 |
 |
|
||||||||||||||
 | ||||||||||||||
|
Never Missing the Target Glyco-chain DNA Microarray
|
|||
|
|||
| Diversity of glyco-chains | |||
|
What is the "glyco-chain," which is included in the name of the Laboratory? Kozutsumi explains, "The glyco-chain is a biological structure comprising a chain of monosaccharides such as glucose, mannose, and galactose. Monosaccharides serve mainly as energy sources for cells. On the other hand, glyco-chains function in totally differently ways. Organisms are clever to utilize glyco-chains for a variety of purposes." At the Glyco-Chain Expression Laboratory, research activities are ongoing to elucidate the mechanisms behind biological processes focusing on biologically essential glyco-chains. Then, what are the functions of glyco-chains? "Glyco-chains are also essential to protein production," says Kozutsumi. Most proteins are produced from a chain of amino acids in the rough endoplasmic reticulum of cells on the basis of DNA base sequence information. During this process, glyco-chains bind to the protein. Proteins cannot reach structural completion or function without being folded with the aid of glyco-chains. Without glyco-chains, most proteins fail to be accurately folded or to perform their functions. Additionally, many of the glyco-chains occur in the form of glycoproteins or glycolipids as bound to proteins and lipids in the cell membrane. In the process of signaling, organisms utilize glyco-chains, which protrude from the cell membrane. However, why is the complex signaling possible by glyco-chains, which comprise only a sequence of monosaccharides? "There are only tens of human monosaccharides. However, monosaccharides take various structures depending on the mode of covalent bonding. Moreover, glyco-chains comprise as many as several hundred monosaccharides for the longest. Furthermore, because each monosaccharide has three or more covalent bonding "hands," glyco-chains may have a branched structure in Y- or a cross-shape, with the acetyl group and the sulfate group added to various positions. The great diversity of glyco-chains can be said to be nearly infinite. Because proteins specifically recognize and bind to the wide variety of glyco-chains, complex signaling is possible." |
|||
| Developing a glyco-chain related DNA microarray | |||
|
"Glyco-chain analysis is very painstaking due to the considerable diversity," points out Kozutsumi. "Our Laboratory was founded with the aim of producing a DNA microarray for elucidating the expression mechanism and function of complex glyco-chains." A DNA microarray comprises many DNA species arranged in a regular pattern on a small substrate. By extracting mRNA from the test cell, labeling it with a fluorescent marker, and reacting it with the DNA species on the substrate, the gene expressed and the level of expression can be determined based on the color and intensity of the resulting fluorescence. Then, what are the differences between the glyco-chain related DNA microarray under development at the Glyco-Chain Expression Laboratory and conventional DNA microarrays? "Our DNA microarray is basically similar to conventional arrays," replies Kozutsumi. "However, because glyco-chains under complex control by multiple genes are targeted, it is necessary to develop an analytical method by which the gene control mechanism can be elucidated directly from the data obtained using the DNA microarray. If things go well, our new DNA microarray will find new applications in fields other than glyco-chains." The sequence of monosaccharides that constitute a glyco-chain is not a direct product from genetic information as with protein amino acid sequences. First, a protein known as glycosyltransferase is produced from genetic information, and this is followed by a complex process by the enzyme to biosynthesize glyco-chains. "Because one glycosyltransferase is functioning for each monosaccharide, any sequence of 10 monosaccharides is associated with 10 genes for the respective types of glycosyltransferase. Additionally, there are different decomposing enzymes involved in individual monosaccharides; glyco-chain expression and functions cannot be elucidated unless a great many genes working against this background are examined (Figure 1). We have been aiming at a DNA microarray that enables dedicated analysis of glyco-chain-related genes." Prior to developing or updating a DNA microarray, the Glyco-Chain Expression Laboratory sends emails to other laboratories engaged in glyco-chain research, including 20 to 30 sites in Japan and several overseas. This is intended to accept their requests for the mounting of genes of interest on the DNA microarray. At the Glyco-Chain Expression Laboratory, joint analyses using this DNA microarray take place in response to requests from other laboratories, both domestic and overseas. The only other laboratory engaged in the development of DNA microarrays for sugar-related genes and analysis using them is in the US. "The quickest, simplest and surest way to examine the expression of glyco-chain related genes is to outsource the analysis to our laboratory," says Kozutsumi in confidence.
|
|||
| Complexity and expectation in glyco-chain research | |||
|
At the Glyco-Chain Expression Laboratory, extensive analyses of glyco-chain expression mechanisms and functions are also ongoing based on analytical results obtained using DNA microarrays. Kozutsumi is interested in, and wants to investigate, immunity-related glyco-chains. However, "glyco-chain research is painstaking," again says Kozutsumi. There are several reasons. In DNA base sequencing, DNA may be amplified by the PCR method, or proteins can be produced in large amounts by introducing genes into Escherichia coli cells. However, it is difficult to apply this approach to glyco-chains. This is because a large number of genes are required for the production of a glyco-chain, and also because not all of them can be introduced to Escherichia coli cells. Such genes are often derived from living tissue or cultured cells, so their availability for analysis is very limited. "Although protein structural analysis is also said to be difficult, it concerns three-dimensional structures. In the case of glyco-chains, even two-dimensional structural analysis for monosaccharide sequencing has yet to be fully automated, much less for three-dimensional structures." Elucidating the three-dimensional structures of glyco-chains will make a significant contribution to drug innovation. Most diseases develop due to protein abnormalities. Because many proteins function as bound to glyco-chains, there is the expectation that drugs capable of efficiently enhancing or weakening a particular function of protein will be designed if the three-dimensional structures of glyco-chains are elucidated. Kozutsumi has another aim in his work. "There are many diseases caused by abnormalities of sugar-related synthesis or metabolism. In the case of a glyco-chain comprising 20 monosaccharides, for example, not all of the 20 genes involved in the respective transfer reactions make similar contributions. There should be a gene that plays a key role in determining the functions of the glyco-chain. If it can be identified, it will become possible to control glyco-chain function and to apply it to the treatment of disease, by adjusting the functions of the key gene and corresponding monosaccharide. I want to elucidate the relevant mechanisms using the DNA microarray." |
|||
| Caterpillar fungus and sphingolipid | |||
"We had another objective in establishing our laboratory," Kozutsumi displayed a brown article (photograph at the beginning of this article). "This is a kind of caterpillar fungus parasitic on cicada larva. It is highly esteemed as one of the three rare items in Chinese medicine."In 1994, Kozutsumi and Dr. Tetsuro Fujita (then professor of Kyoto University) jointly studied on ISP-1, a substance possessing immunosuppressive action, from Isaria sinclairii, a kind of caterpillar fungus, and elucidated its functions. Because its action mechanism differs from that of any known immunosuppressant, ISP-1 attracted attention, and one of the derivatives is now under development for clinical applications. That was the first time Kozutsumi met the caterpillar fungus, and since then he has proceeded with investigations to analyze the structure of ISP-1. "When I was aware of the structure of ISP-1, I was reminded of a structure I had once seen. That was sphingosine, one of the constituents of sphingolipid, the lipid moiety of glycolipid (Figure 2)." Considering that ISP-1 might suppress sphingolipid biosynthesis, Kozutsumi began studying its functions. "Just as expected, ISP-1 proved to inhibit the action of serine-palmitoyl transferase, an enzyme responsible for the first stage of sphingolipid biosynthesis. If cells are treated with ISP-1, sphingolipid decreases and the cells die due to apoptosis." Sphingolipid, a kind of lipid that constitutes the cell membrane, was found to play a critical role in cell signaling in the form of a conjugate with glyco-chains, and has recently been attracting attention. However, its functions remain to be clarified in full. At the Glyco-Chain Expression Laboratory, investigations using ISP-1 are ongoing to determine the outcome of the suppression of sphingolipid biosynthesis, and to elucidate sphingolipid functions. Quite noteworthy results have already been obtained. "We discovered a type of cells that survive upon ISP-1 treatment. An examination revealed increased levels of expression of the gene that causes the phosphate group to be attached to glycerophospholipid. Glycerophospholipid is a major member of the membrane lipid family. For survival in settings with decreased sphingolipid, it seems necessary for the phosphate group to be added to glycerophospholipid. We suspect that this requirement may provide a hint in the elucidation of sphingolipid functions, and are making relevant investigations." To date, a total of nine genes for ISP-1 tolerance have been discovered in yeast, which are being analyzed one by one. Sphingolipid has also been found to mediate cytokinesis. When psychosine, a constituent of sphingolipid, accumulates in a cell, the nucleus splits normally but the cytoplasm does not, resulting in a multinucleate cell (upper panel of cover page). "Psychosine has only one monosaccharide, called galactose, attached to sphingosine. This is sufficient to inhibit cytokinesis. I was very surprised at this finding," says Kozutsumi. "We have demonstrated the causality of both sphingolipid and glyco-chains for various disorders. However, the mechanisms beyond them remain a black box. There is ample room for research into glyco-chains, glycolipids, and glycoproteins. This is why I am so interested in my field of work." Kozutsumi emphasizes, "Our DNA microarray never misses the target in elucidating the genetic control of glyco-chains." The glyco-chain related DNA microarray developed by the Glyco-Chain Expression Laboratory should make a significant contribution to glyco-chain research and help open a new field. 
|
|||
 | |||
|
|||
[ Top ] |
The glyco-chain is a biological structure comprising a chain of monosaccharides such as glucose and galactose, which is also called "the third chain of life." "It is much more complex than the first and second biological structures, i.e., DNA and protein. Even a single glyco-chain is associated with a large number of genes," says Yasunori Kozutsumi, Head of the Glyco-Chain Expression Laboratory at the Supra-Biomolecular System Research Group in RIKEN's Frontier Research System. Bound to cell membrane lipid and protein, glyco-chains are responsible for signaling, an essential process for biological activity. Due to their structural and gene control diversity, however, analyzing the expression mechanism and function of glyco-chains is very painstaking. At the Glyco-Chain Expression Laboratory, researchers are working to prepare DNA microarrays that enable comprehensive analysis of the expression of genes related to the diverse complex glyco-chains, and to develop new analytical approaches that will significantly facilitate glyco-chain research. 
|
Developing Tools for Analyzing Brain Signals
|
|||
|
|||
| Separating the information we want to know | |||
|
The brain is a huge and very complex network consisting of more than 10 billion neurons that can generate electric signals. Individual signals are so weak that it is almost impossible for us to measure them directly from outside without damaging the brain. However, when more than about hundred thousands neurons generate electric signals simultaneously, we can measure them outside the brain on the scalp level as changes in electric potentials. Those are brainwaves recorded by EEG or MEG systems. Such brainwaves are a complex mixture of component waves related to all brain activities including those that control breathing and walking and those that allow us to recognize what we see, smell, or hear, in addition to noises related to heartbeats, eye movements, or blinks. Dr. Andrzej Cichocki from the Warsaw University of Technology, Poland, Head of the Laboratory, came to Japan 10 years ago as an expert in signal processing, and organized a research team for analyzing brainwaves in RIKEN. He is a world-recognized researcher along with Dr Shun-ichi Amari, Director of the RIKEN Brain Science Institute, in the research of Independent Component Analysis (ICA), which is a technique to extract the information you want to know from measured signals consisting of various components. ICA is an advanced mathematical method of analysis of signals and images. For example, you can extract an individual picture image from many overlapping images or noise (Figure 1). Using these unique ICA techniques, what kind of information does the Laboratory for Advanced Brain Signal Processing, headed by Dr. Cichocki, try to extract from brainwaves for subsequent analysis?
 |
|||
| Finding the symptoms of Alzheimer's disease from brainwaves | |||
|
The Laboratory for Advanced Brain Signal Processing developed a new method of finding the symptoms of Alzheimer's disease from brainwaves in very early stage of the disease. Alzheimer's disease is a progressive dementia that results in personality destruction due to the deterioration of the brain caused by nerve cell death. The detection rate of the disease increases with advancing age. Thus, our aging society is sure to see a rapid growth in the number of patients, but we have no ultimate cure for the disease, especially in the advanced stage. However, finding the symptoms of Alzheimer's disease as early as possible is important because some drugs that slow the course of the disease are being developed and they are effective only in the early stage of the disease. Dr. Cichocki says, "Early detection of Alzheimer's disease before reaching the stage where the clinical symptom can be found by 'routine diagnosis' is a big challenge. To cope with this, we have proposed a completely new approach, and are subsequently continuing development improving procedures. This new approach refers to a series of signal processing techniques that can be used to detect specific brainwave patterns or markers that are related to Alzheimer's disease several years before the appearance of any clinical symptoms. This approach will exploit the possibility of regular non-expensive group check-ups for middle-aged and elderly people." New diagnostic techniques for Alzheimer's disease, such as positron-emission tomography (PET) and magnetic resonance imaging (MRI), are currently being developed. Diagnosis by these techniques, however, is too expensive to be economically competitive. In contrast, diagnosis based on the approach that has been developed by the Laboratory is much less expensive because the approach is based on the analysis of data measured with a standard electroencephalograph used in normal hospitals. Furthermore, this approach is simpler and safer because it does not involve administration of a compound that is labeled with traces of radioactive isotope as found in PET measurement, or the powerful magnetic field used in MRI measurement. Successful development would lead to the spread of this approach as a means of regular group checkups. The symptoms of Alzheimer's disease begin with debilitating memory loss, apathy, and a state of depression. However, it is considered that several years before the appearance of early symptoms, nerve cells begin to die in the region called the entorhinal cortex and hippocampus, located deep in the brain. If nerve cells die, the brain generates slightly different electric signals, which should be reflected in a time-varying pattern of frequency or amplitude in brain signals. This pattern can be used as a clue to find the early symptoms of Alzheimer's disease. Firstly, noise should be removed to extract exclusively the information we want to know from EEG brainwaves. Then, ICA is used to separate EEG signals into independent physiologically meaningful components. Finally, the separated components are displayed as neuro-images to indicate the portion in the brain to which each component is attributed (lower picture on the cover page). "Among the separated components, you should find three to four brainwave patterns indicating that some nerve cells are dying in the entorhinal cortex," says Dr. Cichocki. The separated component signals are usually very weak. Thus, to detect those weak components, we had no choice but to use a special technique called "additional averaging," which requires measuring the brainwaves about 200 or 300 times for specific stimuli. However, it takes too much time to be applied to regular group checkups. Using ICA, the Laboratory for Advanced Brain Signal Processing aims at a single-trial brainwave measurement to accurately detect the weak components. To find the brainwave patterns that show a sign of Alzheimer's disease, it is necessary to investigate statistically the difference in brainwaves between healthy age-matched individuals and those who showed clinical symptoms of Alzheimer's disease several years later. However, nobody knows who will show these symptoms several years later. Thus, it is necessary for us to regularly measure EEG brainwaves of all elderly people, and to conduct follow-up reviews to see if they showed a risk of Alzheimer's disease. The Laboratory, with the collaboration of medical institutions, collected about 100 follow-up data samples each, and compared them with the previous data. Thus, we found the brainwave patterns that are unique only for persons who showed symptoms of Alzheimer's disease (Figure 2). This pattern can be used as a clue to find persons who will show symptoms of Alzheimer's disease several years before on-set of the disease with about a 92% probability. "Our current two challenges are (1) to confirm whether nerve cell death due to Alzheimer's disease is actually reflected in the newly discovered brainwave pattern, and (2) to confirm the validity of this diagnostic technique both by analyzing large quantities of EEG data gathered in many places in the world, and by using advanced computer simulation technology." The technique should be available not only for Alzheimer's disease, but also for many other brain-related diseases such as Parkinson's disease. It will also serve to evaluate progression of the disease or to assess the efficacy of drug therapy. "There are a few research groups in the world who are trying to diagnose brain diseases such as Alzheimer's disease by using EEG brainwaves using similar approaches that we do. However, we are leading in this diagnosis because we have the unique ICA and several alternative techniques," says Dr. Cichocki confidently. Furthermore, Dr. Cichocki thinks that the ICA techniques, for diagnostic applications, serve to measure biological signals produced in every part of the human body, such as the heartbeat-related electrocardiogram. "For example, the technique can be used to measure the biological signals of a pregnant woman, which are, in turn, used to separate out only the electrocardiogram signals from a featus in the womb. Thus a slight anomaly can be found."
 |
|||
| Operating a computer by brainwaves | |||
|
The Brain Computer Interface (BCI), which is a technique to directly operate a computer by EEG brainwaves, has recently been studied extensively. For example, some successful experiments are reported where electrodes are embedded into the motor area of a monkey's brain to detect the electrical signals generated by the neurons and thereby move the cursor of a computer, or operate a robot arm. For a person who is suffering from a decline in muscle strength, or who has lost motor control due to damage to the spinal cord, the BCI technique could be an important support to their lives. In the United States, there is a trend to put the embedding of electrodes for BCI into medical practice in five to ten years. On the other hand, the Laboratory for Advanced Brain Signal Processing is making efforts to develop BCI safely and noninvasively without the operation. If physically impaired people (after some training) can move the cursor of a computer from right to left or up and down by their brainwaves, and if they can even click on the items they want, they will have no difficulty in using various computer operations, thus leading to improvement in their quality of life. The Laboratory for Advanced Brain Signal Processing analyzed some measured brainwave EEG data and succeeded in moving the cursor of a computer in the desired, directions with about a 70% probability. Performance improvement and real-time operation are the next challenges for practical applications. In other words, our goal is to develop safe and reliable methods to allow the cursor of a computer to move to the specific directions quickly and reliably once you think that the cursor should move in that direction. The Laboratory aims at operating three-dimensional robot arms in the future. |
|||
| Probing psychological states or the thinking process through brainwave analysis | |||
|
Brain science has greatly improved with the advent of PET and fMRI, which are used, to diagnose brain activity with a higher spatial resolution. However, PET and MRI are inferior in respect to time resolution because they detect the change in blood flow or metabolism associated with nerve cell activities. In other words, even if the brain is excited by an external stimulus, those techniques cannot measure brain activity in real time. Furthermore, the head must be fixed for measurement using an apparatus. In contrast, in EEG measurement, a patient is only required to wear a helmet with electrodes in it. Thus the measurement is available for various types of more natural experiments. Because brainwaves are changes in electric potential, brain activity can be analyzed to a time resolution of 1 ms, or higher. However, conventional brainwave measurements have offered only limited information because it is very hard to separate the component signals you want to know from brainwave signals, and to attribute the signals with higher spatial resolution to a portion in the brain. Brainwave analysis based on the ICA technique will without doubt detect brain activities that are difficult to detect by using other methods, and contribute to the development of computational brain science. The Laboratory, using ICA-supported brainwave analysis, is now advancing the research of detailed classification of psychological and mental states such as sadness, fear, happiness, surprise, irritation, and frustration. If brainwaves allow the estimation of those detailed psychological states, such research may attract much attention, although it has been pushed aside because it has been considered unsuitable as a subject of scientific research. Other examples include the measurement of a driver's psychological state such as fatigue or attention, which will be used to automatically stop the car in case of emergency. Furthermore, the evaluation of the impact of advertising will also be possible. The Laboratory has also been using brainwaves to investigate the brain patterns during various mental and perceptive tasks. The brain integrates different information such as sound, image, and smell to make a judgment, or process signals, paying attention only to the important information. The integration mechanism is still one of the mysteries of brain. The Laboratory is trying to clarify the mystery by investigating the change in component signals with high time resolution during signal processing of different information. Dr. Cichocki relates his dream: "We want to analyze the higher mechanism, develop theoretical models, and apply these models to control a computer or a robot." The tools for analyzing brains signals, which are being developed by the Laboratory for Advanced Brain Signal Processing, will have a significant impact on various fields of science including brain and medical science, and robotics.
|
|||
| |||
|
|||
[ Top ] |
It is the goal of the Laboratory for Advanced Brain Signal Processing to develop "new tools for analyzing brain signals and patterns" by using noninvasive brainwave recordings. They are a complex mixture of true brain signals generated from various brain activities. Thus, the information we want to know is hidden in recordings (measurements) on the scalp by using electro-encephalography (EEG). The Laboratory for Advanced Brain Signal Processing consists of several researchers from various backgrounds such as neuroscience, engineering, and applied mathematics. They are undertaking a joint study to analyze complex brainwaves. In other words, they are making efforts to develop a tool for analyzing brain signals. The tool can be used to find some of the markers of Alzheimer's disease by extracting the information we want to know from brain signals by the unique technique of Independent Component Analysis (ICA). These components can be also used directly to control a computer, or to classify psychological mental states and to model complex mechanisms in the brain.