SIMB 2024

Highlights#

SIMB typically runs in multiple tracks and its not possible to get to everything. Here are my highlights from the meeting. All the speakers from the session that Kou-San and I organized are there, of course.

Natural Products Discovery and MOA#

Natural products represent one of the richest historical sources of drug pharmacophores. They are interesting compounds that come from interesting places including (ahem,ahem) au uncultured microbes. A lot of recent interest has been related to genome-mining - the use of genomes to identify and prioritize gene collectives whose biochemistry creates the compounds of interest. This year, however, the talks that captivated me were a bit different:

There were a few talks on gene cluster discovery and optimizations for discovery and scaling of natural products (Henry Le, Joanne Wong); there were tools for using omics-tools to identify the mechanism of activity of compounds of interest (Brian Bachmann, and Tim Bugni); the application of different tools to understand and thereby engineer components within BGCs (Allison Walker); or the use of new strategies or approaches to identify promising BGCs (Yousong Ding, Kou-San Ju, Emily Mevres.)

My personal favorite talk was from Emily Mevers who isolated a whole slate of compounds from different millipede species. She showed a video of the millipedes being “milked”: the subsequent HPLC of the “milk” contained 4 clean peaks, 3 of which were previously uncharacterized alkaloids. Whoah. Furthermore they had specific activity against a single orphaned neuroreceptor. This underscored the exciting possibility of biodiverse pockets with undiscovered chemistry. Arthropods: the final frontier.

Biofuels#

I finally got a chance to hear the godfather of biofuels, Gregory Stephanopolous, give an opinionated talk on the current state and near term future of biofuels. He started with a thesis based on high level economic and technical trends and then quickly got into the details. He thinks that bioethanol is basically on the way out as electrification will slowly take over most terrestrial electrical systems and the future of biofuels are in the industries that are recalcitrant to direct electrification - in particular fuels for planes and ships. He then laid out an argument for why hydrogen (+C1) is the best, most biologically cost effective electron source for fuel-generating organisms. One of his graphs was of a extrapolated $/per electron. He got a chuckle from the audience when showing that and then saying I don't know what this means to you but to me this means nothing but then explaining how it supported his favored solution which is a paired system where anaerobic organisms use hydrogen and CO2 to create methane/methanol, and that is then used by oxidative organisms to generate long-chain lipids - “microbial fat”.

Unsurprisingly, a Stephanpolous alumni, Ben Woolston gave a great talk out of his lab where he is trying to take the basic insight that Greg laid out above but use co-cultures of aerobic and anaerobic bacteria to get a “one-pot” fat producing solution. It looks very promising, and clearly a smaller process is preferable to a large one.

In a similar vein, Doris Hafenbradl of Electrochaea, gave a talk of her company’s creation of several pilot-scale facilities for the the conversion of H2+CO2 to methane. They have an Archael organism that can natively produce methane at 80% efficiency (!) and they have designed and evaluated plants that can be built to generate methane. The current cost pressure is actually due to the cost of energy to run the plant - in the form of electrolysis of water and in the compression of gas upstream of the reactor. Its wonderful to see more steel in the ground and I wish them great success.

Biocatalysis#

The slow march of Francis Arnold alumni through the biocatalysis institutions continues apace: Yang Yang gave a beautiful talk on radical chemistry bio-engineering that is creating enantiomerically pure non-canonical amino acids with three stereocenters; Xiongyi Huang is designing non-heme metaloenzymes to encroach on the territory of organic chemists. Russel Lewis of Pfizer is optimizing ERED/KRED enzymes for multikilogram enzymatic manufacturing of cancer drug tagtociclib; John McIntosh of Merck gave a masterful recap of efforts to find and optimize enzymes synthesize a therapeutic STING activator.

I don’t think you can go away from these sets of talks and not be completely optimistic on the future of protein engineering.

Tech/hardware#

Quick bullets on tech that I saw and liked:

Zachary Sun of Tierra Biosciences talked about how he has routinized cell-free systems for rapid and reproducible protein production. There is a 2 week turn around time from Amino acid sequence in their portal to lyopholized protein in a tube in your lab. Honestly, I can’t believe I didn’t know about this company already - this really lowers the barrier for protein design companies to test their designs.

Alex Rosay (ex-Zymergen) represented Cascade Biocatalysts. They make polymer beads that immobilize and stabilize enzymes - increasing their useful shelf-life, thereby reducing cost. I’ll keep an eye open for them.

In the showroom, two fermentation device companies caught my eye: Sunflower and Caladan. These companies demonstrate the miniaturization and commodification of awesome, information dense hardware for fermentation. Sunflower has a continual flow fermenter that they claim can get fantastic yields; Caladan sells a desktop instrument with a slick and powerful UI.

Talk Abstracts#

Cell Free Protein Synthesis#

Towards predictable protein synthesis using high-throughput
cell-free synthetic biology workflows and AI/ML techniques
Z. Sun, PhD*, U. Yakubu, PhD, P. Bisesi, BSc, B. Cruz, BSc
and O. Kilian, PhD, Tierra Biosciences, San Leandro, CA, USA

To provide access to the key molecules supporting the central
dogmaof biology, processes have been built out to automate,
democratize andoutsource DNA and RNA. However, access to protein
synthesis still remainslargely limited and custom, requiring
scientific expertise developed through empirical iterations.

Automated cell-free protein synthesis workflows allow for the
centralized production of hundreds of proteins and protein
variants using many synthesis conditions in parallel. In
additionto being able to run many different proteins
belonging to variousclasses at one time, our high-throughput
platform is also able toscreen synthesis conditions in a
combinatorial manner. This allowsfor the empirical testing
of conditions to identify how to producea protein without
having to narrow down the conditions to accommodate manual
protocols. We show the platform can screen various complex
parameters, such as solubility tags, small molecule cofactors,
chaperones, and protein subunits. Additionally, the cell-free
platform allows the testing of many proteins for a newly
developed E. coli lysate that supports the formation of
disulfide bonds. Generating these synthesis conditions on our
platform relies on biochemistry rather than biology, which is
harder to standardize and predict.

We also illustrate that streamlined high-throughput cell-free
protein synthesis can be used to validate LLM-oriented protein
design workflows (lysozymes and other enzymes). We also show
how proteins can scale from smaller, discovery/screening scales,
to gram-scale production with ease. Finally, since the results
are collected at scale from the platform, we cover how AI/ML
learning can be incorporated with protein datasets to improve
protein synthesis predictability.

Cell-Free and Orthogonal Biocatalysis#

• Quintin Dudley

Protein Immobilization for Enzyme Longevity#

• Alex Rosay of Cascade Biocatalysts
• 8 person team.
• beads for inding and stabilizing protiens that increase protine longevity and reusability in industrial processes.

Developing an enzyme immobilization platform for more cost-
effective cell-free biomanufacturing

A. Rosay*, L. Macdougall, Ph.D., A. Chaparro, Ph.D. and
J. Weltz, Ph.D., Cascade Biocatalysts Inc., Denver, CO, USA

Enzymes are nature’s catalysts, enabling selective reactions
from renewable feedstocks under mild conditions and, therefore,
are critical for the decarbonization of chemical manufacturing.
While a few enzymes have already been scaled into industrial
processes such as glucose isomerases or lipases, most enzymes
are often too delicate, expensive, and short lived in cell-free
environments for economical use in large scale chemical
manufacturing processes. Cascade Biocatalysts empowers enzymes
for sustainable chemical manufacturing with our patent-pending
Body Armor for EnzymesTM platform. Our technology immobilizes
enzymes onto surfaces decorated with synthetic polymers to create
rugged, reliable, long-lasting biocatalysts for sustainable,
cost-effective chemical manufacturing.

Enzymes are self-assembling polymers consisting of different
sequences of twenty amino acid monomers. Each enzyme structure,
or fold, is unique and presents a distinct solvent accessible
surface,like a chemical fingerprint. Additionally, recent
computational toolssuch as AlphaFold provide enzyme structure
information when experimental data is unavailable. Synthetic
polymers can stabilize enzymes through covalent and non-covalent
interactions with this solvent accessible surface. Leveraging
unsupervised machine learning, we predict polymer compositions
which stabilize the enzyme to make ultrastable biocatalysts for
various industrial applications. Recent developments in
controlled radical polymerization facilitate copolymerization
of numerous readily available acrylic monomers, enabling enzyme
stabilization through favorable hydrophobic, charge, and covalent
interactions with the enzyme surface. Enzyme stability is not
only highly desirable in industry but provides an easily measured
output of our design build test learn cycle. Both copolymer synthesis
and enzyme stability measurements are amenable to existing high
throughput workflows, allowing rapid iteration of polymer composition
to further improve biocatalyst performance and inform our
computational models. Ultimately, this provides an alternative to
lengthy and expensive enzyme engineering campaigns when stability
issues arise in biocatalyst process development.

Through this enzyme immobilization platform, Cascade has stabilized
a range of different enzymes from various classes, improving enzyme
stability under higher temperatures and solvents relevant for
industrial chemical processes. By increasing enzyme lifespan and
total turnover, Cascade's technology can drop the cost of biocatalysis
and cell-free biomanufacturing catalyzing a green chemical industry.

Millipedes#

• Emily Mevers

Neuromodulating alkaloids from millipede defensive secretions

Natural products, commonly deployed as predator defense, have
formed the basis for today’s therapeutic agents. Millipedes have
evolved specialized glands to store structurally diverse
chemicaldefenses, including cyanide containing metabolites,
alkaloids, and oxidized aromatics. Studies into the secretions
of understudied millipedes, including Ischnocybe plicata,
Andrognathus corticarius, and several species of Brachycybe, has
led to the characterization of over 20 new oxidized monoterpene
alkaloids. These molecules are structurally distinct from all
known millipede defensive agents and represent truly unprecedented
carbon backbones. I. plicata produces oxidized tricyclic
(5,6,6-fused) monoterpenes, the Brachycybe species all produce
bicyclic indolizidine and quinolizidine monoterpenes, while
A. corticarius produces oxidized tetracyclic (6,6,6,5-fused)
monoterpenes. The latter contains eight continuous stereogenic
centers and represents some of the most chemically complex
arthropod-derived alkaloids discovered to date. We have shown that
these compounds are the only components within the defensive
secretions and the purified compounds impede motor function of
common predatorsat ecologically relevant concentrations. Based
on this observation, a subset of these new alkaloids were screened
widely for neuromodulating activity, which revealed they have
potent (Ki 13.6 nM) and selective binding affinity for sigma-1,
an orphan neuroreceptor. This research not only sheds light on the
remarkable chemical diversity of millipede natural products but
also paves the way for a deeper understanding of the ecological
and biosynthetic significance of this family of metabolites.

C1 Chemical Production#

Harnessing acetogenic bacteria for sustainable chemical production from C1 substrates

Harnessing acetogenic bacteria for sustainable
chemical production from C1 substrates
B. Woolston*, Northeastern University, Boston, MA, USA

Single-carbon (C1) compounds including carbon monoxide,
methanol and formic acid have emerged as promising feedstocks
for biofuel and biochemical production. These substrates can
be produced renewably from CO2 through electrocatalysis or
hydrogenation with renewable hydrogen, thus bypassing food
security and land conversion concerns raised over traditional
biofuel feedstocks. Acetogenic microbes are a particularly
attractive class of organisms for the biological upgrading
of C1 substrates, and use the Wood-Ljungdahl pathway (WLP)
to form carbon-carbon bonds between C1 compounds at electron
efficiencies greater than 80%. However, these bacteria present
a number of challenges for metabolic engineering efforts,
in particular an undeveloped synthetic biology toolbox,
an incomplete understanding of their central metabolism,
and the energetic limitations imposed by their anaerobic
lifestyle. In this talk, I will present my group’s work over
the last few years to overcome these challenges. First, I will
describe a dramatically expanded set of genetic tools for the
acetogen Eubacterium limosum, including CRISPR-recombineering
and CRAGE systems that enable rapid generation of genome-wide
precise mutations and integration of large biosynthetic pathways,
respectively. Next, I will show how the use of 13C-metabolic
flux analysis applied to methanol-grown E. limosum has uncovered
an unexpected reaction and helped us understand bottlenecks during
acetogenic growth. Finally, I will describe a novel co-culture
approach we are using to extend the product range achievable by
gas fermentation. In this system, an acetogen and an aerobic
heterotroph are cultured together under microaerobic conditions,
establishing a syntrophic relationship where the aerobe protects
the anaerobe from oxygen in exchange for fixed carbon which it upgrades
to value-added products. Ultimately, the new tools and learnings from
this work will accelerate the deployment of acetogenic bacteria as
part of the sustainable, circular bioeconomy.

Methanoarchaea for chemicals#

Harnessing methanoarchaea for sustainable fuels and chemicals
N. Buan*, University of Nebraska Lincoln, Lincoln, NE, USA

Methane-producing archaea, known as methanoarchaea, are strictly
anaerobic microbes that grow by making methane gas. To do so, they
seemingly “bend the rules” of thermodynamics by using a reversible
energy conservation pathway to synthesize methane from inexpensive
and abundant substrates such as acetate or C1 compounds
(CO, H2+CO2, formate, methanol, etc.). The Buan lab uses
interdisciplinary approaches to study and manipulate energy metabolism
in methanoarchaea to control whether metabolic flux is directed
towards biomass synthesis and/or methane production.

Recently the Buan lab has engineered methanoarchaea to produce
the high-value chemical isoprene as a byproduct in addition to methane.
As methane is both a potent greenhouse gas as well as a renewable fuel,
understanding methanoarcheal biochemistry and physiology is important
for predicting or controlling methane emissions from natural sources,
and for potentially using these organisms supply society’s energy
and transportation fuel needs.

Methanogens go Commerical#

• Electrochaea

D. Hafenbradl*, Electrochaea GmbH
Planegg, Bayern, Germany
Electrochaea’s biological methanation technology produces
methane from CO2 and H2 using a patented methanogenic archaea
strain. The CO2 can come from any source, either a purified
CO2 gas stream or from a feedstock that already contains
other components like CH4 from biogas, or ethanol from a
bioethanol fermentation. As a first step electrical energy
is converted into hydrogen by electrolysis. This hydrogen is
then combined with CO2 and converted via biomethanation into
renewable methane (CH4). The product is a low carbon intensity
alternative to fossil gas. A specific highlight of the
biomethanation technology is that intermittent sources of
renewable energy can be utilized without any negative impact
on the biocatalyst. After demonstrating the technology in industrial
scale pilot plants, Electrochaea is now ready for commercial application.
The market for the low carbon intensity methane is increasing worldwide.
Electrochaea has scaled the technology to commercial levels
(10MWe and 75MWe) with the co-funding of European Commission ́s
Accelerator program and with support of their strategic partners.
The first European commercial project is located in Denmark and will
integrate the biomethanation technology into the Rybjerg Biogas operation
– leading to the first carbon-neutral farm in Europe. In Northamerica,
the province of Quebec has become the frontrunner in enabling renewable
methane production facilities. The presentation will showcase the two
leading commercial projects in Denmark and Canada.

C1 –> Liquid Fuel#

Cost effective production of liquid fuels from renewable feedstocks

Cost effective production of liquid fuels from renewable feedstocks
G. Stephanopoulos*, Massachusetts Institute of Technology, Cambridge, MA, USA

The importance of liquid fuels in transportation is well established, yet,
there are presently no viable options for their cost- effective production
at scale from renewable feedstocks. During the past 15 years we have been
developing in my lab a system for the conversion of gas mixtures of hydrogen
(or CO) and CO2 to oils or alkanes. The two-stage system comprises anaerobic
fixation of CO2 and conversion of the CO2 fixation product (for example, acetate)
to lipids, from which biodiesel, green diesel and SAF can be produced. In
another application, the CO2 fixation product is converted to alkanes.
Our work includes both the engineering of the microbes and development of
a process to achieve gas to liquid conversion in prototype systems. These
systems are scalable, make no use of land (beyond what is needed for
generating renewable electricity for hydrogen production), do not compete
with food and are cost competitive based on high level cost analysis. I will
present the essential features of this process in my talk; full details can
be found in the 5 papers cited.

Protein Ligand Co-Design#

Challenges in Protein-Ligand Co-Design
S. Hassoun*, Tufts University, Medford, MA, USA
There are emerging applications such as bioelectronic sensing
and synthetic biosynthesis where it is desirable to engineer both the
protein and ligand to maximize system performance. A co-design
paradigm, of both ligand and protein, would therefore allow
for simultaneous changes of both system components and the
consideration of how the change of one component affects the function
of the other or the interaction between them. For example, binding
affinity is dependent on both enzyme and ligand, and in the space of
possible co-modification, more favorable pairings may be identified
by matching the binding pocket volume to the size of the molecule and vice versa.

This talk will first address co-design and co-optimization challenges
and offer a conceptual co-design framework that can be utilized to
implement co-design strategies. This framework offers modularity,
effective design space exploration and flexibility in utilizing various
tools to achieve design goals. The talk will then present the application
of this co-design framework to improve the binding affinity between
Type II NADH:quinone oxidoreductase (Ndh2) and quinones. We will
highlight that in modifying both Ndh2 and quinones, the co-design framework
considers a larger number of protein-ligand pairings when compared to
utilizing individual design paradigms where either ligand or protein are
modified. Importantly, we show that the co-design paradigm leads to improved
binding affinities. The talk will conclude with a description of future deep-
learning approaches that can expedite protein-ligand co-optimization.

Biocat for CO2 Capture#

Multi-Pronged Approach to building better Biocatalysts

AI & Computational Biology: a multi-pronged approach to building better biocatalysts
J. Wang*, Lanzatech, Skokie, IL, USA

LanzaTech’s mission is to challenge and change the way the world
uses carbon, enabling a new and sustainable system where carbon
is captured at the source, and further transformed into valuable
and useful commodities at scale. This process is fundamentally powered
by biology, where our microbial biocatalysts convert waste feedstock
into chemicals. The faster we can engineer biocatalysts that can convert
different types of feedstock into different chemicals, the faster we can
achieve direct production of commodity chemicals on a distributed scale
while reducing carbon emissions globally.
To date, LanzaTech has deployed 6 commercial plants, avoided more
than 400,000 tonnes of CO2, and produced more than 12 million US gallons
of ethanol. The optimization of our biocatalysts using computational
and systems biology approaches has played a pivotal role in our journey.
This presentation highlights how LanzaTech has leveraged and integrated
AI/machine learning strategies, metabolic modeling, systems biology, and
data capabilities to accelerate our strain engineering and fermentation
efforts, building out an extensive chemicals pipeline, and promoting
a circular bioeconomy.

AI for Engineering Biosynthesis#

Artificial intelligence for engineering biosynthesis and understanding the chemical ecology of natural products
A. Walker*, Vanderbilt University, Nashville, TN, USA

Advances in DNA sequencing have led to a wealth of newly sequenced
biosynthetic gene clusters (BGCs) from the genomes of isolated microbes
and environmental and human microbiome metagenomes. My lab seeks to
develop artificial intelligence (AI) and other computational methods
to take advantage of these data to learn more about the biosynthesis and
chemical ecology of natural products. There has been a longstanding
interest in engineering biosynthetic pathways to produce novel natural
product- like compound, however these efforts are limited by a lack of
understanding of the functional constraints of biosynthetic pathways.
Analysis of existing BGCs enables the identification of the evolutionary
constraints on BGCs, which can then be applied to guide engineering efforts.
I will discuss my lab’s efforts to discover some of these evolutionary
constraints and to develop AI tools for BGC engineering. My lab is also
interested in the chemical ecology of natural products. Natural products are
known to mediate competition and cooperation between microbes, but the
ecological role of most individual natural products is poorly understood. We
are developing tools for quantifying the relative abundance of different BGCs
in metagenomes. These tools can then be used to develop hypotheses about
the role the product of those BCS play in their community using co-
occurrence networks.

Engineering Across the Tree of Life#

Engineering Across the Tree of Life with Integrated Systems and Synthetic Biology
E. Young*, Worcester Polytechnic Institute, Worcester, MA, USA

Nonconventional organisms are attractive hosts because they possess
the biosynthetic capacity for natural products and can tolerate
production conditions. However, compared to model organisms,
little is known about their biology and few gene expression
parts are available for them. Here, we present a workflow for
integrated systems and synthetic biology that unlocks the
potential of nonconventional hosts. With genome sequencing,
comparative transcriptomics, and genetic parts characterization,
it is possible to enable combinatorial metabolic engineering
in underdeveloped microbes. We demonstrate this in the
ascomycete yeast Debaryomyces hansenii. D. hansenii is osmotolerant
and halotolerant, which confers resistance to harsh fermentation
conditions and enables growth in saltwater. It naturally grows
on the major monosaccharides of lignocellulosic biomass,
specifically glucose, xylose, and arabinose. It produces the
nutraceutical riboflavin and overproduces lipids. We first took
a systems approach and performed transcriptomics under stress
conditions to understand pathway regulation in D. hansenii. Using
metabolic network mapping and machine learning, we showed that
D. hansenii extensively and exclusively rewired its
transcriptome to decrease riboflavin production in favor of
generation of flavin cofactors in response to NaCl stress.
We then took a synthetic biology approach, deriving 20 promoters
and terminators to create a standardized, modular part library.
We used the parts to perform combinatorial metabolic engineering of acetyl-CoA and
pathways for the production of fatty-acid derived molecules,
showing for the first time alkane production from pentoses.
We then generalized these approaches across other branches of
the tree of life – the basidiomycete yeast Xanthophyllomyces
dendrorhous and the gram negative bacteria Pseudomonas putida,
Cupriavidus necator, and Komagataeibacter nataicola. In summary,
this approach highlights the potential for integrated systems and
synthetic biology to uncover novel biology and enable efficient
domestication of nonconventional microbes for the creation
of cell factories.

Modes of NP Mechanism of Action#

Mining apt secondary metabolite producers via ‘perturbagenomics’ based natural product discovery
B. Bachmann*, Vanderbilt University, Nashville, TN, USA

The preeminent role that bioactive secondary exometabolites play
in mediating inter-organismal relationships in complex biological
systems is increasingly apparent in biomedical research. The
unique properties of secondary metabolites are on display when
mediating intercellular transactions in chemical ecologies and
manipulating human biological systems toward positive
therapeutic outcomes. Yet the means by which secondary
metabolites are identified from producing organisms using methods
often lack generality (i.e. use cell line cytotoxicity or
target based screening) or do not incorporate bioactivity
directly (i.e. genome mining methods). Therefore, the full
productive capacity of potential secondary metabolite
producing organisms remains untapped. Herein we describe single
cell systems for detection of molecular perturbagens –
molecules that disrupt intracellular and intercellular
processes and provide information about the systems governing
cellular function and intercellular interactions. Using
spectral flow cytometry, cytometric barcoding, and a
‘drag net’ of specific intracellular markers of global cellular
functions, we can detect the full potential of source extracts
containing perturbagens. Identifying biologically active
metabolites is a central methodology across cell and
molecular biology. The methods we present are generalizable
to both primary and secondary metabolites, and to pure cell
lines and biopsy derived primary samples containing mixtures
of cells. Perturbagenomics, the coupling multiplexed activity
metabolomics with biological source derisking via genome
mining methodologies, represents a potent strategy for
revealing the true potential of potential producers and
discovery of both new and known molecules with new
cellular modalities.

Ribosome tagetting by r-gene mining#

Ribosome targeting metabolite revealed via resistance-gene mining
H. Le*, Hexagon Bio, Menlo Park, CA, USA
The ribosome is essential for protein synthesis and targeting of ribosomal function serves as a platform for many therapies. Recent advances in resistance-gene guided genome mining for a priori target identification has emerged as a useful workflow for discerning mode of action before additional work is performed. In this work, we identify the lig biosynthetic cluster which contains a non-reducing polyketide synthase and the ribosomal protein uL16. Through heterologous expression, we identify that the lig cluster produces Ligustrones. Via chemical genetics and isothermal protein shift assay, we reveal that Ligustrone A targets uL16, likely via RNA intercalation followed by covalent addition to a cysteine in uL16. Further analysis of lig cluster uL16, ligH, has undergone a C105V swap, likely to confer resistance to associated Ligustrone A targeting. Taken together, via resistant-gene guided mining, we uncover a novel metabolite-target pair which targets the ribosome with implications for further surveillance of ribosomal protein inclusions within biosynthetic gene clusters.

NPs from Diseased Environments#

An Integrated Pipeline for the Discovery and Production of Bioactive Natural Products from Disease-Affected Environments
Y. Ding*, University of Florida, Gainesville, FL, USA
An Integrated Pipeline for the Discovery and Production of Bioactive Natural Products from Disease-Affected Environments

Natural product research is currently facing two major challenges,
the low rate of discoveries and the limited chemical supply. Recent
advancements in DNA sequencing have revealed a vast reservoir of
untapped natural products encoded within microbial genomes,
presenting unprecedented opportunities for both the discovery
and scalable production of novel bioactive compounds. Herein,
I present an integrated pipeline for mining and producing
bioactive natural compounds from environments associated with
disease. Specifically, our collaborative efforts have yielded
several natural products with antioxidative and anti-bacterial
activities,
such as looekeyolides, epithiodiketopiperazines, and korormicin,
isolated from samples of coral afflicted by black band disease
and stony coral tissue loss disease. Furthermore, we leverage
combined metagenomic and chemical data from specific ecological
niches to unearth new natural products and their corresponding
biosynthetic gene clusters. Our approach has successfully
identified the gene clusters responsible for dolastatin 10
and its natural analogs, providing key insights into the
evolution and structural conservation of this group of
tubulin-interaction compounds in nature.

Yeast Chemical Genomics#

Yeast chemical genomics as a tool to discover antifungal natural
products with potentially new mechanisms of action
N. Brittin, D. Aceti, D. Braun, J. Anderson, S. Ericksen and T. Bugni*,
University of Wisconsin-Madison, Madison, WI, USA

Drug resistant fungal infections represent a growing threat to human health. A recent report estimates that there are over 6.5 million people affected by invasive fungal infections, annually [1]. Moreover, crude mortality rates are estimated around 4 million, annually [1]. Exceedingly high mortality rates and growing multi-drug resistance underscore the need for new drug classes with new modes of action. As part of an antifungal discovery and development program at the University of Wisconsin-Madison, we have built a discovery pipeline using symbiotic bacteria as a source of small molecules. To prioritize hits effectively, we have developed a prioritization platform that relies on computational mass spectrometry and yeast chemical genomics. These complementary tools provide a route to discover antifungals with new structures and new mechanisms of action. For yeast chemical genomics, we have leveraged DNA-barcoded libraries of Saccharomyces cerevisiae. For non-essential genes, we use a knockout collection and for essential genes, we use a temperature sensitive collection. For hit prioritization, we have used the knockout collection as a tool to identify antifungal hits with mechanisms of action that differ from known antifungals. Leveraging pooled collections of DNA-barcoded knockouts facilitates relatively high-throughput sensitivity profiling across knockouts by leveraging DNA sequencing as opposed to testing individual knockouts for sensitivity to antifungal agents. This presentation will focus on how we optimized the yeast chemical genomics platform as well as hit prioritization from high-throughput screening data. Overall, this approach provides a streamlined method to identify novel antifungal agents and to generate mechanism of action hypotheses early in discovery.
1. Denning, D. W. Global incidence and mortality of severe fungal disease. Lancet Infect Dis, In Press.

NPs in Drug Discovery#

Bioactive natural products from a drug discovery perspective
J. Wong*, Novartis Pharma AG, Basel, Switzerland

Natural products and their structural analogues have historically made major contributions to pharmaceutical discovery by yielding hundreds of approved drugs. However, inherent challenges and an emerging reorientation of pharma industry towards other modalities have induced a decline in the contribution of natural products from the 1990s onwards. In recent years, technological and scientific advancements have enabled new opportunities to leverage their potential with regards to the molecular mechanisms and modalities they are involved in, for the benefit of drug discovery. Recent developments in the field of natural products at Novartis will be illustrated through a few examples.

Genome Mining the Phosphonates#

Discovery of Bioactive Phosphonate Natural Products by Genome Mining
K.S. Ju*, The Ohio State University, Columbus, OH, USA
Phosphonate natural products have a storied history of successful application across medicine and biotechnology. Their potent and highly specific inhibition of essential metabolic pathways have empowered their development as antibiotics, herbicides, and industrial agents. This has inspired renewed interest in phosphonate natural products as new drugs. Here, I describe recent findings from our genome mining campaigns that advance our understanding of the chemical diversity and biosynthetic landscape of microbial phosphonates. These include unusual families of phosphonopeptides that inhibit pathogens of animal, insect, and plant pathogens, new branchpoints in metabolism that drive biosynthesis, and enzymatic means for their structural diversification.

Nonheme iron enzymes#

• Xiongyi Huang

Engineering nonheme iron enzymes to catalyze abiological reactions
X. Huang*, Johns Hopkins University, Baltimore, MD, USA
Enzymes that contain metals or metal cofactors catalyze the most challenging and fascinating transformations on earth, including methane oxidation and nitrogen fixation. To support this broad range of chemical reactions, metalloenzymes have evolved with marvelous structural and functional diversity. Our group leverages the immense synthetic potential harbored in this staggering inventory of metalloenzymes to provide solutions to outstanding problems at the frontiers of chemistry and biology. Central to this research endeavor is to bring new concepts to bioinorganic catalysis by drawing inspiration from mechanistic connections between synthetic and biocatalytic systems. As a demonstration of this design principle, we reprogrammed nonheme iron enzymes to catalyze abiological C(sp3)–H halogenation and pseudohalogenation reactions through iron-catalyzed radical relay, a reaction mechanism that is not utilized by naturally occurring enzymes. We established a high-throughput screening platform based on click chemistry for rapid evolution of the catalytic performance of identified enzymes. Given the prevalence of radical relay reactions in organic synthesis and the diversity of nonheme iron enzymes, we envision that this discovery will stimulate future development of metalloenzyme catalysts for synthetically useful transformations unexplored by natural evolution.

Engineering an ERED and KRED cascade#

Engineering an ERED and KRED cascade for the commercial manufacturing of a selective CDK2 inhibitor
R. Lewis*, A. Vargas, M. Burns, M. Karmilowicz and C. Martinez, Pfizer Inc, Groton, CT, USA

Pfizer has recently developed an investigational cancer therapeutic, tagtociclib, which is a selective inhibitor of cyclin dependent kinase 2 (CDK2). The synthetic route to this compound involved the use of two engineered enzymes: a keto reductase (KRED) and an ene reductase (ERED). In this work, we will showcase the efforts to engineer the two enzymes. The KRED enzyme was engineered using a conventional method of single site saturation mutagenesis followed by combining the best mutations. The ERED enzyme, however, presented more of a challenge and was engineered with the assistance of machine learning techniques. Once optimized, these enzymes were implemented at a multikilogram scale in a one-pot biocatalytic cascade reaction.

A Rieske strategy for biocatalysis#

A Rieske strategy for biocatalysis
J. Bridwell-Rabb*, A. Garcia and J. Tian, University of Michigan, Ann Arbor, MI, USA

Rieske oxygenases represent a largely unexplored strategy for selective C-H bond functionalization. These metalloenzymes rely on a Rieske cluster and a non-heme iron center to perform chemistry. The Rieske cluster delivers electrons to the iron and thereby facilitates the reductive activation of molecular oxygen and the formation of an Fe-based oxidant. This activated species is subsequently used to initiate a diverse set of reactions in numerous pathways. Due to the breadth of noted reaction outcomes, Rieske oxygenases have garnered interest for use as biocatalysts for bioremediation and production of pharmaceutical candidates. However, the application of Rieske oxygenases is limited by a lack of knowledge regarding the structural features that these enzymes use to exert control over selectivity and reaction outcome. To this end, we have built extensive variant libraries to establish an understanding of how a Rieske oxygenase scaffold can be manipulated to catalyze a reaction of interest.

Stereoselective ncAmino Acid Synthesis#

Y. Yang*, University of California Santa Barbara, Santa Barbara, CA, USA
Pyridoxal phosphate (PLP)-dependent enzymes versatile biocatalysts underlying amino acid metabolism with outstanding structural and functional diversity. By merging visible light photoredox catalysis and PLP biocatalysis, we advanced a novel mode of pyridoxal radical biocatalysis which is both new to chemistry and biology. Synergistic photobiocatalysis allowed the repurposing of various pyridoxal phosphate (PLP)-dependent enzymes as radical enzymes, leading to novel radical PLP enzymology. Pyridoxal radical biocatalysis provides convergent, stereoselective, and protecting-group-free access to a range of useful non-canonical amino acids, including those bearing a stereochemical triad and tetrasubstituted stereocenters which remained difficult to prepare by other chemical and biocatalytic means. Such non-canonical amino acids represent essential building blocks in peptide therapeutics, natural products and biomedically useful proteins. Furthermore, we demonstrate that the exploitation of biocatalyst-photocatalyst synergy will afford a wider range of stereoselective intermolecular radical reactions with synthetic utility.

An enzymatic macrocyclization cascade#

J. McIntosh, PhD*, Merck & Co., Inc., Rahway, NJ, USA
Peptide macrocyclization is a ubiquitous reaction both in synthetic chemistry and in natural product biosynthesis. While reactions that lead to macrocyclic structures have been well-described both in synthetic and natural systems, there are relatively few examples of peptide macrocyclizations that have been demonstrated using in vitro biocatalysis at high-levels of productivity. Here we describe the discovery and development of a remarkably green and efficient route to a key macrocyclic peptide intermediate en route to a complex API.

Altering correlated motions outside#

4:00 PM 24-6: Altering correlated motions outside the active site preorganizes catalytic residues to create efficient enzymes
C. Pierce, K. Shi, R. Evans, III and R. Kazlauskas*, University of Minnesota, Saint Paul, MN, USA; G. Casadevall, J. Iglesias and S. Osuna, ICREA & Universitat de Girona, Girona, Catalonia, Spain
Biocatalysis, used in the production of specialty chemicals, pharmaceuticals, and biofuels, relies on efficient enzymes to catalyze chemical transformations. However, current enzyme design struggles to identify substitutions beyond the active site that contribute to catalysis, limiting the development of efficient enzymes for biocatalysis applications. We hypothesized that substitutions outside the active site that enhance catalysis adjust the equilibrium positions of the catalytic residues. This theory suggests that residues outside the active site, whose motions are linked to those of the catalytic residues, may alter catalytic efficiency. To test this hypothesis, we converted hydroxynitrile lyase from the rubber tree (HbHNL) into an efficient esterase. Molecular dynamics simulations revealed motions linked to the oxyanion hole residues to be added to HbHNL and other motions linked to the catalytic aspartate to be removed. We created a variant, HNL7TV, which includes the replacement of one oxyanion hole residue (C81L), three changes within the active site (T11G-E79H-K236M), and four changes outside the active site (two near the oxyanion hole residues (V106F-G176S) and two near the catalytic aspartate (H103V-N104T)). The catalytic efficiency of HNL7TV as an esterase is more than 1400 times higher than the original HbHNL and twice that of the similar esterase SABP2. X-ray structure analysis of an intermediate variant, HNL6V, combined with molecular dynamics simulations, showed that the substitutions moved the main chain positions of an oxyanion residue and the His-Asp residues by approximately 0.7 Å, which restored hydrogen bonds needed for catalysis.
This method of identifying substitutions outside the active site that contribute to catalytic efficiency, using a combination of correlated motions and evolutionary information, has the potential to accelerate enzyme design for biocatalysis.