Sai-Ming Ngai

Computational Identification of Protein Methylation Sites through Bi-Profile Bayes Feature Extraction

Protein methylation is one type of reversible post-translational modifications (PTMs), which plays vital roles in many cellular processes such as transcription activity, DNA repair. Experimental identification of methylation sites on proteins without prior knowledge is costly and time-consuming. In silico prediction of methylation sites might not only provide researches with information on the candidate sites for further determination, but also facilitate to perform downstream characterizations and site-specific investigations. In the present study, a novel approach based on Bi-profile Bayes feature extraction combined with support vector machines (SVMs) was employed to develop the model for Prediction of Protein Methylation Sites (BPB-PPMS) from primary sequence. Methylation can occur at many residues including arginine, lysine, histidine, glutamine, and proline. For the present, BPB-PPMS is only designed to predict the methylation status for lysine and arginine residues on polypeptides due to the absence of enough experimentally verified data to build and train prediction models for other residues. The performance of BPB-PPMS is measured with a sensitivity of 74.71%, a specificity of 94.32% and an accuracy of 87.98% for arginine as well as a sensitivity of 70.05%, a specificity of 77.08% and an accuracy of 75.51% for lysine in 5-fold cross validation experiments. Results obtained from cross-validation experiments and test on independent data sets suggest that BPB-PPMS presented here might facilitate the identification and annotation of protein methylation. Besides, BPB-PPMS can be extended to build predictors for other types of PTM sites with ease. For public access, BPB-PPMS is available at http://www.bioinfo.bio.cuhk.edu.hk/bpbppms.

Comparative proteomic analysis of mesenchymal stem cells derived from human bone marrow, umbilical cord, and placenta: Implication in the migration

Umbilical cord (UC) and placenta (P) have been suggested as alternatives to bone marrow (BM) as sources of mesenchymal stem cells (MSC) for cell therapy, with both UC‐ and P‐MSC possess immunophenotypic and functional characteristics similar to BM‐MSC. However, their migration capacity, which is indispensable during tissue regeneration process, is unclear. Under defined conditions, the migration capacity of BM‐ and P‐MSC was found 5.9‐ and 3.2‐folds higher than that of UC‐MSC, respectively. By the use of 2‐DE and combined MS and MS/MS analysis, six differentially expressed proteins were identified among these MSC samples, with five of them known to be involved in cell migration as migration enhancing or inhibiting proteins. Consistent with their migration capacity, the levels of migration enhancing proteins including cathepsin B, cathepsin D and prohibitin,were significantly lower in UC‐MSC when compared with those in BM‐ and P‐MSC. For the migration inhibiting proteins such as plasminogen activator inhibitor‐1 (PAI‐1) and manganese superoxide dismutase, higher expression was found in the UC‐MSC. We also showed that the overexpression of the PAI‐1 impaired the migration capacity of BM‐ and P‐MSC while silencing of PAI‐1 enhanced the migration capacity of UC‐MSC. Our study indicates that PAI‐1 and other migration‐related proteins are pivotal in governing the migration capacity of MSC.

Proteomic identification of malignant transformation-related proteins in esophageal squamous cell carcinoma

Esophageal cancer (EC) persists to be a leading cancer‐related death in northern China. Clinical outcome of EC is the most dismal among many types of digestive tumors because EC at early stage is asymptomatic. The current study used 2‐DE‐based proteomics to identify differentially expressed proteins between esophageal cancer cell lines and immortal cell line. Fifteen proteins were identified with differences of more than five folds, comprising the down‐regulation of annexin A2, histone deacetylase 10 isoform beta and protein disulfide‐isomerase ER‐60 precursor, and the up‐regulation of heat shock 70 kDa protein 9B precursor, solute carrier family 44 Member 3, heterogeneous nuclear ribonucleoprotein L (hnRNP L), eukaryotic translation initiation factor 4A isoform 2, triosephosphate isomerase1 (TPI), peroxiredoxin1 (PRX1), forminotransferase cyclodeaminase form (FTCD), fibrinogen gamma‐A chain precursor, kinesin‐like DNA binding protein, lamin A/C, cyclophilin A (CypA), and transcription factor MTSG1. Expression pattern of annexin A2 was verified by Western blotting, immunocytochemistry and immunohistochemistry analysis. The implication of these protein alterations correlated to the esophageal malignant transformation is discussed. J. Cell. Biochem. 104: 1625–1635, 2008.

Hepatocellular carcinoma-derived exosomes promote motility of immortalized hepatocyte through transfer of oncogenic proteins and RNAs

Exosomes are increasingly recognized as important mediators of cell-cell communication in cancer progression through the horizontal transfer of RNAs and proteins to neighboring or distant cells. Hepatocellular carcinoma (HCC) is a highly malignant cancer, whose metastasis is largely influenced by the tumor microenvironment. The possible role of exosomes in the interactions between HCC tumor cell and its surrounding hepatic milieu are however largely unknown. In this study, we comprehensively characterized the exosomal RNA and proteome contents derived from three HCC cell lines (HKCI-C3, HKCI-8 and MHCC97L) and an immortalized hepatocyte line (MIHA) using Ion Torrent sequencing and mass spectrometry, respectively. RNA deep sequencing and proteomic analysis revealed exosomes derived from metastatic HCC cell lines carried a large number of protumorigenic RNAs and proteins, such as MET protooncogene, S100 family members and the caveolins. Of interest, we found that exosomes from motile HCC cell lines could significantly enhance the migratory and invasive abilities of non-motile MIHA cell. We further demonstrated that uptake of these shuttled molecules could trigger PI3K/AKT and MAPK signaling pathways in MIHA with increased secretion of active MMP-2 and MMP-9. Our study showed for the first time that HCC-derived exosomes could mobilize normal hepatocyte, which may have implication in facilitating the protrusive activity of HCC cells through liver parenchyma during the process of metastasis.

Enhancement of Auranofin-Induced Apoptosis in MCF-7 Human Breast Cells by Selenocystine, a Synergistic Inhibitor of Thioredoxin Reductase

Thioredoxin system plays an important role in regulation of intracellular redox balance and various signaling pathways. Thioredoxin reductase (TrxR) is overexpressed in many cancer cells and has been identified as a potential target of anticancer drugs. Auranofin (AF) is potent TrxR inhibitor with novel in vitro and in vivo anticancer activities. Selenocystine (SeC) is a nutritionally available selenoamino acid with selective anticancer effects through induction of apoptosis. In the present study, we demonstrated the synergistic effects and the underlying molecular mechanisms of SeC in combination with AF on MCF-7 human breast cancer cells. The results showed that SeC and AF synergistically inhibited the cancer cell growth through induction of ROS-dependent apoptosis with the involvement of mitochondrial dysfunction. DNA damage-mediated p53 phosphorylation and down-regulation of phosphorylated AKT and ERK also contributed to cell apoptosis. Moreover, we demonstrated the important role of TrxR activity in the synergistic action of SeC and AF. Taken together, our results suggest the strategy to use SeC and AF in combination could be a highly efficient way to achieve anticancer synergism by targeting TrxR.

Interaction of hCLIM1, an enigma family protein, with α‐actinin 2

Enigma proteins are proteins that possess a PDZ domain at the amino terminal and one to three LIM domains at the carboxyl terminal. They are cytoplasmic proteins that are involved with the cytoskeleton and signal transduction pathway. By virtue of the two protein interacting domains, they are capable of protein‐protein interactions. Here we report a study on a human Enigma protein hCLIM1, in particular. Our study describes the interaction of the human 36kDa carboxyl terminal LIM domain protein (hCLIM1), the human homologue of CLP36 in rat, with α‐actinin 2, the skeletal muscle isoform of α‐actinin. hCLIM1 protein was shown to interact with α‐actinin 2 by yeast two‐hybrid screening and immunochemical analyses. Yeast two‐hybrid analyses also demonstrated that the LIM domain of hCLIM1 binds to the EF‐hand region of α‐actinin 2, defining a new mode of LIM domain interactions. Immunofluorescent study demonstrates that hCLIM1 colocalizes with α‐actinin at the Z‐disks in human myocardium. Taken together, our experimental results suggest that hCLIM1is a novel cytoskeletal protein and may act as an adapter that brings other proteins to the cytoskeleton. J. Cell. Biochem. 78:558–565, 2000.

Developmental regulation and cellular distribution of human cytosolic malate dehydrogenase (MDH1)

Human cyotsolic malate dehydrogenase (MDH1) is important in transporting NADH equivalents across the mitochondrial membrane, controlling tricarboxylic acid (TCA) cycle pool size and providing contractile function. Cellular localization studies indicate that MDH1 mRNA expression has a strong tissue‐specific distribution, being expressed primarily in cardiac and skeletal muscle and in the brain, at intermediate levels in the spleen, kidney, intestine, liver, and testes and at low levels in lung and bone marrow. The observed MDH1 localizations reflect the role of NADH in the support of a variety of functions in different organs. These functions are primarily related to aerobic energy production for muscle contraction, neuronal signal transmission, absorption/resorption functions, collagen‐supporting functions, phagocytosis of dead cells, and processes related to gas exchange and cell division. During neonatal development, MDH1 is expressed in human embryonic heart as early as the 3rd month and then is over‐expressed from the 5th month until the birth. The expression of MDH1 is maintained in the adult heart but is not present in levels as high as in the fetus. Finally, over‐expression of MDH1 is found in left ventricular cardiac muscle of dilated cardiomyopathy (DCM) patients when contrasted to the diseased non‐DCM and normal heart muscle by in situ hybridization and Western blot. These observations are compatible with the activation of glucose oxidation in relatively hypoxic environments of fetal and hypertrophied myocardium.

Identification of histone methylation multiplicities patterns in the brain of senescence-accelerated prone mouse 8

Histone post-translational modifications (PTMs) are involved in diverse biological processes and methylation was regarded as a long-term epigenetic mark. Though aging represented one of the major risk factors for neurodegenerative diseases, no systematic investigations had correlated the patterns of histone PTMs in the brain with aging and the roles of such concerted histone PTMs in brain aging are still unknown. In this study, enzyme digestion, nano-LC, MALDI-TOF/TOF MS analysis and Western blotting were combined to investigate the defined methylation of core histones (H2A, H2B, H3 and H4) in the brain of 12-month-old senescence accelerated mouse prone 8 (SAMP8). The expression of several modified histones in the brain of 3-, and 12-month-old SAMP8 mice as well as that of the age-matched control senescence accelerated-resistant mouse (SAMR1) was compared. In the brain of 12-month-old SAMP8 mice, seven methylation sites (H3K24, H3K27, H3K36, H3K79, H3R128, H4K20 and H2A R89) were detected and most PTMs sites were located on histone H3. Mono-methylated H4K20 decreased significantly in the brain of 12-month-old SAMP8 mice. Methylated H3K27 and H3K36 coexisted in the aged brain with different methylation multiplicities. Di-methylated H3K79 expressed in the neurons of cerebral cortex and hippocampus. This study showed histone methylation patterns in the aged SAMP8 mice brain and provided the experimental evidences for further research on histone PTMs in the aged brain. We hope these results could initiate a platform for the exchange of comprehensive information concerning aging or neurodegenerative disease and help us interpret the change of gene expression and DNA repair ability at epigenetic level.

The ribosomal protein S26 regulates p53 activity in response to DNA damage

Ribosomal proteins have emerged as novel regulators of the Mdm2-p53 feedback loop, especially in the context of ribosomal stress. RPS26 is a recently identified Diamond-Blackfan Anemia-related ribosomal protein and its role in p53 activation has not been previously explored. In this study we found knockdown of RPS26 induced p53 stabilization and activation via a RPL11-dependent mechanism, resulting in p53-dependent cell growth inhibition. Moreover, RPS26 has the ability to interact with Mdm2 and inhibits Mdm2-mediated p53 ubiquitination that leads to p53 stabilization upon overexpression. Importantly, we discovered that RPS26 knockdown impaired p53’s ability to transcriptionally activate its target genes in response to DNA damage, without affecting its stability. Accordingly, the cells lost the ability to induce G2/M cell cycle arrest. We further found that upon RPS26 knockdown, the DNA damage induced recruitment of p53 to the promoters of its target genes and p53 acetylation were both greatly reduced. In addition, RPS26 can interact with p53 independent of Mdm2 and coexist in a complex with p53 and p300. These data establish a role of RPS26 in DNA damage response by directly influencing p53 transcriptional activity, and suggest that RPS26 acts distinctively in different scenarios of p53 activation. Our finding also implicates p53 transcriptional activity control as an important mechanism of p53 regulation by ribosomal proteins.

Characterization of tissue‐specific LIM domain protein (FHL1C) which is an alternatively spliced isoform of a human LIM‐only protein (FHL1)

We have cloned and characterized another alternatively spliced isoform of the human four‐and‐a‐half LIM domain protein 1 (FHL1), designated FHL1C. FHL1C contains a single zinc finger and two tandem repeats of LIM domains at the N‐terminus followed by a putative RBP‐J binding region at the C‐terminus. FHL1C shares the same N‐terminal two‐and‐a‐half LIM domains with FHL1 but different C‐terminal protein sequences. Due to the absence of the exon 4 in FHL1C, there is a frame‐shift in the 3′ coding region. Sequence analysis indicated that FHL1C is the human homolog of murine KyoT2. The Northern blot and RT‐PCR results revealed that FHL1 is widely expressed in human tissues, including skeletal muscle and heart at a high level, albeit as a relatively low abundance transcript in brain, placenta, lung, liver, kidney, pancreas, and testis. In contrast, FHL1C is specifically expressed in testis, skeletal muscle, and heart at a relatively low level compared with FHL1. The expression of FHL1C transcripts was also seen in aorta, left atrium, left, and right ventricles of human heart at low level. Immunoblot analysis using affinity‐purified anti‐FHL1C antipeptide antibodies confirmed a 20 kDa protein of FHL1C in human skeletal muscle and heart. Unlike FHL1B, which is another FHL1 isoform recently reported by our group and localized predominantly in the nucleus [Lee et al., 1999 ], FHL1C is localized both in the nucleus and cytoplasm of mammalian cell. J. Cell. Biochem. 82: 1–10, 2001.