Development of a Precise Sugar-Chain Profiling Scanner Using a Lectin Arraying Technique

- As a tool for high-speed early diagnosis of cancer -

Key Points

  1. We have arrayed 40 kinds of lectins (glycol-proteins), whose specificity has already been clarified, on slide glasses.
  2. Weak interactions between sugar chains and lectins can be clearly detected in real time.
  3. We hope that our technique will play an important role in sugar chain-related fields of medicine, such as early diagnosis of cancer or rapid sorting of stem cells for regeneration medicine.


The National Institute of Advanced Science and Technology (AIST, President: Hiroyuki Yoshikawa) has developed the first scanner using the evanescent wave-excited fluorescent detection method and a sugar chain profiling system utilizing lectin micro-arrays in the world. This study was conducted as a part of the Glyco-Engineering Project sponsored by the New Energy and Industrial Technology Development Organization (NEDO).

The development was carried out in collaboration with Moritex Co. Ltd. (CEO, Shigeyuki Morita) to fabricate a high-performance lectin array system by combining a specified scanner of Moritex (SCAN III, the former Nippon Laser & Electronics Lab Co., Ltd., incorporated with the Moritex in 2004) with a lectin fixation technique developed by AIST.

Our technique can be applied widely not only to refined products, such as sugar chains, glyco-peptides and antibodies, but also to clinical specimens, such as blood, and thereby the application of glyco-protein engineering to medical technologies can be further accelerated. This work will be published in the latest issue (November) of Nature Methods, and will be presented at the Biotechnology Symposium to be held at the Toranomon Pastral on November 22.

Figure: Lectin micro array Figure: Sugar chain profiling scanner Figure: Profiling data
Lectin micro array
Sugar chain profiling scanner
Profiling data

Background and History of Research Work

After the human genome sequences were determined, proteome analyses attracted attention as post genome studies. However, information polymers called "sugar chains" are attached to most of the proteins working in the human body, and so if merely the bare proteins are investigated, the mechanisms of life functions can never be understood.

Sugar chains are considered to control the cell society by determining the stability of proteins and their destination (where to go in cells) and by regulating the functions of proteins. If the sugar chains are not formed, generation and differentiation do not normally occur, and nerve or kinetic systems also work abnormally.

On the other hand, not all the sugar chains attached to proteins necessarily play an important role; for example, blood-typing sugar chains affect the sustenance of life very little. As sugar chains are deeply related to all life actions, the total picture cannot be explained only by partial phenomena.

Sugar chains on the cell surface are called "the face of cells," and have different characteristics depending on the cell kind and state, respectively. This is because the switch of a gene group synthesizing sugar chains (sugar-chain genes) is different with respect to each cell. Also, bio-markers such as tumor markers can provide very important diagnostic guiding principles for our life, particularly in the present aging society.

Most of these biomarkers have been thought to be sugar chains. Thus, it is important to seek for functional sugar chains closely linking to bio-functions such as cancer and generation among the sugar chains attached to proteins. For such a purpose, rapid, convenient, and precise techniques to analyze sugar chain structures are necessary. However, sugar chains have far more complicated branching structures than the unicursal structures of proteins or nucleic acids.

However, the determined genome sequences do not enable to prediction of the sugar chain structures bound to proteins, and thus direct analyses of sugar chains are essential. Until now, many life science researchers have left analyses until later, because of their difficulty and complication.

Recently, the significance of sugar chains has been recognized, and the worldwide competition in technical developments of sugar chain structural analysis is getting severe. Of all, the U.S. has structured a world-level consortium on the basis of the hefty budget of the NIH (National Institute of Health), and produced marked results.

As one of the driving forces to promote sugar chain analyses, arraying techniques of sugar chains (sugar chain array or tip) have been developed here and many other countries. The sugar chain arrays are very useful for structural analysis of glyco-proteins such as lectins and toxins, but cannot be directly used for structural analysis of sugar chains.

Usually, to directly analyze sugar chain structures, tandem-type mass spectrometry methods are used. However, such methods are not necessarily all-around. In fact, identification of isomers and the analyses of not only whole glyco-proteins and -peptides but also mixed substances are also difficult.

This research was carried out with the support of the "Glyco-Protein Engineering Project" by NEDO.

Details of Research Work

We used a sugar chain profiling method to analyze the chain structures. This method prepares data sets presenting binding states of many kinds of proteins (lectins) which can easily bind to sugar chains. One data set means a set of binding forces between a sugar chain and several dozen kinds of lectins. As the data sets have various structural characteristics of sugar chains, key points of structural information of sugar chains (e.g., degree of branching, binding types, and presence or absence of terminal modification) can be quickly extracted by comparing with those of previously prepared data sets, leading to rapid analyses of many specimens.

As the binding forces between sugar chains and lectins are comparatively weak, after their binding reaction on arrays, if wash-processing (usually done for DNA arrays or antibody arrays) is performed, the sugar chains can easily be detached from the arrays. The binding constants for antigen-antibody reactions are 108-109 M-1, while the constants for lectin-sugar chain reactions are 104-107 M-1.

We noticed a special light, called the "evanescent light wave" (near-field light), leaking in the range of several hundreds nm over the substrate surface, and thought that, if utilizing the wave could be utilized as an excited light for fluorescence, only the fluorescent sugar chains binding to lectins fixed on slide glasses could be detected without washing.

We undertook joint research with Nippon Laser & Electronics Lab Co. Ltd. (at present, Moritex Co. Ltd) because they had the techniques necessary for this research. As a resultant, we have succeeded in a real-time, high sensitive detection of bound sugar chains or glyco-proteins labeled by a green-colored fluorescent reagent, Cy3, on lectin micro-arrays, as shown in the figure.

The binding constants of sugar chains (glyco-proteins) to 40 kinds of lectins can be estimated altogether in theory. Although there are several groups developing lectin micro-array techniques in the world, our micro-array system is unique, because it has high-precision profiling performance, which is able to clearly detect even weak, the binding interactions.


In addition, our lectin micro-array system has a further considerable advantage, that is, precise data concerning sugar-chain specificities of 40 kinds of fixed lectins, obtained by us using a frontal affinity chromatography method we previously developed (see AIST Today (Newsletters) Vol.4, No.1).

Using this technique, we have already determined the binding constants of more than 100 types of standard sugar chains to more than 100 kinds of lectins, and thus we can evaluate the meaning of signals obtained from the lectin arrays effectively by bio-informatics and with high reliability.

As the sugar-chain affinities of the respective lectins are stored as a data base, we can approximately profile the structural characteristics of sugar chains using data about them obtained from the sugar chain profiling scanner. Lectins specifically recognize a part of sugar chain structures, and thus if only using an array of various kinds of lectins, we can predict the structures of sugar chains by analyzing the patterns of data obtained from the scanner.

The important point is that glyco-proteins with different sugar chain structures have different profiles (binding patterns). Not needing to determine whole complex chain structures leads to a drastic speed-up of sugar chain identification. This point is an extremely useful advantage in seeking for sugar chain bio-markers.

If a change in sugar chain structures is induced by diseases, many specimens must be examined to prove it. High speed profiling of sugar chain structures is the main characteristic of the sugar chain profiling scanner we have developed.

▲ ページトップへ