Vol.1 No.2 2008

49/85

Research paper : A systematic analysis of protein interaction networks leading to the drug discovery (S. Iemura et al.)−118 Synthesiology - English edition Vol.1 No.2 (2008) as functional analysis of causal and related genes for cancer, life style disease, neurodegenerative disease, xeroderma pigmentosum, Down syndrome, Behcet’s disease, and essential hypertension. We also discovered several cases in which proteins that were considered to have absolutely no relationship to disease may become totally new drug discovery targets[2]-[14]. I shall describe a representative example of network analysis in the following section.There are huge protein complexes called proteasome in the cell, functioning as factories that degrade unnecessary proteins. The complexes consist of over 60 protein components, and it was long unknown how they were assembled. We discovered the assembly factors that assembled the proteasome. This was academically a significant discovery, where huge protein complexes are constructed with collaborative support of other proteins[4][11]. As shown in Figure 3, an assembly factor called DSCR2 and HCCA3 cooperated and arranged the subunits in ring form. Next, Ump1 and MGC10911 created the ring structure, and and rings joined in a correct alignment. At the same time, this was a new drug discovery target.Proteasome not only has a role of “quality control” or breaking down old and denatured protein, but it also has an important function of controlling diverse vital protein reactions. In cells, when new proteins become necessary for some vital reactions, it may be too late if they are synthesized as need arises. The cell continuously makes proteins that are expected to become necessary until the situation arises, and proteasome continuously degrades them. When the moment arrives when proteins become necessary, the degradation is halted and the necessary proteins can appear immediately.For example, when cells divide, several proteins must cooperate in one direction and work closely together. This is lead by proteasome. Cancer cells that continue to divide indefinitely are thought to require more proteasomes than normal cells. It has been known that drugs that inhibit the action of proteasome possess powerful anticancer effect. However, severe side effects appear when such inhibitory drug is used, since proteasomes are necessary for a normal cell function. Therefore, this inhibitor is used only with special cancers with no other treatment. However, when the function of assembly factor of proteasome that we discovered was inhibited, the amount of newly created proteasomes decreased, and there was hardly any effect on normal cells although it was fatal for cancer cells. Since normal cells do not require as much proteasomes as cancer cells, they are less susceptible to some decrease in proteasome level. Also, we expected that that there will be fewer side effects, unlike complete inhibition of proteasome function. This assembly factor would be a new and more suitable drug discovery target.Since these results could be directly applicable to drug discovery research, we suggested corroboration research with pharmaceutical companies in 2006. As a result, we started a drug discovery research project in which almost all Japanese major and medium pharmaceutical companies participated. The initial grand design of this project was based on protein network analysis of disease related or causative genes or proteins in which each pharmaceutical company was interested, for drug target discovery. However, the project was taken further, and aimed at establishing drug screening platforms under corroboration of pharmaceutical companies based on the information of protein network analysis. When the project finds ‘hit’ compounds, AIST provides the hit information for corroborative companies for developing therapeutic drugs. To do this efficiently, Computational Biology Research CenterNote in AIST Tokyo Waterfront also got involved in the project to bridge hit compounds and combinatorial chemistry by in silico simulation using the hyper parallel computer system. Eventually, to enhance activity of the industry, the project was designed[15] to provide research resource and facility which can not be equipped in each single private sector, such as mass spec facility, large-scale natural compound library and computatonal resources like Blue GeneTerm 1.6 Discussion: strategy toward Full ResearchThe most basic strategy that we implemented when we planned and started the protein network analysis project was: “we shall not aim for eccentric and extraordinary innovation.” Even if we created a wonderful technology or technique, it normally takes at least 10 years before it is standardized as a analysis method and begins to generate data. In fact, it was in the early 1980s that Mr. Koichi Tanaka ionized peptide protein using matrix and conducted mass spectrometry for the first time in the world. This discovery led to the development of MALDI mass spectrometry, which the protein chemists and biologists around the world started to use in the late 1990s to 2000.At the time, we thought spending 10 years developing analytical technology was unrealistic. We adhered to the most realistic, most down-to-earth, and “straightforward” way of doing things, that is, to minimize the greatest bottleneck in current mass spectrometry. The “straightforward” method was to “transfer microquantity of sample to mass spectrometer without loss,” and we focused thoroughly on this issue. We decided not to dip our hands into new innovations such as improvement of mass spectrometer or increasing efficiency of ionization. The mass spectrometer itself was already highly sensitive, and we placed our bet on a hope that if ionization could be accomplished without loss of sample, we could obtain the target sensitivity.If we could achieve ultra high sensitivity, large-scale analysis at high throughput would become possible. The (46)−

※このページを正しく表示するにはFlashPlayer9以上が必要です