The Takeda Award 2001


1. Development of high-throughput DNA sequencers
1.1 Sequencing of genomes Michael W. Hunkapiller produced a fully automated high-throughput multi-capillary DNA sequencer, which has a 10 fold higher productivity than conventional slab-gel DNA sequencers. The genetic information in the DNA is stored as a sequence of bases that encodes a protein and DNA sequencing is the determination of the exact order of the base pairs in a segment of DNA. A genome is a complete set of all genes and other DNA that does not code for proteins in the chromosomes. A genome is not only a complete blueprint of a living cell but also a history of evolution experienced by the organism before, and the sequencing of the genome can lead to both the elucidation of gene function and an understanding of evolution. Hunkapiller recognized the importance of automated and high-throughput analytical instruments for large scale genome sequencing, predicting that an era of the large scale genome sequencing was coming. 1.2 Development of first DNA sequencing technology In 1977, the first techniques for DNA sequencing were reported independently by A.M. Maxam and W. Gilbert(1), and F. Sanger and co-workers(2). Later they received the Nobel Prize for the development of DNA sequencing techniques. The Maxam and Gilbert method involves the use of base selective degradation reactions to produce radio-isotope labeled fragments, which are then submitted to slab-gel electrophoresis to determine the sequence. In the Sanger method, fragments to be sequenced are reacted with radio-labeled short DNA chains (primers) and deoxynucleotides mixed with chain-terminating dideoxynucleotides to produce radio-labeled fragments. Since the reaction with one kind of chain-terminating dideoxynucleotide gives a sequence order concerning one kind of deoxynucleotide, four different reactions with four different kinds of chain-terminating dideoxynucleotides are necessary to determine the complete sequence order of a given fragment. The radio-labeled fragments are submitted to slab-gel electrophoresis and bands are detected by autoradiography (Figure 1. (a)). The sequence is interpreted from the pattern of alternative bands in the lanes corresponding to the terminal base of the fragment. These methods provided a powerful tool for DNA sequencing, but are very tedious and labor-intensive. 1.3 Development of fluorescent detection method An advance in sequencing technology occurred in 1986 and 1987 when L. E. Hood and coworkers at Caltech(3), W. Ansorge and coworkers at EMBL(4), and J. M. Prober and coworkers at DuPont(5) independently reported laser-induced fluorescence detection methods for DNA bases. Their methods involve the use of fluorescent labeled primers to produce fluorescent labeled fragments. Since the fluorescent methods do not involve the use of hazardous radio-isotope labeled primers and time-consuming autoradiographical treatment, they replaced the radioisotope labeling method. Hunkapiller was a post-doctoral fellow in Hood's laboratory at Caltech, where he participated in the development of a fluorescent detection method using four dyes with four different absorption spectra (Figure 1 (b)). In this method, bases are monitored by detecting fluorescent light near the end of the slab-gel while the gel is running. This real time monitoring led to the development of automated high-throughput DNA sequencers. Later, Hunkapiller joined Applied Biosystems, Inc. (ABI), and the Caltech method has been successfully used in the detection systems of the DNA sequencers developed by ABI(6). The productivity of DNA sequencers increased several hundred folds after the development of the fluorescence detection methods for DNA bases. 1.4 Development of high-throughput multi-capillary DNA sequencers However, even after the development of fluorescent detection methods, the productivity of DNA sequencers was not sufficient to decode the human genome in a limited time. Another advance was necessary to achieve higher productivity. This occurred when capillary electrophoresis was introduced for DNA sequencing(7-9). The separation speed is limited in slab gel electrophoresis because of the heat produced when the high electric field is applied to the gel. Since heat elimination is very rapid from the large surface area of a capillary, a higher electric field can be applied to capillary electrophoresis, thus speeding up the separation process. By using a capillary gel, the separation speed is increased about 10 fold over conventional slab-gel systems (Figure 1. (c)). Treating multiple samples at the same time is essential for high-throughput. This can be achieved by employing multi-capillary systems (Figure 1(d)), but a serious problem arises from the detection of DNA bases in multi-capillary systems. The detection of fluorescence from DNA bases is complicated by the scattering of light from the porous matrix and capillary walls. R. Mathies developed a method that uses a confocal fluorescence scanner to avoid light scattering(10). Using this confocal detection method, Molecular Dynamics commercialized the first multi-capillary DNA sequencer, the MegaBACE1000, in 1997. Another solution to this problem was provided independently by Hideki Kambara of Hitachi Corporation and J. C. Dovichi of the University of Alberta. Essentially, their method involves the use of a sheath flow, which is a hydrodynamically focused stream from the outlet of the capillaries, and the direct detection of the fluorescence from the sheath flow in the sheath flow cuvette (Figure.2) (11,12). Since this method does not use capillaries in the fluorescence detection part, the detection does not suffer from light scattering from the capillary walls, and the sensitivity is greater than that of confocal detection. This method also allows the introduction of laser from the side wall of the cuvette, which simplifies the laser introduction apparatus. Hunkapiller was in overall charge of the development of DNA sequencers at ABI. While struggling with the development of multi capillary electrophoresis, he combined the ABI proprietary technologies of high-throughput multi-capillary electrophoresis with fluorescent dye chemistry and an automated sample exchange system, and the sheath flow detection method, acquiring a license from Hitachi and the University of Alberta, and put the automated high-throughput multi capillary DNA sequencer, PRISM3700, on the market. The PRISM3700 uses 96 capillaries and its overall productivity is 10 times greater than the most productive slab-gel DNA sequencers.