Tom Ellis

Boolean logic-gated protein presentation through autonomously compiled molecular topology - Nature Chemical Biology Programming stimulus responsiveness into living systems enables advanced biocomputation. Here, the authors autonomously compile proteins with defined topology that can be site-specifically tethered to and conditionally released from biomaterials and cells following user-specified Boolean logic.

Latest issue of curated #engineeredlivingmaterials feed:
biomed.news/bims-enlima/... #ELMs
via @biomednews.bsky.social

includes research on: boolean logic-gated protein presentation (h/t @coledeforest.bsky.social), bacterial grown living materials (h/t @proftomellis.bsky.social) & more 🧪

Grow Your Own Cloud Grow Your Own Cloud is a green biotech company developing regenerative data solutions through DNA data storage in Plants 🌱

Storing data in plants

growyourown.cloud

Loved the bit about ‘See It Say It Sorted.’ My 4 year old is convinced it is “See It Say It Sausage” and even shouts that on the train when the announcement is made. I can never unhear it.

Intracellularly Coupled Oscillators for Synthetic Biology https://www.biorxiv.org/content/10.1101/2025.09.23.678067v1

Yesterday was the first day of autumn and guess what appeared in the window of the Peninsula hotel in London.

A defined microbial community reproduces attributes of fine flavour chocolate fermentation - Nature Microbiology An in-depth microbiological and metagenomic analysis of Colombian farm and fermentation facilities resulted in the design of a defined microbial community that can reproduce the flavour of fine chocol...

www.nature.com/articles/s41...

Data-driven de novo design of super-adhesive hydrogels - Nature A data-driven approach integrates data mining, experimentation and machine learning to design high-performance adhesive hydrogels from scratch, tailored for demanding underwater environments.

Nature’s glues.

www.nature.com/articles/s41...

astly I also want to shout-out to Morgane Boone and Nico Callewaert whose SECRiFY project at VIB Ghent gave us lots of the inspiration for this work. www.nature.com/articles/s41...

It's been great to see this work develop under Stacey's expertise in our lab and so glad that it's finally out. Klaudia Ciurkot played a big role in applying the POLAR-Seq part so is an author too. We also thank Wolfgang Ott for the ELPs which made great test cases.

There's more to this paper too but I'll leave it here as this is the key to the work. Modular libraries, FACS-based YSD-enabled screening and POLAR-seq will fast-track finding an optimal design. And then this gene cassette design can be kept as the design used for good secretion.

But in all these cases, the protein is still surface-displayed and not actually secreted, so what remained was to show that once the screen had identied a good design, that it was still a good design if it was simply secreted. Thankfully it was!

This was an explosion of data - showing us the best and worst combinations of promoters and signal peptides to use for each target protein to reveal the importance of these modular parts for each case. Very useful data for informing future designs, training AI and maybe uncovering rules.

Stacey and Klaudia applied POLAR-Seq to the libraries of yeasts we were making with different cassette designs for secretion of target proteins. The FACS-based screen sorted the libraries of yeasts into good, medium and bad secretors and POLAR-seq then determined all genotypes.

Combinatorial Design Testing in Genomes with POLAR-seq Synthetic biology projects increasingly use modular DNA assembly or synthetic in vivo recombination to generate diverse combinatorial libraries of genetic constructs for testing. But as these designs ...

But here's where it gets a bit cooler - we recently developed POLAR-Seq in our lab which can be used to directly sequence all the different genetic designs found in a targeted region of the yeast genome in a pool of cells from a combinatorial library. www.biorxiv.org/content/10.1...

So armed with the modular toolkit, the YSD-based method and a FACS-based screen-and-sort approach, Stacey was able to engineer yeast cells to secrete classic enzyme targets like beta-lactamase, as well as complex structural proteins such as Elastin Like Polypeptides (ELPs).

As others have shown before, this YSD-based approach means you can label cells with epitope-binding fluorescent antibodies and then use flow cytometry to sort the ones with the most of your target protein attached to them. These should be the best secretors.

But this required a screening method that could test 1000s of cells for secretion of a target protein. For this she adapted Yeast Surface Display (YSD) to attach epitope-tagged proteins to the outside of the cells that secreted them. This was done in modular format to fit YTK.

We and many others use the YTK MoClo system to assemble genes from parts in yeast and it's ideal for generation of combinatorial libraries. Stacey reasoned that this could be adapted to make libraries of gene cassette designs that can be screened for secretion of a given protein.

This paper is all about taking a combinatorial approach to the age-old biotech problem of how to design a gene cassette to secrete a protein of interest from yeast. What strength promoter to use? Which signal peptide? The best choice varies widely depending on the protein target.

Another new paper from our lab - this time from Anastasiya Kishkevich who has developed "YTK Display-and-Secrete" a library-based method to identifying genetic designs good for protein secretion from S.cerevisiae yeast. Online now at ACS SynBio - pubs.acs.org/doi/10.1021/...

Genetic entanglement enables ultra-stable biocontainment in the mammalian gut https://www.biorxiv.org/content/10.1101/2025.08.13.670093v1

🚨 What do you mean?
🇩🇯 Normal people always
👏 write like this

Jobs We are working to solve the biggest mysteries in bioscience.

We are hiring a scientist to lead our 'Build' team at the Seattle Hub for Synthetic Biology. @alleninstitute.org

Job posting below:
alleninstitute.org/careers/jobs...

Please reach out if you have any questions!

This paper was led by Jane during her PhD and included some of Glen's unpublished PhD work as well as key contributions from Klaudia Ciurkot (POLAR-Seq) and Will Shaw (CRISPR) @willshaw.bsky.social
- Always a pleasure to have amazing people like this in my group and talk about their work.

These datasets and others like it from our lab are now being used to help us train AI models for synthetic yeast genome design that can consider gene context and arrangement effects - features hard to assess using natural genomes. Keep an eye on @biorxiv-synthbio.bsky.social for more soon.

ogether, Jane and Glen's work provide important insights into the effectiveness of gene rearrangement in synthetic genomes for improving fitness and identifying design changes for optimisation. They also offer unique data on how altering gene context changes gene expression.

Nanopore sequencing, key also to POLAR-Seq, gave Glen the genotypes of the yeast strains and allowed us to understand the gene rearrangements in the synthetic genome and hypothesize on why they boost GFP expression, but cannot easily be improved on.

In his PhD, Glen SCRaMbLEd a synthetic yeast chromosome multiple times while selecting for maximum GFP expression from an unSCRaMbLEd plasmid. He saw quick progress on generating strains with 35% more GFP expression but by 3 rounds of SCRaMbLE it maxxed-out at 55% more.

Jane then did further rounds of SCRaMbLE with the best performing of the rearranged strains to see whether fitness could be increased further and saw it maxxing out after just 2 rounds. This quick progress to a limit matched previous unpublished work by ex-lab member Glen Gowers.

This fast and affordable workflow gave us functional fitness data on thousands of rearranged synthetic genotypes - a real treasure trove. For our poor performing design, which had a weak His5 promoter, the dominating rearrangement that led to improved growth was His5 duplication.