SummaryThe application of stem-cell-based therapies in regenerative medicine is hindered by the tumorigenic potential of residual human pluripotent stem cells. Previously, we identified a human pluripotent stem-cell-specific lectin probe, called rBC2LCN, by comprehensive glycome analysis using high-density lectin microarrays. Here we developed a recombinant lectin-toxin fusion protein of rBC2LCN with a catalytic domain of Pseudomonas aeruginosa exotoxin A, termed rBC2LCN-PE23, which could be expressed as a soluble form from the cytoplasm of Escherichia coli and purified to homogeneity by one-step affinity chromatography. rBC2LCN-PE23 bound to human pluripotent stem cells, followed by its internalization, allowing intracellular delivery of a cargo of cytotoxic protein. The addition of rBC2LCN-PE23 to the culture medium was sufficient to completely eliminate human pluripotent stem cells. Thus, rBC2LCN-PE23 has the potential to contribute to the safety of stem-cell-based therapies.

Human somatic stem cells such as human mesenchymal stem cells (hMSCs) are considered attractive cell sources for stem cell-based therapy. However, quality control issues have been raised concerning their safety and efficacy. Here we used lectin microarray technology to identify cell surface glycans as markers of the differentiation potential of stem cells. We found that α2-6Sia-specific lectins show stronger binding to early passage adipose-derived hMSCs (with differentiation ability) than late passage cells (without the ability to differentiate). Flow cytometry analysis using α2-6Sia-specific lectins supported the results obtained by lectin microarray. Similar results were obtained for bone marrow-derived hMSCs and cartilage tissue-derived chondrocytes. Little or no binding of α2-6Sia-specific lectins was observed for human dermal fibroblasts, which are unable to differentiate, suggesting that the binding of α2-6Sia-specific lectins is associated with the differentiation ability of cells, but not to their capacity to proliferate. Quantitative analysis of the linkage mode of Sia using anion-exchange chromatography showed that the percentage of α2-6Sia linkage type was higher in early passage adipose-derived hMSCs than late passage cells. Integrinα5 was found to be a carrier protein of α2-6Sia. Sialidase treatment significantly reduced the differentiation efficiency of bone marrow-derived hMSCs. Based on these findings, we propose that α2-6sialylation is a marker of differentiation potential in stem cells such as adipose-derived hMSCs, bone marrow-derived hMSCs, and cartilage tissue-derived chondrocytes.

Glycans are one of the major building blocks of extracellular vesicles (EVs). However, their roles and applications have not been completely explored. Here, we analyzed the glycome of EVs derived from human induced pluripotent stem cells (hiPSCs) using high-density lectin microarray. The glycan profiles of hiPSC-derived EVs were different from those of non-hiPSC-derived EVs. Moreover, rBC2LCN that shows specific binding to hiPSCs, showed strong specificity for hiPSC-derived EVs but not non-hiPSCs-derived EVs. Further, other hiPSC-specific probes, such as anti-TRA-1-60, anti-SSEA4, and anti-R-10G, exhibited specific, but weaker binding to hiPSC-derived EVs than rBC2LCN. We then developed a sandwich assay using rBC2LCN and a phosphatidylserine receptor, Tim4, to specifically detect hiPSC-derived EVs. The Tim4–rBC2LCN sandwich assay allowed for specific detection of hiPSC-derived EVs but not non-hiPSC-derived EVs, indicating that rBC2LCN could also be used for the specific detection of hiPSC-derived EVs. Together, our findings demonstrate that the characteristic glycan signature of hiPSCs are retained by EVs derived from them. The EV glycome could be novel targets for the identification and characterization of stem cells for use in regenerative medicine.

The use of human pluripotent stem cells (hPSCs) such as human embryonic stem cells (hESCs) and human induced pluripotent stem cells (hiPSCs) in regenerative medicine is hindered by their tumorigenic potential. Previously, we developed a recombinant lectin-toxin fusion protein of the hPSC-specific lectin rBC2LCN, which has a 23 kDa catalytic domain (domain III) of Pseudomonas aeruginosa exotoxin A (rBC2LCN-PE23). This fusion protein could selectively eliminate hPSCs following its addition to the cell culture medium. Here we conjugated rBC2LCN lectin with a 38 kDa domain of exotoxin A containing domains Ib and II in addition to domain III (PE38). The developed rBC2LCN-PE38 fusion protein could eliminate 50% of 201B7 hPSCs at a concentration of 0.003 μg/mL (24 h incubation), representing an approximately 556-fold higher activity than rBC2LCN-PE23. Little or no effect on human fibroblasts, human mesenchymal stem cells, and hiPSC-derived hepatocytes was observed at concentrations lower than 1 μg/mL. Finally, we demonstrate that rBC2LCN-PE38 selectively eliminates hiPSCs from a mixed culture of hiPSCs and hiPSC-derived hepatocytes. Since rBC2LCN-PE38 can be prepared from soluble fractions of E. coli culture at a yield of 9 mg/L, rBC2LCN-PE38 represents a practical reagent to remove human pluripotent stem cells residing in cultured cells destined for transplantation.

Lectins, carbohydrate-binding proteins, are attractive biomolecules for medical and biotechnological applications. Many lectins have multiple carbohydrate recognition domains (CRDs) and strongly bind to specific glycans through multivalent binding effect. In our previous study, protein nano-building blocks (PN-blocks) were developed to construct self-assembling supramolecular nanostructures by linking two oligomeric proteins. A PN-block, WA20-foldon, constructed by fusing a dimeric four-helix bundle de novo protein WA20 to a trimeric foldon domain of T4 phage fibritin, self-assembled into several types of polyhedral nanoarchitectures in multiples of 6-mer. Another PN-block, the extender PN-block (ePN-block), constructed by tandemly joining two copies of WA20, self-assembled into cyclized and extended chain-type nanostructures. This study developed novel functional protein nano-building blocks (lectin nano-blocks) by fusing WA20 to a dimeric lectin, Agrocybe cylindracea galectin (ACG). The lectin nano-blocks self-assembled into various oligomers in multiples of 2-mer (dimer, tetramer, hexamer, octamer, etc.). The mass fractions of each oligomer were changed by the length of the linkers between WA20 and ACG. The binding avidity of the lectin nano-block oligomers to glycans was significantly increased through multivalent effects compared with that of the original ACG dimer. Lectin nano-blocks with high avidity will be useful for various applications, such as specific cell labeling.