Past studies have established distinctive serum peptidome patterns through mass spectrometry that seem to correlate with clinically relevant outcomes. Wider acceptance of these patterns as valid biomarkers for disease required characterization of the individual components and elucidation of the mechanisms by which they are generated. Using an automated, high-throughput peptide extraction and MALDI-TOF MS-based approach, we have shown that a limited subset of serum peptides provides class discrimination between patients with three different types of solid tumors and control individuals without cancer. Sequence analysis revealed that these peptides sort into a small number of clusters, and that most are generated by exopeptidase activities that confer cancer-type-specific differences superimposed on the proteolytic events of the ex-vivo coagulation and complement degradation pathways. Serum peptide profiling by MALDI-TOF MS thus monitors blood proteome metabolomic products and the observed peptides are only surrogates for the real biomarkers: exopeptidases, or better, exopeptidase activities. Our peptidome analysis platform is therefore particularly well suited for detection of cancer as proteases are established components of cancer progression and invasiveness. Building on these observations, we have developed quantitative blood-exopeptidase assays as a novel tool for cancer detection. The prototype test tracks degradation of artificial peptide substrates, under strictly controlled conditions, using the established automated extraction / MS-analysis method to monitor resulting patterns. Each fragment is quantitated by comparison with double-labeled, non-degradable internal standards (synsthesized of all-D-amino acids). The full array of metabolites ('mini-peptidome') is then subjected to multivariate statistical analysis. Using this approach, we have analyzed serum samples of cancer patients and healthy controls and obtained class predictions with 94% sensitivity and 90% specificity. The test effectively bypasses reproducibility problems related to sample collection, storage and handling, as well as with potential variability in endogenous peptide precursor levels. It may lead to a fundamentally new approach to biomarker development.

View full abstract

A complex network of cell-cell communication system by peptide hormones works for maintaining the mammalian homeostatic balance. For further understanding the intricate mechanisms of the regulation, it is important to discover unidentified bioactive peptides. By using our own methods, we have been searching for peptides which may be relevant to circulation control. In the course of these studies, we discovered 3 natriuretic peptides, ANP (1984), BNP (1988) and CNP (1990), and also adrenomedullin (1993). These studies elucidated new regulatory mechanisms of cardiovascular system, and also brought their therapeutic potential on cardiovascular diseases. In 1999, we discovered ghrelin, a novel growth hormone (GH)-releasing peptide as an endogenous ligand for GHS-R, from rat stomach. Ghrelin was a 28-amino acid peptide with a marvelous structure, in which the Ser-3 residue was n-octanoylated. This n-octanoyl modification was essential for the activity. Ghrelin is primarily produced in distinct endocrine cells in the stomach. Ghrelin-producing neurons are also present in the hypothalamic arcuate nucleus, a region that regulates GH release and food intake. In fact, ghrelin stimulates GH release and feeding when administered centrally or peripherally. Beside these effects, ghrelin is also involved in the regulation of cardiovascular system. In the clinical trial, 3-week administration of ghrelin improved left ventricular structure and function, exercise capacity, and muscle wasting in patients with chronic heart failure (CHF). The occurrence of ghrelin in both stomach and hypothalamus will give a new dimension to the regulation of GH release and feeding. Further, GHS-R is widely expressed in peripheral tissues. Ghrelin also thus has multifaceted roles in the cardiovascular and metabolism systems. In my presentation, I will discuss about therapeutic potential of endogenous bioactive peptides.

View full abstract

We received sad news on May 31st, 2007, that Dr. Akira Tsugita who had been playing a leading role of proteomics research globally since its very early stage passed away. Dr. Tsugita became a first Japanese committee member of the international HUPO, when it was established in 2002. At the same time, he established JHUPO and became a first Chairman of JHUPO. Since then, he had devoted much efforts to the development of the JHUPO. It is attributed to his efforts that the JHUPO has evolved such that we can hold a successful 6th JHUPO conference this year. The Tsugita Memorial Lectures are organized in this JHUPO conference to appreciate his great contribution to the JHUPO. In this Lectures, Dr.Hosokawa who had conducted many collaborative works with Dr. Tsugita will talk about his footprint from the early days of molecular biology to the present days through the evolution of the proteomics. Drs. Kunisawa and Takamoto, both of whom studied molecular biology under him will talk about the front-end of the bio database research and the structure proteomics.

View full abstract

Prof. Akira Tsugita deceased in May 2007. He made a number of invaluable contributions to bioscience. We would like to recall his footsteps from the dawn of molecular biology to the contemporary genomics and proteomics. No sooner he earned doctorate in 1958 under Prof. S. Akabori, Osaka Univ., than Tsugita (title omitted hereafter) visited Prof. H. Fraenkel-Conrat, Univ. of Calif., Barkeley, and made complete amino acid sequencing of TMV coat protein, the world second achievement next to Sanger's insulin. From amino acid changes in protein sequence of mutant TMV and later T4 phage, he shed light on genetic informtion transfer and coding. In 1961, upon returning to Osaka, Tsugita developed his academic career on the background of protein chemistry, beginning as Researchfellow for Prof. Akabori, Institute for Protein Research, Osaka Univ., and promoted to full Prof., Osaka Univ. Med. School and spent productive period of time as Dir., Research Institute for Molecular Genetics of the same campus. Between 1972 and '78, while visiting Biozentrum, Univ. of Basel, Switzerland, he discovered a proteinase specific to HIV, which led to a remedy for AIDS, blocking viral maturation by inhibiting the enzyme. In 1978, he was appointed to Member and Prof., EMBL, Heidelberg, Germany and made very important contribution, that is, rapid amino acid analysis. In 1985, he returned back to Japan to take a position of Prof., Science University of Tokyo, meanwhile he organized international meeting of rice genome in Okayama, helped by Drs. Y. Nozu, K. Higo and myself. This triggered international collaboration of rice science research with Asian and Oceanian countries. He also contributed to establish a highly sensitive determination method for N-terminal amino acid of protein. Another important contribution is C-terminal amino acid determination, which is under way for practical use. A. Tsugita organized the 15th International Meeting in Tsukuba serving as Vice President of CODATA as a representative of Japan from 1998 to 2002. Together with Prof. K. Nakamura, he joined Human Proteomics Organization (HUPO) as organizing member of the Committee and participated in the start of Japanese HUPO (JHUPO).
The great footsteps Akira Tsugita left behind will definitely serve as strong impacts for the following generation in the 21st Century, and contribute to ever-growing science, for the welfare of humankind.

View full abstract

Besides his own research area, the late Professor Akira Tsugita played leading roles in the activities of ICSU's committee on Data (CODATA) and Science Council of Japan (SCJ). Tsugita organized Task Groups in CODATA and committees in SCJ to encourage protein scientists. He emphasized that scientific data should be fully and openly accessed, and pushed ICSU and SCJ to make statements against legal restrictions on access to data. His activities of data science will be introduced in the Memorial Session.

View full abstract

Our group has been developing an experimental technology called hydroxyl radical footprinting. This technology is an experimentally probe the solvent accessibility of protein surface by reaction with hydroxyl radical, which is highly reactive and almost same size as water molecule. The technology has been proved to be very successful for probing protein structure changes, protein/DNA binding and protein/protein interactions. These studies are only done with protein with solved structures since it is extremely difficult to interpret the data without knowing structures. The next logical step is to study on the proteins without solved structure. The computational modeling has been coming long way to achieve the place as regular structural biology tool. Ab initio or de novo prediction has achieved the level of accuracy that can be even used for molecular replacement for phasing in some cases. Rosetta is a powerful tool for de novo prediction, developed by Baker group. The critics for this approach are sometimes it predicts completely different fold, thus completely wrong structure. It is true for some cases for blind prediction such as CASP targets, as only criteria for structure selection are energy score and intuition if target sequence has no sequence identity high enough for comparative approach. The experimental data provide significant clues for such cases. The data have to be satisfied by predicted structure. Wrong fold usually has some inconsistency with experimental data. Baker group has been working on use of dipolar coupling data. Our group works with them for incorporating surface accessibility data into Rosetta protocols. For structure prediction, accessibility data indicate crucial information such as hydrophobic core residues. In the case of protein/protein interaction, accessibility data defines interaction interface directly. The study is still in early stage but we are having encouraging data. More detail and some real world work will be presented.

View full abstract

Introduction
Chemical proteomics is considered as one of the effective tool for identification proteins that bind specifically to candidate compounds by means of affinity chromatographic column and mass spectrometry technology in drug discovery. Comprehensive, but medium scale chemical proteomics gives insights to chemists as well as biologists, because this contains tremendous information in terms of chemical structures about biological equivalence, binding proteins and their domains, and protein interactions. Here we have developed a new tool to integrate that information.
Methods
A hundred different compounds were immobilized as affinity columns and HCT116 cell lysates were applied onto each column. After the identification, binding amount was estimated by calculating emPAI values. The emPAI values of basal protein expression levels in HCT116 also were obtained, and then ratio between binding emPAI- and basal emPAI-values to each protein was calculated as enrichment factor. We have applied this approach to mice brains for Alzheimer disease research.
Results
To evaluate finger prints of each compound, first we have made a large 2D heat map including over 100 compounds and 17,000 proteins for giving the bird view of results. As feasibility studies, two series of compounds known biological activities and having similar scaffold structures were added in our experiments. After 2D clustering analysis, these compounds were reasonably grouped as separate clusters in chemical dimension. After functional analysis of binding proteins, some of our compounds seemed to be preferable to ATP binding proteins, therefore we extracted protein kinases from binding protein lists and mapped them onto kinome map(Figure). We compared kinase expression between wild and APP/tau transgenic mice brains after kinase enrichments. Then we also did phosphoproteome analysis between these brains to identify substrates of Alzheimer disease specific kinases.

View full abstract

The identification of target proteins of bioactive compounds such as natural products, old medicines, pending compounds due to side effects, and so on, is now an interested theme in the post-Genome era. Affinity resins have been used for this purpose and exhibited some successful discoveries. However, it is not usually easy to identify them using traditional technology even now, especially for hydrophobic compounds such as orally active compounds. We have focused our efforts on developments of technologies for identification of them using affinity chromatography matrices bearing such compound. Our technologies including recent one that have been developed for these identifications will be presented with some successful examples.

View full abstract

DNA arrays have been successfully used for gene expression profiling of cells at the mRNA level. However, it has certain limitations: since there is not always a direct correlation between mRNA levels and protein expression (Gygi etal, 1999). Gene expression profiling may not represent the actual cellular proteome. Antibody arrays, which are a relatively new technology, are used to profile directly protein expression. They enable the researcher to profile hundreds of proteins and their post-translational modifications. Samples are not restricted to tumor cells and tissues but can be also biological fluids. In the last few years, researchers have developed several different antibody arrays. The common principle of this technology is based on the highly specific recognition between an antibody and its target antigen combined with miniaturization for high throughput screening. In this technology, antibodies are printed on a solid support, and the sample, usually containing a mixture of proteins, is applied on them. The interaction between the antibody and its specific antigen is then visualized by a choice of various detection methods. Identifying the proteins' changes may lead to the discovery of cancer-associated biomarkers that may assist in diagnosis, prognosis, patient monitoring and possibly for therapeutic purposes. Antibody arrays are a relatively new technology that enables one to perform multiplex high-throughput protein expression profiling. In this symposium, I will show the technical advances in slide-based antibody array and their application for biomarker identification within different cancers.

View full abstract

To discover biomarker candidates for diseases such as cancers, we originally developed a novel quantitative proteome method utilizing both mass spectrometry and the stable isotope labeling technique. This method is based on labeling with a chemical reagent, 2-nitrobenzensulfenyl chloride (NBSCl), which selectively reacts to tryptophan residues. Utilization of a pair of light (12C-labeled) and heavy (13C-labeled) NBS reagents, the relative quantitation can be performed from the intensities of generated paired-peaks with 6 Da mass differences. This method has been further improved by an optimization of the protocol, utilization of a new chromatographic enrichment method, and the finding of a novel matrix specifically suitable for NBS-moieties. These improvements resulted in a suppression of sample loss and an increase in sensitivity, and as a result, in an expansion of dynamic range of measurement. By a combination of NBS reagents with HPLC, automatic spotter, MALDI-TOF MS, and analytical software, we constructed a system called "NBS Biomarker Discovery System". Here, we applied this system to clinical samples such as colorectal carcinoma (CRC) and renal cell carcinoma (RCC) and they were compared with their corresponding normal tissue samples. In the case of CRC, for example, 128 proteins were identified as differentially expressed (>1.5-fold) proteins in cancer tissues. Six of them were further characterized and validated by immunohistochemistry and were confirmed to be cancer-localized. The system was also applied to two pairs of cells, highly metastatic cancer cells and their parental cancer cells and more than 100 differentially expressed proteins were identified as well in each pair. Nine of them are common between two pairs and are now under validation. These results indicate that the NBS analytical system is useful to discover drug target and/or biomarker candidates, and that the proteins identified here can be novel cancer biomarkers.

View full abstract

Osteoarthritis (OA), one of the most common skeletal disorders characterized by cartilage degradation and osteophyte formation in joints, is induced by accumulated mechanical stress; however, little is known about the underlying molecular mechanism. Several experimental OA models in mice by producing instability in the knee joints have been developed to apply approaches from mouse genetics. Although proteinases like matrix metalloproteases and aggrecanases have now been proven to be the principal initiators of OA progression, clinical trials of proteinase inhibitors have not been successful for the treatment, turning the interest of researchers to the upstream signals of proteinase induction. These signals include undegraded and fragmented matrix proteins like type II collagen or fibronection that affects chondrocytes through distinct receptors. Another signal is pro-inflammatory factors that are produced by chondrocytes and synovial cells; however, recent studies that used mouse OA models in knockout mice did not support that these factors have a role in the central contribution to OA development. Our mouse genetic approaches found that the induction of a transcriptional activator Runx2 in chondrocytes under mechanical stress contributes to the pathogenesis of OA through chondrocyte hypertrophy. In addition, chondrocyte apoptosis has recently been identified as being involved in OA progression. We hereby propose that these endochondral ossification signals may be important for the OA progression, suggesting that the related molecules can clinically be therapeutic targets of this disease.

View full abstract

In the usual sense, the disease biomarker may imply a diagnostic one. Previously, such a diagnostic biomarker of Alzheimer's disease (AD) was highly desired because the diagnosis of AD was exclusion one. Once dementia was diagnosed, a great number of diseases causing dementia were excluded and finally the diagnosis of AD was made. In these 20 years, the biology of AD has been greatly advanced. The earliest symptom is loss of episodic memory, which is followed within several years by impairment of other cognitive domains. CSF from AD patients contains high levels of phosphorylated and total tau, and low levels of Abeta42. This is the case even in the period of mild cognitive impairment (MCI). Furthermore, refined imaging techniques, especially MRI has made it possible to prevent us from misdiagnosing other diseases as AD. Thus, at well-equipped medical centers, the diagnostic accuracy may have already reached more than 90%. In view of the temporal profile of AD, Abeta accumulation would precede neurofibrillary tangle formation that is associated with cognitive symptoms by more than a decade. As this period accompanies none of the cognitive symptoms, we should develop the method of screening cerebral Abeta accumulation by means of blood test. According to the data based on transgenic mice (Tg2576), plasma (and CSF) Abeta42 levels decrease when Abeta42 starts to accumulate in the brain. If it is true with humans, long-term follow-ups of plasma Abeta42 levels may warn potential Abeta deposition, and such blood-test-positive people should subsequently have sophisticated examinations including MRI or amyloid imaging. These at-risk people could be advised for the appropriate life style, especially foods and exercise, and if needed, may be given gamma modifier. This strategy should at least postpone the development of AD in an individual several years, and in this way, we should make every attempt to decrease the number of AD patients.

View full abstract

The number of protein isoforms in the human proteome has known to be much higher than the number of genes in the human genome. This is in large part due to posttranslational modifications (PTMs) of proteins. Knowledge of these PTMs is extremely important because they may altar physical and chemical properties, folding, complex formation, stability, activity, and consequently, function of the proteins. The Quantum dots are high-intensity photostable fluorophores that have several distinct advantages over standard organic fluorescent dyes. These nanocrystals (commercially available as Qdots from Invitrogen; Carlsbad, CA, USA) consist of a semiconductor core of cadmium selenide (CdSe) coated with a shell of zinc sulfide(ZnS). An additional polymer layer over the Qdot enhances water solubility and enables conjugation to streptavidin or other biomolecules such as IgG. Slight changes in the size of the Qdot semiconductor core change the emission spectra, resulting in a redder color as the size of the core increases. The ability to synthesize different populations of quantum dots with narrow, symmetric and strong emission spectra permits multiplexing, a property that is important for simultaneous detection of several types of PTMs. With the Qdot technology, we developed a new proteomic method for simultaneous detection of various PTMs of proteins on a single Western or immunoblotting analysis. We applied this method to examine the changes of the PTMs of proteins from various tissues and cells in a diabetes model Otsuka Long-Evans Tokushima Fatty (OLETF) rat. Protein carbonyls, ubiquitination and advanced glycation end products (AGEs), were detected simultaneously on a single blot. The advantages of high sensitivity, multiplex labeling, photo- and chemical stability of the Qdots will accelerate information flow in proteomics.

View full abstract

Introduction: For the PTM analysis using MS, protein is usually digested into peptides with proteases. The detection of these peptides is largely depending on their size; larger peptides tend to have low ionization efficiency, whereas smaller peptides can not be well separated by LC. Therefore, the peptide size suitable for detection need to be controlled by choosing appropriate protease to identify the PTM structure of the proteins. In our previous study, the immobilized protease technique was shown to be useful for the protein digestion. Here, we applied this technique to the multiple protease digestion and evaluated its effectiveness in enhancing detection efficiencies of PTM's.
Methods: Three kinds of proteases (trypsin, LysC, V8) were immobilized on the single solid surface, and used to digest BSA, beta casein and ovalbumin as model proteins. To accelerate proteolysis reaction by sample convection, the vibration reaction unit, which was developed in our previous study (Anal. Chem. Insights, 2007, 2, 69-74), was used. Digested proteins were analyzed using LC-MS. The number of the identified peptides, protein score, sequence coverage, and identification of the phosphopeptides were compared.
Results: Digestion using different proteases resulted in the identification of different peptide sequences. Consequently, the combination of these individual results provided higher sequence coverage, suggesting that the protein digestion using multiple proteases is effective to improve the accuracy of protein identification in the MS analysis. The detection of the phosphopeptides changed depending on the proteases used for the analysis; phosphopeptides from beta-casein and ovalbumin were identified only in their LysC and V8 digests, respectively. Therefore, it suggests that the utilization of the multiple kinds of proteases should be considered for the protein phosphorylation analysis.

View full abstract

While phosphoproteins have attracted great interest toward the post-genome research (e.g., clinical diagnosis and drug design), there have been few procedures for the specific separation and enrichment of native phosphoproteins from cells or tissues. Here, we describe a simple and efficient procedure to enrich phosphorylated proteins comprehensively from a complex mixture containing solubilized cellular proteins. This method is based on immobilized metal affinity chromatography using a dinuclear zinc(II) complex, Phos-tag, attached on a hydrophilic vinyl-polymeric bead (TOYOPEARL, Tosoh Corp., Tokyo, Japan). TOYOPEARL is more excellent than other column matrix beads, such as agarose, in physical and chemical properties. The binding, washing, and elution processes were all conducted without a detergent or a reducing agent at pH 7.5 and room temperature. An additive, 0.50 M CH₃COONa, was necessary in the binding and washing buffers (0.10 M Tris-CH₃COOH, pH 7.5) to prevent the nonphosphorylated proteins from binding. The absorbed phosphoproteins were eluted using a mixed buffer solution (pH 7.5) consisting of 0.10 M Tris-CH₃COOH, 10 mM NaH₂PO₄-NaOH, and 0.50 M NaCl. The concentrations of CH₃COONa and NaCl used are only half relative to the conventional protocol of phosphate-affinity chromatography using Phos-tag Agarose. In this study, we demonstrate the separation and enrichment of phosphoproteins from an EGF-stimulated human epidermoid carcinoma A431 cell lysate using the novel phosphate-affinity bead and improved buffer conditions. The total time for the column chromatography (1 mL-beads scale) was less than one hour. The strong enrichment of the phosphorylated proteins into the elution fraction was evaluated using SDS-PAGE and 2-DE (IEF/SDS-PAGE) followed by Western blotting with a biotin-pendant Phos-tag and some antibodies, Pro-Q Diamond phosphoprotein gel staining, and MS analyses.

View full abstract

Protein phosphorylation is a critical regulatory step in signaling networks and is arguably the most widespread protein modification, affecting almost all basic cellular processes in various organisms. Advances in mass spectrometry-based technologies accompanied with phosphopeptide enrichment methods paved the way for high-throughput, large-scale in vivo phosphorylation dynamics. Recently we developed highly specific phosphopeptide enrichment method using hydroxy acid-modified metal oxide chromatography (HAMMOC), where the chemo-affinity of metal oxides to phosphopeptides is enhanced by the addition of aliphatic hydroxy acids[1]. This HAMMOC/nanoLC-MS approach allows the identification of more than one thousand phosphorylation site per single LC-MS run from whole cell lysates without any stimulus.
Several kinase inhibitors have been launched or in the late clinical trial stages as anti-cancer molecular-targeting drugs. However, it often happened that these small molecules also inhibit other kinases, which would lead to unexpected side-effects or synergistic effects. Therefore it is necessary for each drug to examine the in vivo inhibition activity for as many kinases as possible, which is impractical at present. The HAMMOC/nanoLC-MS approach in combination with stable isotope labeling enables to quantify the proteome-wide phosphorylation change induced by the drug treatment. We examined cultured cancer cells treated with an EGFR inhibitor and found that the phosphorylation status of proteins belonging to EGFR pathway was down-regulated by the drug and most of them were not reported as phosphorylated proteins.
Here we will present our recent advances in the improved HAMMOC approach as well as the application to the phosphorylation profiling of cancer cells treated with kinase-targeting drugs.

[1] N. Sugiyama, T. Masuda, K. Shinoda, A. Nakamura, M. Tomita, Y. Ishihama, Mol. Cell. Proteomics, 6, 1103-1109 (2007).

View full abstract

Cell cycle of tumor cells are often dysregulated, and its regulation differs in each tumor cell line. Thus, a single cell line analysis cannot determine whether the changes induced by cell cycle progression are general events occurred in various tumor cell lines or cell-type-specific events observed in a particular cell type. To identify general M-phase specific phosphoproteins, we compared phosphorylation status and changes upon nocodazole treatment in 3 distinct tumor cell lines, HeLa, HCT-116, and NCI-H460, all of which are arrested at M-phase by nocodazole treatment. Phosphopeptides from 3 distinct tumor cell lines with or without nocodazole treatment were enriched by MassPrep^TM (Waters) and analyzed by LC-MS-based protein identification technology and data analysis (Fig.1). In total, 1,570 unique phosphopeptides assigning to 726 phosphoproteins were identified with false discovery rate at 0.8 % in average. Although there are some common phosphoproteins in all 3 tumor cell lines, the identified phosphopeptides varied among cell lines (Fig.2). Distinct phosphopeptides include representative M-phase proteins, such as kinesin family members, microtubule-associated proteins, and nuclear mitotic apparatus protein 1, as well as some signaling molecules, such as Src, raf, Erk, and PDK1. These results suggest that distinct signaling cascades are activated or inactivated in 3 tumor cell lines. In spite of the variability between cell lines and LC-MS analyses, phosphorylation of nucleophosmin S254 and phosphoribosylaminoimidazole carboxylase S27 were identified in all the independent LC-MS analyses of 3 tumor cell lines only when treated with nocodazole. Thus, our results suggest that phosphorylation of these 2 proteins serve as M-phase specific biomarkers, which may be used to monitor the efficacy of M-phase inhibitors treated as anti-cancer drugs.

View full abstract

To clarify roles of an endogenous pain modulatory system of the central nervous system (CNS) in hyperalgesia, we tried to identify qualitative and quantitative protein changes by a proteomic analysis using an animal model of hyperalgesia. Specifically, we first induced functional hyperalgesia on male Wistar rats by repeated cold stress (specific alternation of rhythm in temperature, SART). We then compared proteomes of multiple regions of CNS and the dorsal root ganglion between the hyperalgetic rats and non-treated ones by 2-D PAGE in the pI range of 4.0–7.0. We found that SART changed the proteomes prominently in the mesencephalon and cerebellum. We thus analyzed the two brain regions in more detail using gels with narrower pI ranges. As a result, 29 and 23 protein spots were significantly changed in the mesencephalon and the cerebellum, respectively. We successfully identified 12 protein spots by a MALDI-TOF/TOF MS and subsequent protein database searching. They included unc-18 protein homolog 67K, collapsing response mediator protein (CRMP)-2 and CRMP-4, which were reported to be involved in neurotransmitter release or axon elongation. Interestingly, mRNA expression levels of these three proteins were not changed significantly by the induction of hyperalgesia. Instead, we found that the detected changes in the protein spots are caused by the post-transitional modification (PTM) of proteolysis or phosphorylation. Taken together, development of the hyperalgesia would be linked to PTM of these three CNS proteins. Regulation of the PTM may be therapeutic targets to treat hyperalgesia.

View full abstract

Based on systematic protein-protein network analysis, our laboratory has started new chemical-biology project. Over the last decade, we discovered new interactions of disease related or causative proteins by large-scale protein-protein network analysis, leading to uncover precise molecular mechanisms of several diseases. The primary goal of the project is to establish efficient and versatile drug discovery platform targeting crucial and vital interactions for disease treatment using second generation natural chemistry libraries. And secondly, we evaluate the newly discovered drugs in vitro and vivo as chemical probes. To this aim, we develop and integrate ultra high sensitive mass spec facilities, chemo-informatics platform, large scale natural chemical sources, combinatorial chemistry and fluorescence imaging technology.

View full abstract

Human protein research on the genomic scale is currently becoming more plausible to conduct due to the accumulation of both genome sequencing data1 and full-length cDNA clones. If many proteins are handled simultaneously, it is assumed that analysis of many reactions such as enzyme-substrate reactions, protein-protein interactions, protein modifications, processing by peptidases, etc., could be carried out more easily and effectively. In vitro proteome research, which handles expressed proteins on a large scale, has the advantage of being able to handle proteins that exist in a restricted fashion in vivo, such as proteins with tissue-specific, stage-specific or intrinsically rare expression. Recently we constructed both resources and an infrastructure for genomic scale protein expression with multi-use purpose. To build this foundation, we adopted the so-called Gateway technology, which can adequately handle multiple expression modes, and planned the accumulation of Gateway entry clones as key resources. We produced about 60,000 entry clones, including full-length ORF, processed ORF, and Domain ORF type. By using these resources, we search the new target for drug discovery (ex. protein-protein interaction ) and construct the screening system for inhibitor of protein-protein interaction. In this talk, we would like to introduce the importance of the construction of protein expression resources and infrastructure for genomic scale protein expression.

View full abstract

Poor in targets for drug discovery is a significant problem to develop clinical drugs, today. Thus, novel targets and/or strategies for drug screening are greatly required. Protein-protein interaction is one of exciting strategies to overcome the difficulty in drug development. On the other hand, regulation of protein-protein interaction with small molecule compounds has been considered to be difficult because larger spaces take part in the protein-protein interaction than those were expected to be covered by small molecule compounds. To challenge this opposite concept, we developed screening method to estimate protein-protein interaction applicable for drug screening. There are several methods to observe protein-protein interaction, however, rapidity and easiness are important for high through-put screening. As the results, we chose alpha-screen and bimolecular fluorescence complementation (BiFC) assay as the first screening method (1-2). We have constructed cell-based and in vitro BiFC screening system, and carried out 10 screenings using about 130,000 samples in each screening. As the results of screening, although, the hit ratio is extremely low, we have found 4 active substances. Including low selective regulators for protein-protein interaction, the average molecular weight was about 800. These results suggested that regulators of protein-protein interaction should have relatively large molecular weight as expected. Thus, natural compounds are considered to be suitable sources for protein-protein interaction regulators. 1) T. K. Kerppola, Visualization of molecular interactions by fluorescence complementation. Nat Rev Mol Cell Biol., 7, 449-456 (2006). 2) M. L. MacDonald, J. Lamerdin, S. Owens, B. H Keon, G. K. Bilter, Z. Shang, Z. Huang, H. Yu, J. Dias, T. Minami, S. W. Michnick and J. K. Westwick. Identifying off-target effects and hidden phenotypes of drugs in human cells. Nat. Chem. Biol., 2, 329-337 (2006)

View full abstract

In silico analysis using the computational techniques of protein structure prediction, protein-protein interaction prediction, chemoinformatics, ligand docking and virtual screening is an integral part of drug discovery. In chemical biology project, in silico analysis contributes to the following issues,
- Prediction of structural model of protein-protein interaction identified by a protein interaction network analysis system utilizing mass spectrometry.
- Discovery of seed compounds that regulate biological system using virtual screening from commercially available chemical libraries.
- Design and optimization of seed compounds or natural products using structure based approach through cooperation with combinatorial library synthesis and natural product library groups.
We propose an in silico analysis to discover seed compounds that regulate protein-protein interaction related to diseases such as cancer and other lifestyle-related diseases.

View full abstract

Naturally occurring cyclic depsipeptides exhibit various biological activities. It would be efficient for drug discovery if a variety of their analogues can be synthesized in a high-throughput manner. Beauveriolides III, isolated from the culture broth of Beauveria sp. FO-6979, leads to a reduction in the number and size of lipid droplets in macrophages without cytotoxic effect and also inhibits cholesteryl ester (CE) synthesis, leading to suppression of foam cell formation in macrophages [1]. In order to elucidate the structure activity relationships and to discover a new drug lead, we have studied a method for rapid synthesis of beauveriolide III and its analogues. Initially, the unnatural amino acid and aliphatic acid were asymmetrically synthesized as building blocks of beauveriolide III. Solid-phase peptide synthesis by Fmoc strategy provided a linear cyclization precursor, which was allowed to undergo macrolactamization in solution phase leading to beauveriolide III.
To synthesize a variety of beauveriolide analogues, we assembled the three synthetic blocks including unnatural amino acids and aliphatic acids using a split and mix method on a polymer-support. Macrocyclization in a parallel manner, followed by preparative HPLC provided a 200-member combinatorial library of beauveriolide analogues in pure form.
The inhibitory activity of the beauveriolide analogues against CE synthesis in mouse peritoneal macrophages was evaluated. The linear cyclization precursors did not indicate the biological activity. Therefore, the macrocyclic ring structure plays a crucial role in inhibitory activity. We also found that one of the analogues exhibits 10 times more potent activity than the original natural product [2].
[1] Namatame, I.; Tomoda, H.; Ishibashi, S.; Omura, S. Proc. Natl. Acad. Sci. USA 2004, 101, 737.
[2] Nagai, K.; Doi, T.; Sekiguchi, T.; Namatame, I.; Sunazuka, T.; Tomoda, H.; Omura, S.; Takahashi, T. J. Comb. Chem. 2006, 8, 103.

View full abstract

Structural analysis of O-glycopeptides is much more challenging than N-glycosylation due to a number of difficulties; There are more than fifteen glycosyl enzymes (GalNTs) in human for initiating O-glycosylation of proteins, and no consensus sequence for O-glycosylation has been described. The assignment of attachment sites of the glycan is further complicated by the fact that there may be many sites of O-glycan attachment adjacent to each other and that these are frequently located in a region of peptide sequence in which many serine and threonine residues are found together, all of which may or may not be glycosylated. In addition, proline residues are often present in the sequence and hamper the sequencing with MS/MS, due to the high efficiency of cleavage at the amide bond on the N-terminal of proline residue in collision-induced dissociation (CID). As a result, quite a small number of attachment sites have been experimentally determined and recorded in the protein database to date.
In our O-glycoproteome project, various chemical methods such as b-elimination followed by Michael addition of different kinds of substituents and partial deglycosylation with trifulromethanesulfonic acid were examined, but they were inadequate to solve the problems when combined with CID. Electron capture dissociation and a similar technique electron transfer dissociation (ETD) produce significantly different types of fragment ions than other fragmentation modes including CID, which introduces internal vibrational energy. In ECD and ETD, labile post-translational modifications such as phosphorylation, glycosylation and others remain attached to the backbone during ECD or ETD MS/MS experiments allowing determination of the site and identity of post-translational modifications. Indeed, we have determined a number of new O-glycosylation sites of plasma proteins with the aid of ETD.

View full abstract

Discovery of cancer biomarker is a major objective of clinical proteomics; molecular biomarkers allow for detection of early-stage cancer, for classification of tumors, and for monitoring their progression, regression following treatment and/or recurrence. Glycoproteins are one of the ideal candidates for biomarkers because most secretory and cell surface proteins are glycosylated, and their glycan structures frequently and drastically change during tumorigenesis as well as normal cell differentiation. In addition, the core proteins are often expressed as cell specific. In fact, most cancer-related biomarkers available to date, such as prostate specific antigen, carcinoembryonic antigen, and alpha-fetoprotein, are glycoproteins. Thus far, we developed LC/MS-based glycoproteomic analysis method, termed IGOT (isotope-coded glycosylation site-specific tagging), and applied it for large-scale profiling of N-glycosylated proteins of C.elegans and mouse tissues; about 1,450 sites on 829 proteins and 4,500 sites on 2,300 proteins were determined, respectively. Using the large dataset of N-glycoproteins, we constructed a glycoprotein database, JCGGDB, and the partial data will become available soon at the Japan Consortium for Glycobiology and Glycoscience. Recently, to apply the IGOT procedure for biomarker discovery, we incorporated a step of differential stable isotope tagging for quantitative analysis. The isotope tags could be incorporated into the core peptide portion by both enzymatic and chemical methods. In this presentation, the methods of quantitative glycoproteomics and its application will be introduced.

View full abstract

Glycosylation regulates the biological functions of most proteins. This regulation is caused by not only addition of glycans to the peptides but also alteration of the glycan structures in the status of the cells. Therefore, detection of such alterations is very useful for understanding real-time condition of cells, for example, diseases. Especially, it is widely known that cancer cells produce different glycan structures from normal cells. Analytical method for determination of complicated glycan structures using quite a small amount is the key to developing novel biomarkers. To aim at highly sensitive and simple determination of glycan structures, we have developed negative-MALDI-QIT-TOFMSⁿ analysis after labeling with pyrene derivatives of neutral and acidic oligosaccharides. Introduction of pyrene to oligosaccharides prominently enhances productions of negative ions even from neutral oligosaccharides. MSⁿ analysis of negative ions is very powerful for identification of branched and isomeric oligosaccharides. It was also demonstrated that pyrene-derivatization improves ionization of glycopeptides as well as oligosaccharides but suppresses ionization of peptides. Thanks to the important effects we could detect gycopeptides in practical samples in which only signals of peptides were obtained without prene-derivatization. We have applied this method to various glycoproteins prepared from patient's sera and cleared altered glycosylation. This study is supported by Japan Science and Technology Agency and New Energy and Industrial Technology Development Organization.

View full abstract

Glycosylation that has an important role in cellular functions is regulated by various factors such as nucleotide sugars, their transporters, glycosyltransferases, glycosidases, and cell surface receptors. Nucleotide sugars are donor substrates of glycosyltransferases and their availability and localization regulate glycosylation levels. Previous reports showed the levels of certain nucleotide sugars by HPLC. However, the separation of each nucleotide sugar from nucleotides was incomplete. Here we report a high resolution method by using ion-pair reversed-phase HPLC to determine nucleotide sugars involved in glycosylation simultaniously. A series of 9 nucleotide sugars (CMP-NeuAc, UDP-Gal, UDP-Glc, UDP-GalNAc, UDP-GlcNAc, GDP-Man, GDP-Fuc, UDP-GlcA and PAPS) and 12 relating nucleotides were perfectly separated on a ODS column and were detected by UV detector. The developed method gave a chromatogram with high sensitivity for cellular samples, which contained structurally similar nucleotide sugars. Using this method, we were able to determine the levels of nucleotide sugars in mouse embryonic fibroblast cells and to observe their changes in various cellular conditions. Interestingly, levels of UDP-GlcNAc, UDP-GalNAc, and CMP-NeuAc were markedly varied in the exponential and stationary phases of cellular conditions. To investigate the upstream signaling and gene expression that would affect the changes of nucleotide sugars, we performed DNA-microarray using RNA from cells in each phase. Several genes involved in nucleotide sugar metabolism such as N-acetylglucosamine kinase, Nagk, and GlcNAc 2-epimerase, Renbp, were markedly up-regulated in the stationary phase, compared to the exponential phase. In our newly developed method we could be able to determine the levels of most of all kinds of nucleotide sugars simultaneously, which enable us to show an overview of cellular glycosylation status in glycome.

View full abstract

Malignant transformation is frequently accompanied by a drastic alteration in the structures of oligosaccharides at the cell surface. To understand the nature of the mechanism of alteration of oligosaccharide structures, we have analyzed and compared the structures of glycosphingolipids (GSLs) of colon cancer cells and normal colon epithelial cells from 18 patients. To determine accurate structures of GSLs of colon cancer and normal colon epithelial cells, these cell populations were purified from colon cancer tissues and surrounding normal colon tissues using the epithelial cell marker CD326. GSLs were extracted from the cells and carbohydrate moieties were released by endoglycoceramidase. The oligosaccharides were labeled by pyridylamination, and structurally characterized using a two-dimensional HPLC mapping technique, electrospray ionization tandem mass spectrometry, enzymatic cleavage and methanolysis. This analysis demonstrated that GSLs in normal colon epithelial cells are, in most cases, composed of neutral GSLs having type 1 chains with acidic GSLs being very minor components. The predominant neutral GSL of normal colon epithelial cells is Le^a, together with significant amounts of lactose and Le^b. The expression pattern of GSLs of colon cancers cells are different and vary case to case. However, human colon cancer cells from most cases consist of more various GSLs than normal colon epithelial cells having Le^a, Le^x, LST-c and GM3 as major components, and Le^y, Le^b, sialyl Le^x, sialyl Le^a, IV⁶NeuAcIV²Fuc-nLc₄, VI³NeuAc-nLc₆, and VI⁶NeuAcIII³Fuc-nLc₆ as minor components. These results indicate that levels of GSLs having type 2 chains and sialylated GSLs are elevated in carcinogenesis. Furthermore, we found cancer cells from two patients which presented a variety of sulfated GSLs.

View full abstract

Patients with inflammatory bowel disease (IBD) share several immunologic similarities with rheumatoid arthritis (RA). RA patients have significantly increased levels of serum agalactosyl IgG. However, the oligosaccharide structure of igG and its relationship to disease activity and prognosis of IBD patients have not yet been investigated. Our aim was to investigate the clinical significance of analyzing the oligosaccharide structure of serum IgG in IBD patients. IgG oligosaccharide structures were analyzed using high performance liquid chromatography in 60 patients with Crohn's disease (CD), 58 patients with ulcerative colitis (UC), 27 healthy volunteers (HV), and 15 colonic inflammations not related to IBD (disease controls; DC). The ratio of agalactosyl fraction in the fucosylated IgG oligosaccharides (G0F/G2F) of CD and UC was significantly greater than that of HV and DC. The percentage of subjects with a high G0F/G2F in CD, UC, HV, and DC was 72%, 33%, 0%, and 0%, respectively. G0F/G2F in active patients was significantly higher than that in patients in remission. G0F/G2F was also significantly higher in patients with extensive disease than in patients with inflammation in the limited lesion. G0F/G2F had higher sensitivity to diagnose IBD compared with anti-Saccaromyces cerevisiae antibody (ASCA) judged by receiver operating characteristic (ROC) curve and area under the curve (AUC) analysis. Moreover, G0F/G2F correlated with the prognosis of UC patients: the patients with a high G0F/G2F showed significantly lower frequency of maintaining long-term remission than the patients with a low G0F/G2F. The activity and mRNA level of beta-1,4-galactosyltransferase in B cells and plasma cells of CD were significantly decreased than those of UC. Collectively, the agalactosyl fraction in the fucosylated IgG oligosaccharides was increased in IBD. G0F/G2F is a potentially effective diagnostic marker of disease activity in both CD and UC, and of the clinical course in UC.

View full abstract

Our previous studies found that the concentration of fucosylated haptoglobin had increased in the sera of patients with pancreatic cancer compared with those of other types of cancer and normal controls. Haptoglobin, an acute phase protein, has four potential N-glycosylation sites, although it remains unknown which site is responsible for the change in fucosylated N-glycans. In this study, site-specific N-glycan structures of haptoglobin in serum obtained from patients with pancreatic cancer or chronic pancreatitis were analyzed using liquid chromatography-electrospray ionization mass spectrometry. Mass spectrometry analyses demonstrated that concentrations of total fucosylated, di-, tri- and tetra-branched glycans of haptoglobin were increased in the sera of pancreatic cancer patients. Tri-antennary N-glycans containing a Lewis X-type Fuc were markedly increased at the Asn211 site of haptoglobin N-glycans. This approach could make a differential diagnosis for tumor-forming pancreatitis form pancreatic cancer in one complicated case. Furthermore, we recently identified a factor to stimulate the production of fucosylated haptoglobin in the conditioned medium of a pancreatic cancer cell line. Thus, the present study provides evidence that site-specific analyses of fucosylated N-glycans may be useful as a novel tumor marker for pancreatic cancer.

View full abstract

Peptidomics has received more attention in the search for biomarkers, as almost all the potential biomarkers reported through serum or plasma analysis have turned out to be proteolytic fragments of proteins. Peptidomics is also promising as a tool to discover bioactive peptides which are beyond the reach of current proteomics. But there are many problems to be solved in peptidomics. Both generation and degradation pathways of the peptides are present together in the cells, tissue and plasma, and endogenously generated peptides are easily converted into secondary products by proteolysis during extractions and preparations. Proteins are simultaneously digested by proteases, providing numerous peptides. Therefore, it has long been recognized that the real status of endogenous peptides are extremely difficult to analyze and identify. However, recent progress in peptidomics allowed us to identify peptides in their endogenous forms and to apply this technology to the biomarker searches, even though there were still partially digested peptides. On the other hand, peptidomic analysis has not succeeded in the identification of new bioactive peptide in the mammals, but a recent study made it possible by using an improved peptide preparation method. In this symposium, I would like to summarize problems in peptidomics, and then introduce new technologies and recent progress in the discovery of biomarkers and bioactive peptides, which will be detailed by the speakers.

View full abstract

Proteomics has generated great anticipation in the cancer field as it may hold promise to establish simple blood tests through mass spectrometric (MS)-based profiling of patient proteomes/peptidomes. However, several concerns have been voiced regarding biological (patient demographics, environmental, dietary and psychological factors), technological (sample collection, processing and storage, analytical chemistry) and data mining artifacts that may introduce bias. We have established an automated platform to measure serum peptides using magnetic beads for selective capture and release, followed by a MALDI-TOF MS read-out. It is more sensitive than SELDI chips as spherical particles have larger combined surface areas than small-diameter spots. Hundreds of peptides can be detected in a droplet of serum; use of robotics facilitates throughput and ensures reproducibility. Unfortunately, serum preparation, handling and storage remain problematic, especially among archived samples. Thus, we have made great efforts to identify and remedy potential sources of bias, such as blood collection tubes, clotting times and temperature, number of freeze-thaw cycles, surface chemistries of magnetic beads, MALDI sample crystallization, laser irradiation, and various automation steps. In addition, adequate signal processing of the spectra (e.g. peak alignment, binning, etc.) proved more difficult and rate limiting than the actual statistical analysis. We therefore developed a minimal entropy-based algorithm that simplifies and improves alignment of spectra. Overall, our current SOP and informatics pipeline has greatly reduced handler variability and induced error typically associated with peptide measurements of clinical samples. A different problem of oncopeptidomics is that most of the serum peptidome as measured using high-throughput, low-resolution MS is the result of exopeptidase activities targeted at coagulation and complement activation byproducts. Formation of these peptides can be deregulated in disease and is easily affected by sample collection and handling. We have therefore designed and optimized a reliable in vitro activity test, uniquely suited to probe the altered balance of exopeptidases and/or their modulators in the blood of cancer patients.

View full abstract

Taking advantage of the sequence databases that have now accumulated, proteomic or peptidomic analyses of human body fluids such as blood, urine, bile, etc. have become one of the most important issues relating to the identification of biomarkers or disease markers. We have established the workflows using an¹⁸O-labeling method for comparative analysis of two paired protein pools or quantitation of a marker candidate peptide in urine or bile. The mass spectra containing¹⁸O-incorporated ion species could be precisely interpreted by ISOTOPICA (1), a software application that allows the relative abundances of narrow-split paired peaks to be resolved.
1) Fernandez-de-Cossio J., Gonzalez L.J., Satomi Y., Betancourt L., Ramos Y., Huerta V., Besada V., Padron G., Minamino N., Takao T. Rapid Commun. Mass Spectrom. 18, 2465-2472 (2004); Nucleic Acids Res. 32, 674-678 (2004).

View full abstract

Lung cancer is classified into small cell lung cancer and non-small cell lung cancer. Small cell lung cancer is a major cause of cancer deaths and accounts for 15 to 20% of all lung cancers. Because this cancer is initially highly responsive to chemotherapy, early detections are desired. Serum tumor markers are non-invasive diagnostic tools for malignant tumors and they are commonly used for the screening of cancer and as an indicator of the treatment effect. In this study, in order to develop new tumor marker for small cell lung cancer, secreted proteins from small cell lung carcinoma cell lines are analyzed in proteomic approach. Lung carcinoma cell lines were cultured in serum-free media and conditioned media were collected. Proteins fractioned with RP-HPLC were digested by protease and peptide sequence was determined using tandem mass spectrometry (MS/MS). About 50 proteins were detected in cultured media using MALDI-TOF MS/MS. Among these proteins, proneurotensin / neuromedin N (proNT/NMN) was determined in cultured medium. Moreover, proNT/NMN was detected in 4 of 9 small cell lung carcinoma cell lines but not from 9 non-small cell lung carcinoma cell lines. These results indicate proNT/NMN has a high potential as a specific tumor marker of small cell lung cancer. ProNT/NMN is the common precursor of the two related neuropeptides, neuromedin N (NMN) and neurotensin (NT). It is well known that neuroendocrine carcinomas (including small cell lung carcinomas) secrete various neuropeptides such as neurotensin or gastrin-releasing peptide. However, neuropeptides are known to be unstable in the blood and they are very difficult to establish as tumor markers. Because pro-form hormone, such as pro-gastrin releasing peptide (ProGRP), is more stable in the blood, it is likely that proNT/NMN has great potential as a tumor marker in a similar way to proGRP.

View full abstract

Serum proteins reflect physiological or pathological states in humans and are an attractive option for use in the discovery of disease biomarkers. However, the large dynamic range of serum proteins makes any quantitative analysis very challenging because a large number of low abundance proteins are often masked by a few high abundance proteins. Furthermore, analyses of peptides, including the cleaved fragment of proteins, are difficult because of carrier protein binding.
We have established a novel strategy, called the K-method, for extracting peptides from serum with high efficiency and high reproducibility, as compared to the typical peptide extraction method that uses organic solvent precipitation and ultrafiltration. Using the K-method, we extracted low molecular weight proteins/peptides from 0.01 mL of serum, followed by fractionation by RP-HPLC. The 60 fractions that were separated subsequently underwent comparative analysis by using MALDI-TOF MS, enabling us to observe a total of more than 1,500 peptides. To examine the reproducibility of our method, we evaluated four sera from healthy volunteers. More than 90% of the peaks were commonly observed in three of the four sera. Our results suggest that high reproducible extraction and fractionation enables us to treat more than 90% of these peptides as targets of comparative analysis. These results indicate that the reproducibility of the peptide extraction and RP-HPLC fractionation is extremely high.
We applied this method to the sera from eight colon cancer patients and eight healthy volunteers. Using MALDI-TOF MS/MS, we discovered and identified four potential biomarkers, which had p values less than 0.005. One of the peptides, the concentration range of ng/mL, was identified as a cleaved fragment of a cellular protein that has previously been reported to increase in cancer tissue. These results indicate that our new method has the potential to play a prominent role in novel biomarker discoveries.

View full abstract

Peptidergic hormones and neurotransmitters are known to be produced by the specific cleavage of their precursor proteins that per se have no biological functions. The neutrophil-activating peptides we recently identified, however, are the peptides cleaved from mitochondrial proteins by proteolysis [1,2]. Therefore, we designated functional peptides hidden in protein sequences "cryptides" and named them "mitocryptides" because they are hidden in mitochodorial protein sequences [3]. Some of these peptides activate Gi type of G proteins directly, and neutrophils are suggested to be stimulated by the direct (i.e., not via GPCRs) activation of G proteins. These peptides had features, in common, in their distributions of charged and hydrophobic amino acid residues, but homologies in their primary structures were not apparent. In the present study, we predicted functional cryptic peptides that activate G proteins, based on the distribution of charged and hydrophobic residues. Receptors for these peptides were also investigated by the direct cross-linking experiments between peptides and their targeted proteins. The finding of functional cryptic peptides is expected to lead to the identification of novel signaling mechanisms where such peptides are involved in the regulation of bio-functions
.
[1] Mukai, H., Hokari, Y., Seki, T., Nakano, H., Takao, T., Shimonishi, Y., Nishi, Y., and Munekata, E. (2001) Peptides: The Wave of the Future, 2001, 1014. [2] Mukai, H., Matsuo. Y., Kamijo, R., and Wakamatsu, K. (2004) Peptide Revolution: Genomics, Proteomics & Therapeutics, 2003, 553. [3] Ueki, N., Someya, K., Matsuo, Y., Wakamatsu, K., and Mukai, H. (2007) Biopolymers (Pep. Sci.), 88, 190.

View full abstract

Peptidomics refers to a comprehensive study of peptides that are naturally cleaved from larger proteins by endogenous proteases. Of all naturally occurring peptides, bioactive peptides such as peptide hormones and neuropeptides have received the most attention in life sciences, since they function as intercellular signaling molecules to regulate physiological processes. While an increasing number of endogenous peptides are identified by peptidomics, sequence information alone does not tell us which peptides are bioactive. To facilitate the discovery of bioactive peptides, we focused on C-terminal amidation, a post-translational modification shared by many known bioactive peptides. We peptidomically analyzed peptides released by a human cell line and identified two novel amidated peptides, designated neuroendocrine regulatory peptide (NERP)-1 and NERP-2(1). NERPs are derived from the neurosecretory protein VGF that was originally thought to associate with neuronal differentiation (2,3). The endogenous presence of NERPs in rat brain was confirmed by mass spectrometric characterization of immunoreactive substances. Since immunoreactive NERPs were present in the rat hypothalamic neurons producing the antidiuretic hormone vasopressin, we investigated biological functions of NERPs in the context of vasopressin physiology and obtained evidence that they are suppressors of vasopressin release. It should be noted that NERPs would evade proteomic analysis because they harbor several basic amino acid residues. To the best of our knowledge, NERPs are the first mammalian bioactive peptides discovered by peptidomic studies. Overall, this study illustrates one solution to identify novel bioactive peptides in "omic" studies. References: 1) Yamaguchi, H., Sasaki, K. et al. J. Biol. Chem. 282, 26354-60 (2007). 2) Levi, A., Eldridge, J.D. & Patterson, B.M. Science 229, 393-5 (1985). 3) Nagasaki, K., Sasaki, K. et al. Neurosci. Lett. 267, 177-80 (1999).

View full abstract

HKUPP (Human Kidney and Urine Proteome Project) has been recently approved as one of HUPO-sponsored scientific initiatives on October 2007, in which more than 150 participants collect and share proteomics data on the kidney and urine in order to understand functions of the kidney and mechanisms of disease progression, and finally to discover biomarkers and target molecules for new therapeutics of kidney diseases. Based on our continual discussion through foregoing symposia and workshops held in HUPO Annual Congresses and American Society of Nephrology Renal Weeks, we are planning to draft HKUPP Standards and Guidelines for Urinary Proteome Analysis. These guidelines are really needed for current situation under which rapidly increasing number of investigator are trying or applying urinary proteomics in aims of biomarker discovery. In this presentation, I will outline the current concerns pertaining to urine proteomics with special emphasis on variability of urine proteome, and provide our initial HKUPP Standards and Guidelines for collection, handling, storage and treatment of urine to perform urinary proteome analysis in the more standardized way. Our aim is neither to use these standards and guidelines as the rules and regulations for publication in proteomics journals nor to discourage analysis of the archival urine samples collected for a long time of period. In the standards and guidelines, we keep in mind that they must be: i) generalized for all of most of techniques; ii) generalized for all or most of laboratories and iii) balanced between perfection and practicability.

View full abstract

Biomarkers in some studies are profiling-based and have limitation in clinical use, particularly due to the lack of reproducibility because of variations in sample preparation and difficulty in assay development. We have analyzed serum proteins of liver disease patient cohorts by 2D-PAGE and 2D-LC MALDI-TOF MS with respect to expression levels and different isoforms generated by alternative splicing, cleavage by proteases or posttranslational modifications. In clinical proteomics, key issues including sample collection and preparation prior to proteomic analyses or diagnostic assays have been addressed.
1. Depletion of major plasma proteins: The proteins and peptides circulating in serum/plasma are thought to be a source of potential biomarkers for the disease. However, the major plasma proteins abundant in the blood mask the presence of these circulating biomarkers indicative of the disease state, making it difficult to discover such biomarkers in the blood.
2. Evaluation of biomarkers: Several plasma/serum proteins are changed in terms of concentrations and composition with the physiological state of individuals. As accuracy of measurement of biomarker proteins is required, intra-individual deviation should be independent of the methods of sample collection and preparation.
3. Utilization in clinical laboratory: We must standardize the methods of sample collection and storage in routine work at clinical laboratory for detection of protein/peptide biomarkers.
The object of this presentation is to review the methods of sample preparation and storage for serum proteomics and discuss the strategies to improve the detection of proteins in low abundance which have the potential to discriminate between health and disease conditions. A further object is to review those points to consider for the development of clinical assays for proteomic biomarkers.

View full abstract

Proteomic profiling of mouse liver microsomes was performed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and nano-liquid chromatography tandem mass spectrometry (nano-LC-MS/MS) for the pilot study of Membrane Proteome Initiative (MPI) of AOHUPO. The microsomes were prepared by two-step centrifugation of liver homogenate, firstly the homogenate was centrifuged at 15,000 g for 15 min to yield supernatant and secondly the supernatant was centrifuged at 132,000 g for 60 min to yield microsomes in a pellet. The microsome pellet was suspended in 0.1 M sodium carbonate solution containing 0.5 mM PMSF and 10 ug/mL of aprotinin and leupeptin to yield highly purified microsomes (J. Cell Biol., 1982, 93, 97-102) in a pellet by centrifugation at 132,000 g for 60 min. The purified microsomes were then treated with ice cold acetone-methanol (8:1) solution (Anal. Biochem., 1999, 273, 313-315) to remove lipids prior to SDS-PAGE with a gradient gel (5-20%). The proteins in the microsomes were finely separated into more than 50 bands which were visualized by Coomassie Blue staining. The protein bands were cut out and submitted to in gel digestion with trypsin. Peptides in the tryptic digests were separated by nano-LC followed by MS/MS to identify the protein(s) in each of the bands. More than two hundreds proteins were identified and most of those proteins have been reported to be membrane associated proteins such as microsomal glutathione S-transferase 1, cytochrome P-450 and low density lipoprotein receptor-related proteins. This platform could be used for ER proteomics of Jurkat cells of a human T lymphoma cell line to find marker proteins for ER stress. We show the proteomic profiling of microsomes from Jurkat cells and discuss the standardization of sample preparation for ER proteomics.

View full abstract

Ribonucleic acids (RNAs) are continuing to attract increased attention as they are found to play pivotal roles, often as a part of ribonucleoprotein (RNP) complex, in a variety of cellular processes such as transcription and translation. Thus, isolation and characterization of novel RNP complexes, which contain micro or small non-coding RNAs, are crucial to understand the biological systems. The analysis of RNAs in RNP complexes depends mainly on the techniques in genomics, which include the process of reverse transcription from RNA to cDNA. Therefore, they often suffer from the error of the reverse transcriptase reaction arising from both RNA secondary structures and posttranscriptional base modifications, which is difficult to characterize by standard RNA analytical approaches. LC-MS based analytical platform for tandem mass spectrometry data could offer a number of advantages for RNA analysis for its potential to provide efficient solutions to analysis of the sequence and modification of RNA using limited amounts of sample with mapping interactions between proteins and RNAs. Here we describe our attempts to develop a mass spectrometry-based analytical platform for small RNAs and RNP complexes with efficient method for sample preparation. Our approach is essentially based on the instrumentation for shotgun proteomics, where the direct nanoflow LC is coupled with high resolution Q-TOF/IT-FT hybrid mass spectrometer. In addition, we developed a unique computational tool to correlate a tandem mass spectrum with a nucleotide sequence of RNAs, thus allowing automated MS-based identification of RNAs in complex biological mixtures.

View full abstract

A huge amount of formalin-fixed paraffin-embedded (FFPE) tissues have been reserved in hospitals and medical research centers. Until recent years those materials have been utilized only for limited purposes such as DNA analyses and immunohistochemistry verification studies. We have developed application of Liquid Tissue technology, which solubilizes proteins in FFPE tissue samples, for biomarker discovery and verification. An use of human FFPE tissues has a great advantage because of their already known clinical outcomes. AS a protocol in Tokyo Medical University Hospital, tumor blocks are formalin-fixed immediately after their surgical excision. We have been experiencing a promising results from proteomic analyses of FFPE tissues following their laser microdissection (LMD), LT solubilization, and trypsin digestion. A standard preparation protocol for FFPE samples will be presented for further proteomic analyses, and some recent results obtained from discovery cancer biomarkers investigations and their verification/validation using multiple-reaction monitoring (MRM) MS-based assays will be exemplified.

View full abstract

Glycan analysis by mass spectrometry (MS) is an emerging field in glyco-proteomics, poised to meet the technical demand for elucidation of the structural complexity and functions of the glycan components of molecules. Considering the divergence of the mass spectrometric methods employed for glycan analysis, it is necessary to establish sample preparation and technical standards.
In the present study of the HUPO HGPI (Human Proteome Organization Human Disease Glycomics/ Proteome Initiative, http://www.hgpi.jp/), the same samples of transferrin (Tf) and immunoglobulin G (IgG), both of which have N-glycans, were analyzed for their glycan structures and their relative abundances in 20 laboratories, and the chromatographic and MS analysis results were reported.
The blood samples were obtained from three healthy Japanese donors with the permission from the Medical Ethics Committee of Osaka University Graduate School of Medicine. Tf and IgG were purified from individual serum by immunoaffinity with rabbit polyclonal antibody against human Tf, and by protein G affinity chromatography, respectively, and then lyophilized. The purity of distributed samples was validated by SDS-PAGE under reducing conditions. The stability of these materials during transport was guaranteed by test incubation, during which neither degradation nor modification of glycan moieties was observed after one week of storage at 37 degrees centigrade.
Glycomic and glyco-proteomic methodologies of the type assessed in this study, have the potential to provide large volumes of quantitative data for diagnostics and, more broadly, for underpinning functional glycomics and systems biology research.
A part of this work was supported by the 21st century COE Program of Osaka University from the Japan Society for the Promotion of Science (JSPS) and by the JSPS Core-to-Core program.
REFERENCE
Wada Y, et al. Glycobiology. 2007, 17:411-22.

View full abstract

Recent advances in the proteomic technology in search of biomarker or biomarker signature that accurately identifies the clinical syndrome would allow for improved diagnosis and disease monitoring. Diagnosing Multiple Sclerosis(MS) and other demyelinating diseases such as neuromyelitis optica (NMO) is not always easy clinically and the pathogenesis of the two demyelinating diseases remains mystery. Here we adopted CLINPROT system, a magnetic bead-based purification followed by MALDI-TOF MS to profile human cerebrospinal fluids proteins and peptides. CSF samples from patients with definite MS (remission and relapse), NMO (remission and relapse) and primary progressive multiple sclerosis (PPMS) were analyzed. We used a reagent set of chemically coated magnetic beads, reversed phase ClinProtTM and a- cyano-4-hydroxycinnamic acid as the matrix solution. The eluted samples were then dropped onto a MALDI sample plate (AnchorchipTM), and spectra were obtained by an Autoflex II and a subsequent tandem MS analysis was performed by Ultraflex (Bruker Daltonics). The criteria for peak detection were: signal-to-noise ratio >5, 2-Da peak-width filter. The pretreated data were graphed as spectra and evaluated by statistical analysis using the ClinProToolsTM software (Bruker Daltonics).Reproducible profiles were obtained as clear signals and approximately fifty peaks were detected from each of CSF samples. A differential distribution of samples from MS and NMO both in remission was noticeable, while samples from PPMS were not separated effectively using the same platform. One of the key variables contributing to the classification of the samples with an m/z of 3,511 was defined as c-terminal fragment (182-212) of neuroendocrine peptide 7B2 by the tandem MS analysis. The application of CLINPROT system for CSF samples holds the potential to advance our understanding of the biochemical basis of MS and NMO.

View full abstract

Epithelial-mesenchymal interactions play a pivotal role in hair morphogenesis. We hypothesized that the dermal papilla (DP) cells may secrete soluble molecules that induce the differentiation of epidermal cells into hair follicle in vivo as well as in vitro. To comprehensively identify these molecules, we performed a secretomic study by comparing the secreted proteins of cultured rat DP cells with those of rat fibroblasts.
These two types of cells were serially subcultured in serum-containing growth media as previously reported and then cultured in a special serum-free media for additional 48 h. The 48h-culturing serum-free media were harvested as conditioned media (CM) that contained proteins secreted by each of the cell types and subjected to scretome analysis. The CM proteins were separated by 2D PAGE. Comparison of the protein distribution profiles between the two types of cells enabled us to successfully identify 25 proteins that were exclusively expressed in DP cells. Among them MMP13 was worthy of further characterization because its expression was exceedingly high in DP cells and tissue inhibitor of metalloproteinases 1 (TIMP1) was one of the preferentially expressed proteins in DP cells.The expression of MMP13 mRNA was also markedly upregulated in DP cells compared to fibroblasts.
To investigate the function of MMP13 in the DP cell-dependent hair induction, we transplanted DP cells and rat epidermal cells into the back of nude mice in a manner that the transplants contained varied amounts of TIMP1. The results of the graft reproducibly showed that the hairs induced by the graft cells were suppressed in a TIMP1 dose-dependent manner. These data indicated that MMP13 activity is necessary for sustaining the hair inducing ability of DP cells and consequently identified MMP13 as a protein that functions in the process of hair induction from epidermal cells by DP cells.

View full abstract

Urinary stones of >60% urolithiasis patients contain calcium oxalate as the major mineral component (Tsujihata M, Int. J. Urol. 15, 115-120, (2008)). Supersaturation of calcium oxalate in urine is one of the causes of urinary stones, but it does not result in accumulation of calcium oxalate deposits in non-urolithiasis patients, suggesting that there should be some physiological systems that prevent growth of calcium oxalate stones in urinary tracts. Calcium oxalate stones contain a low amount of proteins, some of which have been shown to stimulate or prevent stone formation in vitro and in vivo (Watts RWE, Q J Med, 98, 241-246 (2005)). Thus identification of proteins involved in calcium oxalate stones would provide key information on the pathophysiology of urinary stones, but the proteins have only partly been characterized possibly due to their low abundance and high heterogeneity, in addition to solubility problems of the stones. In the present study, we have solubilized calcium oxalate stones using EDTA and obtained their protein profiles by shotgun analysis with nano-LC/MALDI-TOF/TOF mass spectrometry. As a result, more than 90 kinds of proteins including well-known stone components such as uromodulin, osteopontin, prothrombin, and S100A8/A9 were identified in calcium oxalate stones from 9 urolithiasis patients. In addition, protein profiles of the individual stones were not fixed, and could be classified into at least two groups, based on the major protein components. These results suggest that multiple systems, composed of different sets of proteins, are involved in prevention of calcium oxalate stone formation in the kidney.

View full abstract

Formation of metaphase chromosome is essential for higher eukaryotes to ensure proper and accurate segregation of genomic information into two daughter cells. In order to clarify the structural basis of chromosome formation, we have carried out proteome analyses of human metaphase chromosomes and identified more than two hundreds of chromosome proteins. Comparative proteome analyses using the chromosomes from different preparation methods as well as from different cell lines eliminated proteins that are not essential for the higher order structure of chromosomes. Different types of electrophoresis were employed for the efficient and high-resolution separation of constituent proteins. Furthermore, absence of chromosome titin, which is a giant protein with the molecular mass of 3MDa and had been reported as a protein contributing to chromosome higher-order structure, was confirmed by proteomic analysis in addition to transcript analysis. Finally we established a chromosome four layer model where all chromosome proteins are classified into one of four different groups. Our recent functional analyses on chromosome proteins in vivo suggest the proteins in the different groups have different roles.

View full abstract

Under the assembled proteomic analysis system, we have been studying the role of tumor related gene functions in cancer cells with two quantitative proteomic strategies. In this study, we focused on the function of neuronal tumor suppressors in NGF induced neuronal model cell PC12. To obtain a new set of functional protein information related to the biological processes, we identified specific proteins in PC12 cells after treatment of NGF and a siRNA of NF1 neuronal tumor suppressor by assembled proteomic strategies, such as differential analysis with iTRAQ (4-Plex) methods and differential phospho-protein analysis with 2D-DIGE coupled with proQ Diamond methods. At the moment more than 5000 non redundant cellular proteins were identified from the combined data by both of LC-MALDI and ESI-MS/MS analysis, and out of about 1500 proteins semi-quantitatively identified, 72 proteins were uniquely identified as differentially expressed in response to NGF and NF1 siRNA stimulations in PC12 cells. Parallel analysis was performed with affinity cellular mapping to identify 58 NF1 specific associating proteins by iTRAQ analysis, and the data obtained from all of the analysis were merged together and the most interesting signal molecules being expected to be drug targets were extracted after several mining protocols. We focused the novel neuronal cell survival, motility, and differentiation related protein cascades. The biological validation was performed by the 2D-western blotting and Time laps and confocal fluorescent microscopic analysis with using the specific antibodies and their inhibitors. The results suggest that the proteins identified in this study are strongly related to neuronal cellular differentiation, survival, and pathogenesis of NF1 with a specific kinase cascade such as CDK5, GSK-3b and Rho kinase, and assembled proteomic analysis are useful for the comprehensive functional proteome analysis to understand cellular biological mechanisms.

View full abstract

Systemic lupus erythematosus (SLE) is a systemic autoimmune disease that affects multiple organs. Involvement of the central nervous system (CNS) in SLE is called CNS lupus. In the CNS lupus, the disease state of neuropsychiatric manifestations without clear organic abnormalities, that is, acute confusional state, anxiety disorder, mood disorder, psychosis, and cognitive dysfunction, is further classified into neuropsychiatric lupus (NPSLE). Recently, existence of autoantibodies to neuronal cells, called anti-neuronal antibodies, has been reported in NPSLE. Even though they may be related to the pathogenesis and/or diagnosis of NPSLE, further investigations have been hampered because target antigens for the anti-neuronal antibodies have not been identified. Therefore, we here tried to identify the antigens using 2-dimensional electrophoresis and western blotting. We first separated proteins extracted from cultured human neuroblastoma cells (SK-N-MC) and those from cultured human leukemia cells (HL-60) separately, We transferred them onto the PVDF membranes, and then reacted them with anti-neuronal antibody-positive serum samples. By comparing the two panels and selected protein spots which were positive only in the panel of SK-N-MC as candidate for neuronal cell-specific autoantigens. We detected eleven protein spots and successfully identified five of them by MALDI-TOF/TOF mass spectrometry They are tubulin beta, heterogeneous nuclear ribonucleoprotein H, nucleophosmin, pyruvate dehydrogenase beta subunit and thioredoxin peroxidase 4. We are now promoting preparation of recombinant proteins for them to investigate the autoantibodies to them in larger numbers of patients together with further identification of the remaining candidate antigen spots.

View full abstract

We previously identified a plasma biomarker set that was able to distinguish pancreatic cancer patients from healthy controls with high accuracy (Honda et al., Cancer Res., 65:10613-22, 2005). Here we report the preliminary results of an ongoing validation study using 549 plasma samples [pancreatic cancer (n=227), gastric cancer (n=68), colorectal cancer (n=31), other cancers (n=2), benign pancreatic disorders (n=21), other benign disorders (n=8), and healthy controls (n=192)] collected at seven medical institutions. All the samples were blinded and processed randomly using a robotic magnetic bead-based hydrophobic interaction chromatography system. The processed samples were randomly allocated to 16 wells of 384-well MALDI plates, and MS spectra were obtained within the range 1000-30,000 m/z using a high-resolution hybrid quadrupole MALDI-TOF-MS. As a result of the highly repetitious measurement protocol, the correlation coefficient (CC) and coefficient of variance (CV) values for all the corresponding MS peaks (>2200 peaks per sample) were constantly kept within 0.993 and 0.0392, respectively, throughout the experiments. We identified the amino acids sequence of previously two reported biomarkers (17252 m/z and 8765 m/z). In addition, we confirmed that the intensities of two biomarker proteins combination (17252 m/z and 8765 m/z) differed between pancreatic cancer patients (n=227) and healthy controls (n=186) with high statistical significance [Mann-Whitney U test, P=1.11X10E-42 and AUC (area under curve) value >0.888]. In order to analyze a sufficient number of plasma samples and establish the clinical utility of the biomarkers, prospective collection of well-characterized clinical samples will be necessary. A collaborative study involving 7 medical institutions throughout Japan is now underway. This collaborative study is part of the "Third-Term Comprehensive Control Research for Cancer" conducted by the Ministry of Health, Labor and Welfare of Japan.

View full abstract