Research

We use modular parts libraries and combinatorial pathway engineering to make a host overproduce a target molecule.

Debaryomyces hansenii. Grows in extremely salty water on lignocellulosic biomass and overproduces fatty acids, which can support on-demand manufacture of oleochemicals like omega-3 fatty acids, alkanes, or lubricants.

Xanthophyllomyces dendrorhous. Grows on lignocellulosic biomass and overproduces terpenes, a diverse class of molecules that has long held promise for making medicines, fuels, and chemicals.

Taxus chinensis. In collaboration with Prof. Susan Roberts we extend this to plant cell culture, engineering suspension cell lines to overproduce paclitaxel, including CRISPR-guided control of supporting pathways.

Co-advised PhD work with the Roberts lab at WPI.

We have developed a unique fungal-bacterial system that can send signals centimeters underground. The system uses bacteria with a biosensor to detect a chemical, and then fungal mycelia to send the detection signal to the surface. This required a great deal of genetic circuit optimization to make the system work in actual dry soil. We have developed a method of sequential screening to filter out candidate genetic circuits and plan to use this to develop diverse biosensors that work in actual soil.

This work was funded by DARPA through the BioReporters and Ceres programs. US Patent 12,571,789 B1 has been awarded, and the results are in a preprint.

With RTX BBN Technologies.

We build the genomic foundation and modular parts that make a new host engineerable — nonconventional yeasts, diverse bacteria, and plant cell cultures. Our key strategy is merging bioinformatics and development of modular genetic parts libraries to facilitate high-throughput combinatorial pathway engineering in these organisms.

Parts collections. Xanthophyllomyces dendrorhous, Debaryomyces hansenii, Pseudomonas putida, Cupriavidus necator, and Komagataeibacter nataicola.

Genomes and transcriptomes. Kregervanrija delftensis, Ogataea polymorpha, probiotic yeasts and Candida auris (with Reeta Rao), and oleaginous yeasts by comparative transcriptomics.

Software and bioinformatics. We have developed processes to integrate genome sequencing into strain development workflows to verify accuracy of genetic engineering. This same approach can be used to identify genetic engineering from complex environmental samples in biosecurity contexts. The same assembly work reads yeasts that were never engineered at all: with Reeta Rao we resolved the genetic basis of probiotic yeast phenotypes and the virulence and drug tolerance of Candida auris. With Claudia Vickers we applied PRYMETIME to engineered yeast, showing that plasmid integration diversifies a strain. We also curated synthetic biology knowledge from the literature into a searchable system, combining text mining with curation principles drawn from the SBOL iGEM data set, so that published designs and data could be found and reused.

Automation. We work on laboratory automation for genetic engineering, and build it as shared infrastructure through the CREATE Biofoundry — the NSF Center for Robust, Equitable and Accessible Technology for Next-Generation BioFoundries, with Mark Blenner at the University of Delaware and Howard Salis at Penn State. CREATE builds distributed automation and design capability for proteins, biosensors, and bacteriophage products, and puts those tools within reach of institutions that could not run a foundry of their own — primarily undergraduate institutions, HBCUs, minority-serving institutions, and women's colleges across the Northeast and Mid-Atlantic.

Scale. Automation is necessary for combinatorial engineering, and fermentation capacity is necessary to turn a strain into a process. The WPI BioHub, one of five NSF biofoundries nationally, provides fermentation from 1 L to 100 L, process development, and trained personnel, so that a product developed at the bench can be scaled in Central Massachusetts. It was launched with $5.2M from the Massachusetts Technology Collaborative and built on the WPI Cell Engineering Research Equipment Suite (CERES) and Bioprocess Center.

Knowledge curation with Chris Myers and a multi-institution team.

We build bacterial biosensors that detect a chemical underground and use fungal mycelia to carry the signal to the surface, over centimeters of soil.

Making this work in actual dry soil took a great deal of genetic circuit optimization. Comparing circuit output across labs and instruments requires calibrated fluorescence measurement. We use Pseudomonas putida for the sensing function because it tolerates the chemical stress of a field environment. This work was funded by DARPA through the BioReporters and Ceres programs, and US Patent 12,571,789 B1 has been awarded. The results are in a preprint.

Draws on Genetic Circuits · Organism Onboarding

Feedstock cost determines whether a bio-based product is viable. We work with hosts that grow on inexpensive inputs: lignocellulosic biomass, water too salty for most organisms, and CO₂.

Debaryomyces hansenii overproduces fatty acids while growing in extremely salty water, a route to alkanes and lubricants. Xanthophyllomyces dendrorhous makes terpenes from lignocellulose. Cupriavidus necator fixes CO₂ directly. Scaling any of these from a strain to a process requires fermentation capacity, which the BioHub provides.

Draws on Organism Onboarding · Metabolic Engineering · Biofoundries

Paclitaxel is still sourced from yew trees or made semi-synthetically. With the Roberts lab we engineer Taxus chinensis suspension cell lines to overproduce it, including CRISPR-guided control of the supporting pathways that compete for the same precursors.

D. hansenii produces fatty acids, a route to omega-3 fatty acids, and the terpenes from X. dendrorhous are precursors to a large class of drugs.

We surveyed the field in Genetic engineering of host organisms for pharmaceutical synthesis.

Draws on Organism Onboarding · Metabolic Engineering

Bacterial nanocellulose is pure and strong, and bacteria grow it directly. We have characterized cellulose-producing Acetobacteraceae and shown that the carbon source changes the material: Komagataeibacter hansenii grown on arabitol yields optically clear cellulose.

Onboarding these organisms enables genetic tuning of the material. Cupriavidus necator extends this to materials built from CO₂.

We surveyed the field in Synthetic biology for bio-derived structural materials.

Draws on Organism Onboarding · Metabolic Engineering

The same genome reading that verifies a strain we built can also determine whether an unknown sample was engineered.

We built assembly methods that recover engineered constructs intact from mixed samples, and contributed to ensemble detection that flags engineering signatures across many target organisms without requiring an expert to review each case.

Reading a genome requires collecting a sample first. The biosensors and circuits we build for soil report from the field without a sample.

Draws on Biofoundries · Genetic Circuits