Since its founding, the Link lab has been interested in applying the tools of protein engineering and bioconjugate chemistry to engineer peptides and proteins with conformational constraints. Conformational constraints within a polypeptide can lead to improvements in properties such as protease stability, thermostability, and binding affinity. With the development of bioorthogonal chemistry in the early 2000’s, the toolbox for conformational constraints expanded. We have used a combination of protein engineering and azide-alkyne click chemistry to carry out “protein stapling” on small proteins (Abdeljabbar et al. Chem. Comm. 2014). We have also used olefin metathesis chemistry to constrain short engineered peptides that function to turn on apoptosis in cells (Link and Zhang, US patent 9,464,125). We have also looked to nature for inspiration in strategies for conformationally constraining peptides. This has led to our program on lasso peptides, an ever expanding class of natural products defined by their slipknot-like topology (see figure). Our group started out working on the antimicrobial lasso peptide microcin J25 as an interesting substrate for engineering by directed evolution (Pan et al. JACS 2011) and with unnatural amino acids (Piscotta et al. Chem. Comm 2015). We became interested in the variety of different lasso peptides present in nature and developed the first algorithm for large-scale genome mining of lasso peptides (Maksimov et al. PNAS 2012). We continue to look to lasso peptides as a source of new antibiotics, interesting new enzymology, and even as building blocks for molecular machines. Please read below for more details on these specific projects.

Many of the early examples of lasso peptides exhibit antimicrobial activity. Microcin J25 and capistruin, for example, kill some Gram-negative enterobacteria by inhibiting RNA polymerase. We are leveraging our genome mining algorithm and insights into structure-function relationships to discover novel antimicrobial lasso peptides. We are also interested in determining their mode of action.

Our interests in enzymology were sparked when we discovered a new enzyme, lasso peptide isopeptidase, that is widespread among alpha-proteobacteria encoding lasso peptides (see figure). This enzyme specifically recognizes the knotted shape of the lasso peptide and hydrolyzes it to form a linear peptide (Maksimov and Link JACS 2013). In addition, we have a long-standing interest in understanding the enzymes that make lasso peptides from their linear, ribosomally-synthesized precursor (Cheung et al. ACS Cent. Sci. 2016). Finally, with more and more genomes being sequenced, it’s becoming clear that lasso peptide gene clusters are colocalized with other interesting enzymes, including those that install posttranslational modifications. We are interested in studying these modifications and the impact that they have on lasso peptide function.

Mechanically interlocked molecules, including rotaxanes and catenanes, are the foundation of nanoscale molecular machines, a topic recognized with the Nobel Prize in chemistry in 2016. Lasso peptides are nature’s example of this type of molecule. Lasso peptides are [1]rotaxanes, and are comparable in size to synthetic rotaxanes. We have recently become interested in building up more complex architectures from lasso peptides, such as a [3]catenane and [4]catenanes from a variant of microcin J25 (see figure, Allen and Link JACS 2016). We have also studied the unthreading (Allen et al. ACS Chem. Biol 2016) and switching behavior (Zong et al. JACS 2017) of lasso peptides as a way to incorporate molecular motions into lasso peptide-derived architectures.