70 | December 2021 www.facs.website shuttled towards the kidneys and urinary bladder from liver. To further increase glycan complexity, the biodistribution in mice based on the usage of heterogeneous glycoproteins 1e and 1h, containing additional branching, was also studied25. Although glycoprotein 1e followed a pathway similar to 1d, it showed higher accumulation in intestines compared to 1d. The glycoprotein 1h accumulation compared glycoproteins 1f-g was rather directed to the kidneys and then excreted via urinary bladder. Moving away from biodistribution, we next focusedon the investigationof the tumor targeting capacity of artificial glycoproteins. According to the in vivo imaging data (Fig. 3D)28, homogeneous a(2,3)-sialic acid terminated glycoprotein 1f was the most promising, as it accumulated at a significantly higher level in theA431 tumor than the other tested glycoproteins 1a, c, d, and g. With the aimof improving upon this modest target, the focus of a later studywas onartificial glycoproteins composed of heterogeneous glycan assemblies. To test differential tumor selectivity, several heterogeneous glycoproteins were injected into mice implanted with 3 different tumors (HeLa, DLD-1, and U87MG).29 In this study, the most significant observation was that the a(2,3)-sialo and a(2,6)- sialo terminatedglycoprotein 2a showed selective accumulation toHeLa tumors, while it showed no targeting to theDLD-1andU87MG tumors. Apossible explanation for this result is that HeLa cells are known toexpress bothgalectin-1and siglec-3 lectins, which are receptors for a(2,3)-sialic acid and a(2,6)-sialic acid, respectively. In contrast, DLD-1 andU87MGonly express galectin-1 lectin. Given this promising result, to multiply technique to more selectively target tumors, we developed another approach using higher-order heterogeneous glycoalbumins that were conjugated with four kinds of complex N-glycans.33 As depicted in Figure 3D-(III), the in vivo data revealed that the higher-order glycoalbumin 3b displayed the strongest targeting towards SW620 tumors than lower-order glycoalbumin 3a. Overall, the useof glycanpattern recognition for organsor cancer cells targeting represents anovel and promising strategy for the development of diagnostic, prophylactic, and therapeutic agents for various diseases. Moreover, the use of glycan targeting would have significant advantages over current techniques (i.e. antibody targeting), such as shorter accumulation times and lower immunogenicity. Biocatalysis and Mild Biocompatible Reactions Translating abiotic metal catalysts into in vivo synthetic chemistry could encounter numerous challenges regarding their biocompatibility, stability, and reactivity in the complicated biological environment. To solve these issues, we have been actively involved in research related to artificial metalloenzymes (ArMs). ArMs are created by incorporating synthetic metal complexes into protein scaffolds, thereby combining the advantageous features of organometallic and enzymatic catalysts and facilitating the design of novel biocatalysts to perform new-to-nature reactions. To combine with glycan targeting, we were mainly interested in developing ArMs from human serum albumin. Owing to previous direction from the late Prof. Koiji Nakanishi, we decided to utilize coumarin derivatives as an anchor for the hydrophobic binding pocket of albumin (Fig. 4A). With the use of a varying series of coumarin-metal complexes, we developed four kinds of ArMs (mainly ArM-Au and ArM-Ru) (Fig. 4B)32,36-41. An important observation from these studies was the discovery that a combination between the deep hydrophobic binding site of albumin and the negatively charged surface of albumin naturally repels entry to hydrophilic metabolites (i.e., glutathione (GSH)). As a result, using a 1,6-heptadiene-based substrate, metastasis activity was shown to proceed even in the presence of up to 1000 x equivalents of GSH additive.36 As depicted in Figure 4B-I, we also have developed several biocompatible organic reactions that are applicable to these ArMs. In a preliminary screen to identify amide bond formation, we were surprised to see that the Au(III) complexes coordinated with 2-benzoylpyridine could generate amide via propargyl esters.32,39,41 Forming an activated ester intermediate (via Au binding) likely leads to amide bond formation via a nucleophile amine. Using this chemistry, we have shown that fluorescent labeling of proteins is possible using propargyl ester-based probes and the ArM-Au-1. On the other hand, the ArM-Au-2 containing Au(I) complexes coordinated with an N-heterocyclic carbene ligand can perform hydroamination to synthesize phenanthridinium derivatives with an excellent turnover number (Fig. 4B-II).38 In particular, the phenanthridiniummoiety has attracted a great deal of attention because of its presence in the scaffolds of several DNA-intercalating agents with antitumor properties. Importantly, hydroamination is not catalyzed by any known naturally occurring enzymes, highlighting the significance of the ArM-Au-2 catalyzed hydroamination under physiological conditions. In addition to that, ring-closing metathesis (RCM) for olefins and ene-ynes can be catalyzed smoothly by the ArM-Ru-1, which incorporates the 2nd generation Hoveyda catalyst.36,37,41 RCM iswidely recognized as a powerful method for creating heterocycles and phenyl moieties that are most significant structural componentsof pharmaceuticals. Lastly, ArM-Ru-2 can effectively catalyze alkylation with nucleophilic moieties such as thiol, hydroxyl, and amino groups in biomolecules using a benzyl fluoride substrate via a quinone imine intermediate.40 Overall, theArM-Au-2 and theArM-Ru-1 couldbe utilized as powerful biocatalysts for application in therapeutic in vivo drug synthesis. As for in vivo imaging and drug conjugation, the ArM-Au-1 and the ArM-Ru-2 are competent for these tasks, respectively. With these new developed biocatalysts and biocompatible reactions, we would like to adapt and apply these technologies for innovative applications. In one of our endeavors, we looked specifically at development of ArM-based biosensors, which would offer a unique path for tailoring against difficult-to-detect metabolites. Ethylene gas is an essential plant hormone that plays amajor role in regulating aspects of growth, immunity, and senescence. With this in mind, our group has investigated the creation of an ethylene-sensing ArM biosensor.37 As depicted in Figure 4D, the basis of this approach is to use the albumin scaffold to solubilize and protect a quenched ruthenium catalyst complex. In the presence of ethylene, cross metathesis is then occurred, leading to the removal of the quencher and the emission of a fluorescent signal. Using the ArM ethylene probe (AEP), the AEP was used to detect changes in ethylene biosynthesis specifically in the outer pericarp of kiwifruit. Since this process is typically unregulated during the ripening process, comparative studies showed an increase in pericarp fluorescence for ripening kiwifruits (Fig. 4E). Since chemotherapy is not perfectly specific for cancer cells, it has significant side effects on healthy cells. Therefore, another practical application is to employ the ArM-Au-2 as a trigger to control the release of bioactive drugs to improve the defect of chemotherapy.38 As shown in Figure 4F, the ArM-Au-2 successfully implemented drug synthesis from a non-active prodrug to achieve cancer therapy in a cell-based assay, suggesting the potential of the gold ArM to be a therapeutic ArM for in vivo anticancer application. Therapeutic in vivo Synthetic Chemistry by GArMs Given our interest in both glycan targeting and biocatalysis, a natural course of action eventually led to combining both aspects of targeting glycoproteins and ArMs to establish the concept of glycosylated artificial metalloenzymes (GArMs). The ultimate goal of this endeavor will be to eventually establish effective and biocompatible therapeutic ArMs, which can then be conferred with organ/tumor targeting properties by simply decorating the protein scaffold with an appropriate glycan assembly. Our first attempt at developing GArMs came during a study to determine whether specific tissues of mice could be targeted for in vivo labeling (Fig. 5A-I).32 In this work, glycosylated ArM-Au-1 with the intent to label targeted cells in vivo with propargyl ester-based probes. As depicted in Figure 5A-II, mice were first intravenously injected through the tail vein with a GArM. Then, a near infrared fluorescent propargyl ester (Cy7.5-PE) was injected. As shown in the imaging results, preferential organ labeling could be achieved depending on the identity of the attached glycans; α(2,6)-sialic acid terminated glycans targeted the liver, while Gal terminated glycans targeted the intestines. In the controls, localization of fluorescence labeling was not exhibited in targeted organs. Given the promising results, more recently, we represented research on selective cell tagging (SeCT) therapy in vivo via the GArM-Au-1 (Fig.
RkJQdWJsaXNoZXIy NDU2MA==