a single fluorescent protein, sfGFP. The probe retains a three-coordinate configuration to identify Pb2+ and has a high affinity for it with a Kd' of 1.48 ± 0.05 × 10-17 M. It effectively transfers the conformational changes of the peptide to the chromophore upon Pb2+ binding, leading to fast fluorescence quenching and a sensitive response to Pb2+. The probe offers a broad dynamic response range of approximately 37-fold and a linear detection range from 0.25 nM to 3500 nM. More importantly, the probe can resist interference of metal ions in living organisms, enabling quantitative analysis of Pb2+ in the picomolar to millimolar range in serum samples with a recovery percentage of 96.64%-108.74 %. SIGNIFICANCE: This innovative probe, the first to employ a single fluorescent protein-based probe for ultrasensitive and precise analysis of Pb2+ in animal and human serum, heralds a significant advancement in environmental monitoring and public health surveillance. Furthermore, as a genetically encoded fluorescent probe, this probe also holds potential for the in vivo localization and concentration monitoring of Pb2+.

LSAM: L2-norm self-attention and latent space feature interaction for automatic 3D multi-modal head and neck tumor segmentation.

Objective.Head and neck (H&N) cancers are prevalent globally, and early and accurate detection is absolutely crucial for timely and effective treatment. However, the segmentation of H&N tumors is challenging due to the similar density of the tumors and surrounding tissues in CT images. While positron emission computed tomography (PET) images provide information about the metabolic activity of the tissue and can distinguish between lesion regions and normal tissue. But they are limited by their low spatial resolution. To fully leverage the complementary information from PET and CT images, we propose a novel and innovative multi-modal tumor segmentation method specifically designed for H&N tumor segmentation.Approach.The proposed novel and innovative multi-modal tumor segmentation network (LSAM) consists of two key learning modules, namely L2-Norm self-attention and latent space feature interaction, which exploit the high sensitivity of PET images and the anatomical information of CT images. These two advanced modules contribute to a powerful 3D segmentation network based on a U-shaped structure. The well-designed segmentation method can integrate complementary features from different modalities at multiple scales, thereby improving the feature interaction between modalities.Main results.We evaluated the proposed method on the public HECKTOR PET-CT dataset, and the experimental results demonstrate that the proposed method convincingly outperforms existing H&N tumor segmentation methods in terms of key evaluation metrics, including DSC (0.8457), Jaccard (0.7756), RVD (0.0938), and HD95 (11.75).Significance.The innovative Self-Attention mechanism based on L2-Norm offers scalability and is effective in reducing the impact of outliers on the performance of the model. And the novel method for multi-scale feature interaction based on Latent Space utilizes the learning process in the encoder phase to achieve the best complementary effects among different modalities.

Molecular basis for substrate recognition and transport of human GABA transporter GAT1.

γ-Aminobutyric acid (GABA), an important inhibitory neurotransmitter in the central nervous system, is recycled through specific GABA transporters (GATs). GAT1, which is mainly expressed in the presynaptic terminals of axons, is a potential drug target of neurological disorders due to its essential role in GABA transport. Here we report four cryogenic electron microscopy structures of human GAT1, at resolutions of 2.2-3.2 Å. GAT1 in substrate-free form or in complex with the antiepileptic drug tiagabine exhibits an inward-open conformation. In the presence of GABA or nipecotic acid, inward-occluded structures are captured. The GABA-bound structure reveals an interaction network bridged by hydrogen bonds and ion coordination for GABA recognition. The substrate-free structure unwinds the last helical turn of transmembrane helix TM1a to release sodium ions and substrate. Complemented by structure-guided biochemical analyses, our studies reveal detailed mechanism of GABA recognition and transport, and elucidate mode of action of the inhibitors, nipecotic acid and tiagabine.

A novel strategy of engineering genetically encoded probe for ultrasensitive sensing Hg2+ with unusual planar trigonometric coordination configuration.

At present, few genetically encoded fluorescent probes are currently available for the analysis of toxic heavy metal ions, and most have poor performance that cannot meet the requirements of sensitive and dynamic detection in living cells. In this study, we designed a single fluorescent protein-based probe sfGFP-MerBD, which can specifically response to Hg2+ with high binding affinity and wide dynamic range. More importantly, the developing probe can timely and reversibly monitor changes of Hg2+ concentration in living mammalian cells. The excellent performance of this probe is largely due to the recognition element of the probe, MerBD, which adopts an unusual planar trigonometric coordination configuration with Hg2+, and the coordination can cause enough conformational change to influence the fluorescence of skeleton protein sfGFP coupled with it. The small peptide MerBD was delicately designed based on the three-dimensional structure of metalloprotein MerR. This novel design strategy solves the challenging problems that there are few natural functional proteins in the process of constructing fluorescent probes for toxic metal ions and some functional proteins cannot be directly used as recognition elements. Based on the new strategy, more genetically encoded fluorescent probes of toxic heavy metal ions can be efficiently constructed and applied in the future.

Automated multi-modal Transformer network (AMTNet) for 3D medical images segmentation.

Objective.Over the past years, convolutional neural networks based methods have dominated the field of medical image segmentation. But the main drawback of these methods is that they have difficulty representing long-range dependencies. Recently, the Transformer has demonstrated super performance in computer vision and has also been successfully applied to medical image segmentation because of the self-attention mechanism and long-range dependencies encoding on images. To the best of our knowledge, only a few works focus on cross-modalities of image segmentation using the Transformer. Hence, the main objective of this study was to design, propose and validate a deep learning method to extend the application of Transformer to multi-modality medical image segmentation.Approach.This paper proposes a novel automated multi-modal Transformer network termed AMTNet for 3D medical image segmentation. Especially, the network is a well-modeled U-shaped network architecture where many effective and significant changes have been made in the feature encoding, fusion, and decoding parts. The encoding part comprises 3D embedding, 3D multi-modal Transformer, and 3D Co-learn down-sampling blocks. Symmetrically, the 3D Transformer block, upsampling block, and 3D-expanding blocks are included in the decoding part. In addition, a Transformer-based adaptive channel interleaved Transformer feature fusion module is designed to fully fuse features of different modalities.Main results.We provide a comprehensive experimental analysis of the Prostate and BraTS2021 datasets. The results show that our method achieves an average DSC of 0.907 and 0.851 (0.734 for ET, 0.895 for TC, and 0.924 for WT) on these two datasets, respectively. These values show that AMTNet yielded significant improvements over the state-of-the-art segmentation networks.Significance.The proposed 3D segmentation network exploits complementary features of different modalities during the feature extraction process at multiple scales to increase the 3D feature representations and improve the segmentation efficiency. This powerful network enriches the research of the Transformer to multi-modal medical image segmentation.

New Fructokinase, OsFRK3, Regulates Starch Accumulation and Grain Filling in Rice.

Plant fructokinase (FRK) guarantees the growth and development of higher plants by participating in carbohydrate metabolism. In this study, a new fructokinase, OsFRK3, was identified using bioinformatics analysis, enzyme assay, bacterial growth assay, and yeast complementation test. Then, we created OsFRK3 knockout transgenic lines (osfrk3-1 and osfrk3-2) by the CRISPR/Cas9 technology. We found that the 1000-grain weight decreased notably (approximately -3.6% and -6.1%, respectively) in osfrk3-1 and osfrk3-2. Evidently decreased grain width, grain thickness, and endosperm filling rate were detected in the osfrk3 mutants (osfrk3-1 and osfrk3-2) compared with those of the WT. In addition, the content of seed total starch was significantly decreased by 3.42 and 4.80% in osfrk3 lines, compared with that in the WT. The level of maltose was significantly reduced in the mutants, while that of sucrose and fructose was obviously increased in the mutants. The transcript levels of OsGBSS1, OsBEIIb, OsPFP1ß, and OsAGPL1 were significantly decreased in the osfrk3 mutants. These results suggest that OsFRK3 may positively regulate the accumulation of starch through influencing the sugar metabolism.

Atractylenolide III ameliorates Non-Alcoholic Fatty Liver Disease by activating Hepatic Adiponectin Receptor 1-Mediated AMPK Pathway.

Background: Nonalcoholic fatty liver disease (NAFLD) is the most frequent cause of chronic liver diseases worldwide. At present, there are no effective pharmacological therapies for NAFLD except lifestyle intervention-mediated weight loss. Atractylenolide III (ATL III), the major bioactive component found in Atractylode smacrocephala Koidz, has been shown to exert anti-oxidant, anti-tumor, anti-allergic response, anti-bacterial effects and cognitive protection. Here we investigate the therapeutic potential and underlying mechanisms of ATL III for the treatment of NAFLD. Methods: Male C57BL/6J mice were fed a high-fat diet (HFD) and treated with ATL III. Lipid accumulation was analyzed by Oil Red O staining in liver tissues and free fatty acids (FFAs)-treated hepatocytes. AMP-activated protein (AMPK) and sirtuin 1(SIRT1) signaling pathways were inhibited by Compound C and EX527 in vitro, respectively. Small-interfering RNA (siRNA) was used to knockdown adiponectin receptor 1 (AdipoR1) expression in HepG2 cells. Results: ATL III treatment ameliorated liver injury and hepatic lipid accumulation in the HFD-induced NAFLD mouse model as demonstrated by that ATL III administration significantly reduced serum levels of alanine aminotransferase, glutamic oxaloacetic transaminase, triglycerides, total cholesterol and low-density lipoprotein. Furthermore, treatment with ATL III alleviated hepatic oxidative stress, inflammation and fibrosis in the HFD feeding model. To study the underlying mechanisms, we performed Computer Aided Design assay and found that open-formed AdipoR1 and adiponectin receptor 2 were the potential receptors targeted by ATL III. Interestingly, HFD feeding or FFAs treatment only reduced hepatic AdipoR1 expression, while such reduction was abolished by ATL III administration. In addition, in vitro treatment with ATL III activated the AdipoR1 downstream AMPK /SIRT1 signaling pathway and reduced lipid deposition in HepG2 cells, which was diminished by silencing AdipoR1. Finally, inhibition of AMPK or SIRT1, the AdipoR1 downstream signaling, abolished the protective effects of ATL III on lipid deposition and oxidative stress in FFAs-treated HepG2 cells. Conclusion: Our findings suggest that ATL III is a therapeutic drug for the treatment of NAFLD and such protective effect is mediated by activating hepatic AdipoR1-mediated AMPK/SIRT1 signaling pathway.

Cross-sectional study of the educational background and trauma knowledge of trainees in the "China trauma care training" program.

BACKGROUND: Since the trauma knowledge of trauma providers correlates with the outcomes of injured patients, this study aims to assess the socio-demographic characteristics and levels of trauma knowledge of trainees in the China trauma care training (CTCT) program in addition to their post-course test results to provide support for the development of trauma care training programs and trauma systems in China. METHODS: A cross-sectional study was conducted by collecting demographic information, hospital-related information and trauma knowledge of the trainees from 19 regions in China. All participants were assessed by questionnaires collecting the socio-demographic data, the trauma care knowledge levels and the information of the hospitals. RESULTS: There were 955 males (78.9%) and 256 females (21.1%) enrolled. Among them, 854 were physicians (70.5%), 357 were registered nurses (29.5%). In addition, 64 of them also played an administrative role in the hospitals (5.3%). The score of the trainees who were members of the emergency department staff (72.59 ± 14.13) was the highest among the scores of all the personnel surveyed, followed by those of the trainees from the intensive care unit (ICU) (71.17 ± 12.72), trauma surgery department (67.26 ± 13.81), orthopedics department (70.36 ± 14.48), general surgery department (69.91 ± 14.79) and other departments (69.93 ± 16.91), P = 0.031. The score of the professors (73.09 ± 15.05) was higher than those of the associate professors (72.40 ± 14.71), lecturers (70.07 ± 14.25) and teaching assistants (67.58 ± 15.16), P < 0.0001. The score of the individuals who attended experts' trauma lectures (72.22 ± 14.45) was higher than that of individuals who did not attend the lectures (69.33 ± 15.17), P = 0.001. The mean scores before and after the training were 71.02 ± 14.82 and 84.24 ± 13.77, respectively, P < 0.001. The mean score of trauma knowledge after the training of trainees from different provinces and with different educational backgrounds was higher than that before the training, with a statistically significant difference (P < 0.05). CONCLUSIONS: The level of trauma knowledge of trauma care providers was associated with their department, professional position and previous participation in related academic conferences. Trauma care experience and participation in academic lectures and training program including CTCT may effectively improve individuals' level of trauma knowledge.

Correction: Wentao Wu, et al. Evolution Analysis of the Aux/IAA Gene Family in Plants Shows Dual Origins and Variable Nuclear Localization Signals. Int. J. Mol. Sci. 2017, 18, 2107.

The authors would like to insert some websites and citations in the following sentence, "First, the complete proteomes of these species were downloaded from the Phytozome website (Version 11; Available online: www.phytozome.org)" in the "Materials and Methods" section on page 13, paragraph 3.1 of their paper published in the International Journal of Molecular Sciences [1].[...].

Evolution Analysis of the Aux/IAA Gene Family in Plants Shows Dual Origins and Variable Nuclear Localization Signals.

The plant hormone auxin plays pivotal roles in many aspects of plant growth and development. The auxin/indole-3-acetic acid (Aux/IAA) gene family encodes short-lived nuclear proteins acting on auxin perception and signaling, but the evolutionary history of this gene family remains to be elucidated. In this study, the Aux/IAA gene family in 17 plant species covering all major lineages of plants is identified and analyzed by using multiple bioinformatics methods. A total of 434 Aux/IAA genes was found among these plant species, and the gene copy number ranges from three (Physcomitrella patens) to 63 (Glycine max). The phylogenetic analysis shows that the canonical Aux/IAA proteins can be generally divided into five major clades, and the origin of Aux/IAA proteins could be traced back to the common ancestor of land plants and green algae. Many truncated Aux/IAA proteins were found, and some of these truncated Aux/IAA proteins may be generated from the C-terminal truncation of auxin response factor (ARF) proteins. Our results indicate that tandem and segmental duplications play dominant roles for the expansion of the Aux/IAA gene family mainly under purifying selection. The putative nuclear localization signals (NLSs) in Aux/IAA proteins are conservative, and two kinds of new primordial bipartite NLSs in P. patens and Selaginella moellendorffii were discovered. Our findings not only give insights into the origin and expansion of the Aux/IAA gene family, but also provide a basis for understanding their functions during the course of evolution.

MicroRNA-216b-5p Functions as a Tumor-suppressive RNA by Targeting TPT1 in Pancreatic Cancer Cells.

MicroRNAs (miRNAs) are increasingly recognized as being involved in pancreatic cancer progression by directly regulating the expression of their targets. In this study, we showed that miR-216b-5p expression was significantly decreased in pancreatic cancer tissues and cell lines. In addition, low miR-216b-5p expression was significantly associated with large tumor size and advanced TNM stage. Meanwhile, both Kaplan-Meier and multivariate survival analysis showed that decreased miR-216b-5p expression was associated with overall survival. miR-216b-5p over-expression repressed pancreatic cancer cell proliferation and induced cell cycle arrest and cell apoptosis in vitro and inhibited tumorigenesis in vivo. The translationally controlled tumor protein (TPT1) was identified as a novel direct target of miR-216b-5p. miR-216b-5p up-regulation suppressed TPT1 expression. Moreover, TPT1 mRNA expression levels were increased in pancreatic cancer tissues, and were inversely correlated with miR-216b-5p expression. TPT1 down-regulation had similar effects as miR-216b-5p up-regulation on pancreatic cancer cell progression. The restoration of TPT1 reversed the effect of miR-216b-5p on pancreatic cancer cell progression. Furthermore, we found that miR-216b-5p up-regulation suppressed Pim-3, Cyclin B1, p-Bad and Bcl-xL protein expression. However, the effect of miR-216b-5p up-regulation was partly reversed by TPT1 up-regulation in vitro. Taken together, our findings suggested that miR-216b-5p functions as a potential tumor suppressor by regulating TPT1 in pancreatic cancer cells, and it may represent a potential therapeutic target for patients with pancreatic cancer.

Association of elevated risk of pancreatic cancer in diabetic patients: A systematic review and meta-analysis.

Pancreatic cancer has a five-year overall survival rate <5%, a situation that has not improved since for 40 years. Diabetes mellitus including type 2 diabetes mellitus (T2D) is a suspected risk factor for the development of pancreatic cancer and nearly 45% of the pancreatic cancer cases are likely to present as new onset diabetes cases; however, the nature of association between T2D and pancreatic cancer is still controversial. In this meta-analysis, we examined the association specifically of T2D with pancreatic cancer and the influence of insulin therapy. PubMed, EMBASE, Scholar, Web of Science and Scopus databases were searched to identify clinical and patient oriented studies that examined the incidence of diabetes in pancreatic cancer patients and vice versa, over the last 10 years. All the authors independently screened the articles, and a collective decision was reached about the studies included in the meta-analysis. Parameters analyzed included, the Incidence of diabetes in pancreatic cancer patients; duration history of T2D in pancreatic cancer patients; influence of insulin therapy in T2D patients on pancreatic cancer incidence. Eleven studies with a total of 14,399 patients, of whom 4,080 were T2D-positive and 9,721 were non-diabetic were included in this meta-analysis. T2D duration history was significantly related to pancreatic cancer incidence and insulin therapy effects. In conclusion, recent-onset T2D is probably a manifestation of pancreatic cancer whereas long-term T2D is likely a risk factor for this cancer. Insulin therapy appears to decrease the incidence of pancreatic cancer.

MiRNA-22 inhibits oncogene galectin-1 in hepatocellular carcinoma.

Hepatic stellate cells (HSCs) induce immune privilege and promote hepatocellular carcinoma (HCC) by suppressing the immune system. On the other hand, galectin-1 and miRNA-22 (miR-22) are dysregulated in HCC and serve as prognostic indicators for patients. In this study, therefore, we measured galectin-1 and miR-22 expression in HSCs isolated from HCC tissues (Ca-HSCs), and in normal liver tissues (N-HSCs) as a control. We also investigated the apoptosis rate among T cells and the production of cytokines (IFN-γ and IL-10) in HSCs co-cultured with T cells. And we used immunohistochemical staining to tested for correlation between galectin-1 expression, CD3 expression and clinicopathological features in 162 HCC patients. Our results showed that galectin-1 expression was much higher in Ca-HSCs than in N-HSCs. Overexpression of galectin-1 promoted HSC-induced T cell apoptosis and cytokine production (IFN-γ and IL-10), while miR-22 expression inhibited it. Galectin-1 expression correlated negatively with miR-22 expression in HSCs. High galectin-1 and low CD3 expression levels were associated with poor prognosis in HCC patients. These results suggest that the immunosuppressive microenvironment promoted by HSC-derived galectin-1 in HCC can be inhibited by miR-22. Galectin-1 and miR-22 could potentially serve as prognostic markers and therapeutic targets in HCC.

Effects of Raf kinase inhibitor protein expression on pancreatic cancer cell growth and motility: an in vivo and in vitro study.

PURPOSE: Raf kinase inhibitor protein (RKIP) is a tumor suppressor that inhibits cell growth and metastasis of malignant tumors. Pancreatic cancer is a leading cause of cancer death with a low survival rate. RKIP expression and its role in tumorigenesis and metastasis in pancreatic cancer are poorly understood. The aims of our study were to assess the effects of RKIP on pancreatic carcinoma cells in vitro and in tumor tissues in vivo. METHODS: This study included 84 patients with histologically confirmed pancreatic adenocarcinoma. The expression levels of RKIP were measured in pancreatic cancer tissues and adjacent normal tissues using real-time PCR and immunohistochemistry. Overexpression plasmid of RKIP was transfected into SW1990 and AsPC-1 cell lines, and the effects on cell proliferation were studied using a Cell Counting Kit-8 assay. MEK1/2 and ERK1/2 were detected by Western blot and immunofluorescence assay. RESULTS: Results showed a reduced expression of RKIP in pancreatic carcinoma tissues compared with adjacent normal tissues, which closely correlated with patient outcomes. Overexpression of RKIP suppressed cell proliferation and promoted apoptosis in cultured SW1990 and AsPC-1 cell lines. Transwell assay showed RKIP can inhibit cell migration and invasion, and in vivo RKIP can suppress tumorigenesis by diminishing the volume of the tumors. CONCLUSIONS: In conclusion, expression of RKIP is closely correlated with the survival of pancreatic cancer patients. RKIP can inhibit pancreatic adenocarcinoma cells proliferation, activities of migration and invasion, through downregulating Raf-1-MEK1/2-ERK1/2 signaling pathway.

Par-4 downregulation confers cisplatin resistance in pancreatic cancer cells via PI3K/Akt pathway-dependent EMT.

Cisplatin (CDDP) efficiency in pancreatic cancer therapy is limited due to development of drug resistance. However, the comprehensive mechanisms remain largely unclear. In this study, we first established a CDDP-resistant pancreatic cancer cell line-BXPC-3/CDDP from its parental cell line-BXPC-3. The results showed that CDDP resistance in BXPC-3/CDDP cells correlates with changes in cellular EMT phenotypes. Prostate apoptosis response-4 (Par-4) expression at both mRNA and protein levels were reduced in CDDP-resistant BXPC-3/CDDP cells compared with that in BXPC-3 cells. Ectopic expression of Par-4 reversed EMT and CDDP resistance in BXPC-3/CDDP cells. In BXPC-3 cells, knockdown of Par-4 expression induces EMT and CDDP insensitivity, however, these effects were blocked by inhibition of PI3K/Akt pathway using LY294002. Furthermore, Par-4 knockdown could significantly stimulate PI3K/Akt signaling in BXPC-3 cells. In vivo studies, xenograft BXPC-3 tumors were sensitive to CDDP treatment. Treatment with CDDP alone had little effect on the growth of Par-4 siRNA-transfected BXPC-3 tumors in nude mice and the survival rate compared with control. Inhibition of PI3K/Akt pathway using LY294002 reversed CDDP resistance in Par-4 siRNA-transfected BXPC-3 tumors. In conclusion, these results indicate that Par-4 downregulation confers CDDP resistance via PI3K/Akt pathway-dependent EMT in BXPC-3 cells. Therefore, Par-4 may be a potential target for overcoming CDDP resistance in pancreatic cancer.