Ming Xian

Late-stage peptide modification is a powerful strategy for diversifying and expanding peptide bioactivity. Although numerous methods have been developed for peptide side-chain functionalization, chemical approaches for selective backbone editing remain scarce. Inspired by the natural enzyme-mediated conversion of cysteine residues into thiazoles during peptide biosynthesis, we developed a traceless “label-and-edit” strategy for the backbone modification of cysteine-containing peptides. This approach begins with the selective labeling of cysteine thiols using 2-bromoacetophenone, followed by photoinduced generation of a thioaldehyde/enethiol intermediate and subsequent intramolecular SH/amide cyclization to afford thiazole-containing peptides. The transformation is compatible with a broad range of synthetic and natural peptides, highlighting its potential as a versatile platform for late-stage peptide backbone editing and diversification.

Precisely controlling shape and crystallinity is central to defining structure-dependent properties in colloidal nanocrystals (NCs), which is particularly important yet challenging for silver (Ag) NCs. Here, we report catalytic regulation that facilitates cluster-mediated growth pathways to enable the scalable synthesis of single-crystalline Ag-NCs with programmable shapes. A copper (Cu)-assisted, one-pot heating-up process produces highly uniform Ag-NCs with tunable diameters (5–11 nm) and morphologies ranging from truncated-octahedral (“mecon”) to spherical geometries, while maintaining gram-scale batch output, air tolerance, and compatibility with continuous flow synthesis. Spectroscopic and mass spectrometric analyses reveal a stepwise growth pathway involving discrete Ag cluster intermediates, whose formation and interconversion are accelerated by trace Cu species acting as catalysts. Density functional theory calculations further show that Cu modulates surface and stacking-fault energetics, promoting single-crystalline growth with well-defined facets. By introducing a quantitative shape descriptor, we establish predictive relationships linking NC morphology to plasmonic response, supported by discrete dipole approximation simulations. These results define a general synthetic principle in which catalytic species and dynamic ligand environments cooperatively govern cluster-phase evolution, enabling precise and scalable control of NC structure and function.

Protein S-persulfidation is a key post-translational modification in redox signaling. In this process, hydropersulfides (RSSH) and hydrogen persulfide (H₂S₂) act as regulatory sulfane sulfur species, transferring their electrophilic sulfane sulfur (S⁰) atoms to protein thiols (P-SH) to form protein persulfides (P-SSH). Due to the high reactivity and instability of RSSH/H₂S₂, controllable donors are essential chemical tools for inducing protein S-persulfidation. Existing donors are primarily disulfide-based; however, this structural feature leads to a major limitation, as disulfide-based donors are known to undergo unavoidable disulfide exchange reactions with cellular thiols (e.g., glutathione (GSH), cysteine (Cys)) and subsequently lose their ability to release the desired RSSH/H₂S₂. In this work, we proposed that thiosulfoxides [R-S(═S)-R'] could serve as equivalents of RSSH/H₂S₂ for protein S-persulfidation and that controlled thiosulfoxide formation is an alternative way for the design of RSSH/H₂S₂ donors. To prove the hypothesis, we studied spontaneous [2,3]-sigmatropic allyl disulfide rearrangements with a variety of allyl disulfide substrates. We identified structural elements that promote the rearrangement to form thiosulfoxide adducts and demonstrated that the resulting transient thiosulfoxide intermediates were effective S⁰ transfer agents. Based on this discovery, we developed stable allyl disulfide reagents (such as 3i) that could be selectively activated by cellular thiols via the inevitable disulfide exchange, followed by a [2,3]-sigmatropic rearrangement cascade to induce protein S-persulfidation. Our results present a conceptually new way to deliver reactive sulfane sulfur species in biological systems.

Hydrogen sulfide (H₂S) is an important gasotransmitter that has shown many physiological effects, ranging from anti-inflammation to antioxidation. To advance research on H₂S, donor compounds that can slowly release H₂S in biological conditions while producing minimal bioactive byproducts are essential. Herein, we report the evaluation of thiosaccharides as hydrolysis-based H₂S donors. These compounds were found to slowly produce H₂S over days, in aqueous buffers and in cells. Their H₂S release rates could be affected by the hydroxyl protection groups of thiosaccharides, with faster release by electron-donating groups and slower release by electron-withdrawing groups. We also demonstrated the vasodilatory effect of 1-thioglucose using arterial rings isolated from adult ewes, which is likely due to H₂S release. Altogether, thiosaccharides might be suitable slow-releasing H₂S donors for biological applications.

Enveloped viruses are significant zoonotic disease threats with the potential to cause global pandemics. We identified a class of small-molecule sulfur-containing antiviral compounds (XM series) that broadly inhibit enveloped viruses. Multidisciplinary approaches revealed that XM compounds alter the viral membrane lipid chemical composition, enhance membrane order within the hydrophobic bilayer, and increase membrane phase transition temperatures. This mechanism inhibits membrane fusion and viral entry, while leaving the viral glycoproteins and genomes largely unaffected. Leveraging these unique properties, we develop a proof-of-concept whole inactivated influenza virus (IIV) vaccine using XM-01 (XM-01-IIV). In a mouse model, XM-01-IIV elicit significantly enhanced neutralizing antibody responses against hemagglutinin and neuraminidase compared to traditional paraformaldehyde-inactivated vaccines. Further, XM-01-IIV reduces morbidity and mortality following influenza challenge, achieving protection comparable to live virus vaccination. This promising class of broadly acting antivirals can be highly impactful in the development of highly potent inactivated vaccines for enveloped viruses.

Late-stage peptide modification is an important strategy for exploiting and expanding peptide bioactivity. While numerous methods exist for side-chain functionalization, chemical approaches for backbone editing remain limited. Inspired by the natural, enzyme-mediated conversion of cysteine to thiazole within peptides, we developed a traceless “label-and-edit” strategy for the backbone modification of cysteine-containing peptides. This method involves selective, biocompatible labeling of cysteine thiols with bromoacetophenone, followed by photoinduced thioaldehyde/enethiol formation and subsequent SH/amide cyclization to generate thiazole-containing peptides. A broad range of synthetic and natural peptides are compati-ble with this transformation, demonstrating its value as a versatile platform for late-stage peptide modification.

Phthalimide-based disulfides are used to readily prepare trisulfides on both small molecule and protein substrates.