RiPPs are ribosomally synthesized and post-translationally modified peptide natural products.  These natural products are produced in every domain of life, but are most prevalent in bacteria and other microorganisms, where they are usually biosynthesized as antimicrobial self-defence molecules.  Despite their vastly different chemical structures, all RiPP biosynthetic pathways share several characteristics that make these pathways attractive targets for molecular engineering approaches aimed at producing new to nature peptides with biological activity.  First, the entire pathway (both the substrate peptide and all the biosynthetic enzymes) is genetically encoded, which provides access to large libraries of peptide/enzyme variants that can be screened for novel functions/activities using cell-based engineering/directed evolution approaches.  Second, RiPP biosynthetic enzymes have relaxed substrate specificity – a useful property for engineering novel biological catalysts.  We are very interested in the mechanistic origins of the inherently relaxed substrate specificity of RiPP biosynthetic enzymes, which we suspect are strongly governed by transient and dynamic protein-peptide interactions in many systems.

As a model system, much of our research has focused on the class II lanthipeptide synthetase, HalM2 (publications 18, 23, 24, 28, 30, 32, 33, and 35 on the publications page).  This enzyme possesses two separate active sites that catalyze two distinct chemical transformations: 1) the dehydration of serine and threonine residues in the HalA2 precursor peptide and 2) the intramolecular conjugate addition of cysteine thiols onto the dehydrated sites to form thioether peptide macrocycles.  In lanthipeptides, these thioether rings are crucial for biological activity (most characterized lanthipeptides are antibiotics). A remarkable property of HalM2 is its ability to harness this relaxed substrate specificity, while maintaining biosynthetic fidelity – namely, the ability to make a single final product with a specific thioether topology even though many hundreds of isomers are theoretically possible.

To untangle the origins of this biosynthetic fidelity, we have pioneered the application of mass spectrometry based approaches to profile the complex reaction cascades and conformational dynamic properties of lanthipeptide synthetases.  To summarize briefly, we use mass spectrometry to extract kinetic parameters for the multistep reaction cascade catalyzed by HalM2.  Because each HalA2 peptide intermediate has a distinct mass, we can use MS to simultaneously follow the changes in concentration of many different species as HalM2 converts them into the final product.  We use numerical simulation to evaluate different kinetic models that describe the data.  We also use tandem mass spectrometry to fragment each reaction intermediate into pieces.  By measuring the masses of these peptide fragment ions, we can map the regiochemistry of the modifications and we can build detailed biosynthetic maps for the multistep maturation.  With this quantitative approach, we can systematically compare the catalytic properties of different lanthipeptide synthetases or the effects of enzyme mutations on catalytic efficiency and biosynthetic fidelity in order to better understand how the multifunctional enzyme is directing the sequence of reactions.

Figure 1. A) General pathway for post-translational maturation of RiPP natural products, highlighting the modular, genetically encoded peptide and the presence of modifications enzymes with relaxed substrate specificity.  B) Chemical mechanism of class II lanthipeptide synthetases, which install thioether macrocycles in the precursor lanthipeptide. C) The post-translational modifications catalyzed by HalM2.  D) Mass spectra kinetic data showing the conversion of the HalA2 starting material into the fully modified product containing 7 dehydrated Ser/Thr residues and 4 thioether rings.  E) Numerical simulation of the mass spectral kinetic data from (D).

A major hypothesis to emerge from our kinetic studies of lanthipeptide synthetases is that biophysical process such as enzyme conformational changes and enzyme-peptide binding interactions play a crucial role in regulating the rates of the chemical transformations measured in our kinetic studies.  Understanding these interactions may help to better manipulate the system of reactions.  However, defining these critical biophysical interactions is highly challenging in RiPP biosynthetic systems, because both the lanthipeptide synthetase and the precursor peptide are large and structurally dynamic.  These properties make traditional structural biology approaches such as nuclear magnetic resonance (NMR) and X-ray crystallography challenging.  To solve this pressing issue, we have again turned to mass spectrometry.  We have employed hydrogen deuterium exchange mass spectrometry (HDX-MS) as a tool to study the structural dynamics of lanthipeptide synthetases.  To the best of my knowledge, our study was the first detailed application of HDX-MS to investigate a natural product biosynthetic enzyme.  HDX-MS reports on the rates of solvent deuterium exchange into the protein backbone.  The exchange rate is highly sensitive to the extent of local hydrogen bonding.  Thus, the HDX-MS measurement directly probes local protein secondary structure and is highly sensitive to the structural perturbations induced by substrate binding.  Using HDX-MS, we have shown that highly dynamic regions of HalM2 are indeed critical for both substrate binding and enzymatic function.  We have also mapped out the (previously unknown) HalA2 peptide binding interface, have located allosteric networks that trigger high level enzymatic activity, and have shown that lanthipeptide synthetases with unique functional properties also have unique conformational dynamics.  We predict that this powerful, label-free biophysical tool will allow us to make important insights into RiPP biochemistry – far beyond what we have already discovered with the lanthipeptide synthetases.

Figure 2.  Hydrogen-deuterium exchange mass spectrometry provides insights into HalM2 dynamics. A) Regions of HalM2 that become more structured (blue) and more flexible (red) upon HalA2 peptide binding mapped onto a homology model of HalM2.  Most of these structural elements (shown in panel B) are unique to class II lanthipeptide synthetase enzymes. Mutations in these dynamic elements perturbed substrate binding and or dehydration/cyclization catalysis.

While HDX-MS provides valuable information on structural dynamics at local (peptide-level) spatial scales, it does not provide direct measurements of changes in tertiary structure that could be important for understanding how the enzyme orients the bound peptide for modification by the two HalM2 active sites.  Moreover, very little is currently known about the nature of lanthipeptide synthetase/peptide interactions.  To begin examining these properties, we employ native ion mobility mass spectrometry (IM-MS).  Using very gentle ionization conditions, we can transfer proteins from aqueous solution into the gas phase environment of the mass spectrometer in a manner that preserves the native fold of the protein (Figure 3).  We can then characterize the conformational landscape of the folded protein using ion mobility – a technique that separates ions on the basis of their shape (e.g. their conformation).  Using this approach, we have shown that HalM2 adopts two major native conformations.  In addition, we can keep the HalM2:HalA2 (enzyme:peptide) non-covalent complex intact.  Using ion mobility, we have shown that HalA2 binding triggers a conformational change in HalM2. Strikingly, we have also discovered that the conformational landscape of HalM2 is sensitive to the post-translational modification state of the HalA2 peptide.  Namely, as more modifications are installed into the peptide, the biophysical interactions with the enzyme change in a manner that influences the global and local structure of the enzyme.  This has provided some of the first evidence that the structure of RiPP biosynthetic enzymes can be influenced by the structure of the precursor peptides they modify.  More generally, this provides strong support to our long-standing hypothesis that enzyme-peptide interactions change during the course of peptide maturation.  Ultimately, measurements like this on other biosynthetic enzymes will be important for defining how molecular motions contribute to the sequence of events in natural product biosynthesis.

Figure 3.  A) Native mass spectrum of folded HalM2 and the HalM2:HalA2 complex with the charge state of each ion indicated.  B) Ion mobility characterization of the 22+ ions of HalM2 and the HalM2 complexes with the HalA2 leader peptide (see Fig. 1C) and full length HalA2.  Ion mobility separates ions on the basis of their conformation.  HalM2 exists in two conformations (compact and open) and peptide binding alters this conformational landscape.  C) The sequential installation of thioether rings into the HalA2 precursor alters the global conformation of the enzyme (as measured by in mobility) as well as the HDX dynamics in a helical region of the enzyme (peptide ³³⁵ARELTQSVF, colored orange) at the interface of the dehydratase/cyclase domains.  Our current model is that this helical region helps to guide the peptide to the cyclase active site.  As thioether rings are installed, these interactions weaken, the helical region becomes more unstructured, and the enzyme shifts to a more open conformation.