Some reflections on a career in science and a note of thanks to the contributors of this Special Issue

It is a great honor for me that Biophysical Reviews publishes this special issue in honor of my 70th birthday on a theme relating to my field of scientific research. I would thank the many colleagues who have agreed to participate and who have submitted articles about their recent studies. In this commentary, I will briefly introduce my research activities and outline the philosophy that has driven them.

When I started my academic career, I had two supervisors, Dr. Akiyoshi Wada (Professor Emeritus, Dept. of Physics, Faculty of Science, University of Tokyo) and Dr. Koji Okano (Professor Emeritus, Dept. of Applied Physics, Faculty of Engineering, University of Tokyo). The former, Prof. Wada, is one of the pioneers of biophysics in Japan, and he is famous because of his early conceptualization of the technology for rapid DNA sequencing (realized in the 1970s, a date much earlier than the commencement of the human genome project (Ito 2005)). Prof. Wada once wrote a book with the inspiring title “Physics Crosses Borders” (Wada 2005), insisting that biological science should be reorganized with the physical aspects covering material science, electronics, informatics, and computer science. In my opinion, this prediction has since been proven to be correct. I was a Ph.D. student in his laboratory from 1975 to 1979 and was engaged in developing new devices to measure dielectric relaxation spectra of biopolymers in the time domain (Nakamura et al. 1981a, 1981b). I learned much from Prof. Wada, in particular, that any new study requires new technologies and methods, both in terms of hardware and software. I also made several computational studies of the electrostatic properties of proteins (Wada and Nakamura 1981), which examined the origins of the experimental dielectric relaxation spectra of biopolymers and also went on to develop my own molecular graphics software programs to display the results (Nakamura et al 1985). My other supervisor, Prof. Okano, is a theoretical physicist in the field of soft-matter physics and liquid crystals, and from him, I learned a very great deal about the physical and mathematical concepts governing how an ordered system is produced (Nakamura and Okano 1983).

Dielectric relaxation spectroscopy provides unique information on polymers, such as their dipole moments and fluctuations. It is, in fact, a powerful method for homogeneous systems but is not very informative for inhomogeneous systems like proteins in an aqueous solution. On the contrary, nuclear magnetic resonance (NMR) spectra give much more precise information about the tertiary structures and the physical properties of proteins. In our early studies in the NMR field, we developed our own algorithms for the determination of distance geometry using simulated annealing in the 4-dimensional space (Nakai et al. 1993) and used these to determine several protein structures (Ogata et al. 1992) and protein-DNA complex structures (Ogata et al. 1994). In addition, the precise pK_a values of individual aspartates and glutamates of RNase-HI protein were determined by using ¹H–¹³C heteronuclear two-dimensional NMR (Oda et al. 1994). We found the cooperative features of the ionizable residues, and they could be both simulated and analyzed by a simple electrostatic model based on solving the Poisson-Boltzmann equations numerically, the method of which was originated by my own algorithm that achieved an accurate solution based on Green’s theorem (Nakamura and Nishida 1987). Dr. Kengo Kinoshita later applied this approach to the development of the eF-site database (eF-site 2022), displaying and predicting the functional sites on the protein electrostatic molecular surfaces (Kinoshita and Nakamura 2003).

From July to November 1985, I left Japan to engage in research at the Astbury Department of Structural Molecular Biology with the University of Leeds, UK, as a visiting researcher, and there I learned protein design from the late Dr. Sandy Geddes using molecular graphics and molecular simulation (which were not matured technologies at the time). After returning, I later focused on computational studies at the Protein Engineering Research Institute (PERI) in Osaka, Japan, developing a molecular dynamics (MD) program, PRESTO, initially designed for a vector machine (Morikami et al. 1992). We recognized that stable and meta-stable protein conformations have a free-energy minimum or local minima positioned within a relatively wide phase space. In order to look for those stable and meta-stable structures, enhanced conformation sampling technology was necessary, and Drs. Akinori Kidera and Nobuyuki Nakajima invented their own method, multi-canonical MD (Nakajima et al. 1997), the codes of which were first installed in the PRESTO program. Enhanced sampling procedures have now become standard methods for determining distributions of PMFs (potential of mean forces) as the free energy landscapes (FELs). We applied such procedures to reveal structural ensembles of short peptides (Nakajima et al. 2000) and that of a β-hairpin (Kamiya et al. 2001). More recently, Dr. Junichi Higo and his colleagues succeeded in obtaining accurate FELs for docking flexible peptides to acceptor proteins (Higo et al. 2011; Umezawa et al. 2012; Kasahara et al. 2018; Higo et al. 2020a, b) using even more sophisticated sampling technologies (Higo et al. 2013; Higo et al. 2020a, b). Those studies revealed the mechanism of induced folding of intrinsically disordered regions of signal proteins, and the method was also applied to ligand docking simulations for drug development (Higo et al. 2022; Fukunishi et al. 2022). The concept of landscapes is useful to provide an overview of entire systems, not only within the area of molecular folding but also in the fields of immunology, as the fitness landscape (Furukawa et al. 1999), and also, as recently shown, within the domain of systems biology (Okada 2022). The enhanced sampling method was also applied to QM (quantum mechanics)/MD simulations to draw the FEL of cis–trans isomerization of the proline dipeptide in water, where the peptide was treated by ab initio QM methods (Yonezawa et al. 2009).

Although classical MD seems to be well-matured, the issues of the boundary condition and the force fields used in the MD algorithm are still complicated problems that have not been solved completely. My colleague Dr. Ikuo Fukuda has long been engaged in the former issue by developing his own algorithm, the zero-multipole summation method, as an effective Non-Ewald method (Fukuda et al. 2014; Fukuda and Nakamura 2022). For the latter problem, Dr. Yu Takano and his colleagues recently found the origin of the inconsistency in H-bond energies within α-helices calculated by classical molecular mechanics (MM) against those calculated using more accurate QM by using the negative fragmentation approach (NFA) (Kondo et al. 2019; Takano et al. 2022). In the near future, an effective modification of the available MM force fields is expected.

Aside from direct “first principle” physical calculations, my group has also been interested in bioinformatics-based studies, which use informatics methods to tease out original principles from information stored within databases. For example, with regard to the phenomenon of the low probability of occurrence of Ile within α-helices, my colleague, Dr. Koji Furukawa, rationally engineered a small globular protein by changing an Ile to a Leu residue within an α-helix, and his NMR observation indicated a lack of structural uniqueness caused by greater side-chain conformational entropy exhibited by Leu. Namely, this “negative” protein design principle suggested that some Ile residues are necessary to confer the protein with structural uniqueness (Furukawa et al. 1996). Another example was made by Dr. Ashwini Patil, who was a Ph. D. student in my laboratory at Osaka University. One day during laboratory tea time, I asked a naïve question to her: “Why are hub proteins able to interact with a great many (although finite) number of the partner proteins?” It was not a trivial question, for as far as we believed at that time, specific molecular recognition was caused by interaction between rigid single tertiary structures. In answering this question, she went on to show that disordered domains of hub proteins conferred an ability to interact with multiple proteins in large interaction networks (Patil and Nakamura 2006), an idea that has now become widely accepted (Patil et al. 2010).

Apart from the above scientific studies, I have been engaged in database activities related to the establishment of the PDBj (Kinjo et al. 2012, 2017; Kurisu et al. 2022; PDBj 2022) and the wwPDB (Berman et al. 2003; wwPDB consortium 2019; Burley et al. 2022; wwPDB 2022), helping to maintain and improve the archives of biological macromolecular structures as well as developing and employing methods to keep the data quality high. Dr. Daron M. Standley, who once joined the PDBj, developed a unique service to predict the unknown functions of proteins from their sequences and structures and applied this method in a scientifically objective manner to validate the results of a very large Japanese structural genomics program known as “Protein 3000” (Standley et al. 2008).

Although officially retiring in 2018, since then, I have been acting as the lead administrator of a number of government granting schemes and also working as the Editor-in-Chief of the official journal of the Biophysical Society of Japan, “Biophysics and Physicobiology” (Nakamura 2021; BPPB 2022). It is my sincere hope that both these later administrative and journal-related activities and my earlier scientific contributions have helped to make a positive contribution, especially to the promotion and conduct of “open science” and “open data” in the national and international research arenas.

Declarations

Ethical approval

Not applicable.

Consent to participate

The author agrees to participate in the publication.

Consent for publication

The author agrees to publish the manuscript.

Conflict of interest

The author declares no competing interests.