Per-O-methylated β-cyclodextrin (TMe-β-CD) forms a very stable 2 : 1 inclusion complex with water-soluble 5,10,15,20-tetrakis (4-sulfonatophenyl) porphyrin (TPPS) in water. The supramolecular complex provides strong hydrophobic cavity to the porphyrin scaffold, which is similar to the environment of heme in heme proteins. In our laboratory, per-O-methylated β-CD dimer having pyridine linker (Py3CD) was synthesized to make a biomimetic model compound of oxygen-binding heme protein like hemoglobin (Hb) and myoglobin (Mb). The inclusion complex of Py3CD with iron complex of TPPS (FeTPPS) is the first and only-one biomimetic complex that functions in water. The complex, named hemoCD, showed a very high CO binding affinity. When hemoCD was injected to rodent animals after exposure to CO gas, hemoCD captured CO during circulation and was excreted in urine without showing any toxic effect. These properties are quite suitable to the use of an antidote against CO poisoning. Our laboratory has started the drug development to implement hemoCD as a CO antidote for clinical use.

Radiotheranostics integrates diagnostic imaging and radionuclide therapy using radiopharmaceuticals and is a promising approach for personalized medicine. Central to this paradigm is the development of theranostic pairs—diagnostic and therapeutic radiolabeled analogs that exhibit comparable biodistribution. However, constructing halogen-based pairs such as ¹⁸F/²¹¹At remains challenging because of their divergent chemistry and the pronounced in vivo lability of astatine-carbon bonds. Here, we describe a neopentyl labeling strategy designed to address these limitations. The neopentyl group leverages steric shielding to suppress nucleophilic dehalogenation and incorporates hydrophilic functionality to mitigate metabolic degradation. Using stable precursors, including carbamoyl difluoromethanesulfonate (CDf) esters, we established an efficient synthetic route enabling high-yield radiolabeling. The resulting ²¹¹At-labeled compounds demonstrated high in vivo stability and biodistribution closely matching their radioiodinated counterparts, outperforming conventional aromatic labeling approaches. Application of this platform to a PSMA-targeting ligand achieved robust tumor uptake and significant therapeutic efficacy in a prostate cancer model. In parallel, an efficient synthesis of the corresponding ¹⁸F-labeled analog was developed, permitting automated production without HPLC purification. Collectively, the neopentyl labeling platform provides a versatile and stable scaffold for generating ¹⁸F/²¹¹At theranostic pairs and may facilitate clinical translation of targeted alpha therapy.

Oligonucleotide therapeutics have recently emerged as an important modality in drug discovery. The use of artificial nucleic acids is essential for improving both nuclease stability and duplex-forming ability. We previously demonstrated that oligonucleotides incorporating 2′-O,4′-C-methylene-bridged nucleic acid (2′,4′-BNA) exhibit high duplex-forming ability toward complementary RNA. In recent years, we have developed several novel bridged nucleic acids, including 2′-O,4′-C-spirocyclopropylene-bridged nucleic acid (scpBNA), 2′-O,4′-C-spirocyclopentylene-bridged nucleic acid (scpBNA2), and alkyl-substituted guanidine-bridged nucleic acids (GuNA [R]). These analogs provide RNA-binding affinity comparable to or greater than that of 2′,4′-BNA, while dramatically enhancing resistance to nuclease degradation. In particular, scpBNA and scpBNA2 were found to reduce the hepatotoxicity of antisense oligonucleotides. These artificial nucleic acids are expected to contribute to the development of highly potent and safer oligonucleotide therapeutics.

Clinical development of antisense oligonucleotides (ASOs) is frequently constrained by the efficacy-safety dilemma, in which increased target affinity is accompanied by enhanced off-target toxicity. We propose that this apparent trade-off arises from an equilibrium specificity limit inherent to conventional thermodynamically controlled ASO-RNA recognition. To address this limitation, we developed BROTHERS^TM technology, a non-equilibrium-inspired molecular design framework that mitigates the efficacy-toxicity coupling by integrating two orthogonal principles. 1) Static shielding suppresses non-specific interactions of single-stranded ASOs, while 2) dynamic selection enables kinetically driven target RNA recognition via a toehold-mediated strand displacement (TMSD) reaction. In vivo studies demonstrate that BROTHERS^TM-based ASOs (BROs) achieve potent and selective gene silencing while markedly reducing toxicities observed with conventional single-stranded ASOs. These results indicate that non-equilibrium control of molecular recognition can effectively alleviate the efficacy-safety dilemma, establishing BROTHERS^TM technology as a robust platform for next-generation antisense therapeutics.

Our group has been developing organic electrochemical reactions (since 1994) and liquid-phase peptide synthesis (since 2002) mostly independently of each other. The merger of electrochemistry and peptides had been a long-cherished goal of our group, yet apart from the limited example of electrochemical disulfide bond formation, the two had never encountered. Herein, we describe electrochemical peptide synthesis using triarylphophines as recyclable “coupling reagents,” where electrochemistry and peptides are merged.

In the field of synthetic organic chemistry, the standard approach to reducing reagent consumption is to develop new catalysts that promote the desired reaction. This is also true for amide bond formation and various catalysts have been reported so far, achieving remarkable outcomes. Another fascinating-yet-challenging approach is to make coupling reagents catalytic. In situ (one-pot) regenerable coupling reagents would be ideal, yet ex situ (two-pot) recyclable ones would also contribute to reducing the consumption. When it comes to recycling coupling reagents, it is crucial to avoid applying thermal energy and/or “second” coupling reagent; otherwise, it does not constitute a fundamental solution. In this context, we focused on oxidative (electrochemical) an amide bond formation using triphenylphosphine as a coupling reagent precursor. During amide bond formation, triphenylphosphine oxide is accumulated as waste, which can be recovered and reduced back (recycled) to triphenylphosphine. Tris(4-methoxyphenyl)phosphine was found to be a superior coupling reagent precursor to non-substituted triphenylphosphine in electrochemical amide bond formation, which was further enhanced by using iodide mediator. We have successfully synthesized three bioactive peptides, including leuprorelin (9-mer), bradykinin (9-mer), and icatibant (10-mer), without the use of typical coupling reagent.

Cyclic disulfide peptides are attracting significant attention as a new modality in drug development because their rigid structures enable selective binding to target molecules and confer high resistance to metabolic enzymes. Therefore, new methods for constructing disulfide bonds can contribute to more efficient preparation of these peptides. In this paper, we report the development of disulfide bond-forming methods using 3-nitro-2-pyridinesulfenyl (Npys) compounds and their application to the synthesis of disulfide peptides. The paper covers two topics. First, we developed a one-pot solid-phase disulfide ligation (SPDSL) method, which readily affords disulfide-linked products composed of two thiol-containing components. Based on this SPDSL strategy, we further demonstrate a disulfide-driven synthesis of cyclic peptides from two different peptide fragments. Second, we identified methyl 3-nitro-2-pyridinesulfenate (Npys-OMe) as a mild oxidative reagent that promotes intramolecular disulfide bond formation between two thiols within a peptide. As an application of Npys-OMe, we show that disulfide bond formation can be achieved on thiol-containing peptidyl resins. In addition, a water-soluble Npys derivative functions as a disulfide-forming reagent in aqueous buffer, enabling the direct oxidation of thiol-containing peptides prepared by native chemical ligation.

Mirror-image proteins (D-proteins) comprise D-amino acids and achiral glycine, which are capable of assembling into mirror-image architectures of native L-proteins. Although D-proteins cannot be obtained through recombinant technology, synthetic proteins with more than 100 residues have been prepared using advanced peptide chemistry techniques, including solid-phase peptide synthesis and native chemical ligation. To date, a number of mirror-image versions of target proteins have been synthesized and used to screen the therapeutic potential of mirror-image peptides (D-peptides), nucleic acids (L-nucleic acids), and natural products (mirror-image enantiomers). We have been exploring the looking-glass world to expand the repertoire of protein-based therapeutics through mirror-image screening. To this end, we have established processes for synthesizing nanobodies (VHH antibodies) and monobodies, which contain three variable regions on stable scaffolds to bind epitopes on target proteins. Despite their identical sequences, the mirror-image forms of nanobodies and monobodies are significantly less immunogenic in mice than their L-form counterparts. Scaffolds of mirror-image proteins have also been chemically modified to study their pharmacokinetic properties and potential for bioconjugation.

In recent years, manufacturing technologies for peptides and oligonucleotides have advanced considerably; however, research and development efforts have remained predominantly focused on upstream processes, particularly synthetic methodologies. In industrial practice, downstream operations, including purification and lyophilization, often constitute major bottlenecks to productivity, and meaningful, sustainable improvements at scale are unlikely without technological innovation in these areas. To address this challenge, we have focused our efforts on developing downstream processes beyond purification, with an emphasis on continuous purification and mixer-type lyophilization technologies. In this study, we describe the key features of these two technologies and evaluate their implementation in the large-scale manufacturing of a cyclic peptide, in comparison with conventional approaches. Application of these technologies resulted in an approximately 1.5-fold increase in overall yield and improvements in product quality. Additionally, the time required for downstream processing was reduced to nearly one-quarter of that associated with the traditional workflow. Collectively, these findings demonstrate that innovation in downstream operations can substantially enhance both productivity and product quality in peptide and oligonucleotide manufacturing.

X-ray crystallography remains a powerful technique for determining protein three-dimensional structures with high spatial resolution and high throughput, even though cryo-electron microscopy has become widely used. In conventional X-ray crystallography, most obtained structures were static. However, serial femtosecond crystallography enables the visualization of dynamic structures with high temporal and spatial resolution by combining reaction triggers such as light excitation and substrate mixing. In addition, we have recently established and begun operating an X-ray crystallography experimental station at the fourth-generation synchrotron radiation facility NanoTerasu. In this article, we introduce recent topics in X-ray protein crystallography, including not only experimental advances but also widely used protein structure prediction and molecular design tools.