Call for Ideas

We’re excited to present the many excellent submissions generated by Astera’s recent Call for Ideas, which sought innovative solutions to address scaling challenges in synthetic biology.

Bounty-Winning Statements

To determine our five $1,000 bounty winners, each submission was anonymized and then scored by three reviewers, using a rubric that incorporated four criteria, listed in order of descending weight:

  1. Relevance and Impact: Does the idea address the specific scaling challenges outlined in the call? Does the idea have the potential to significantly advance the field of synthetic biology?
  2. Feasibility and Tractability: Is the proposed solution technically feasible and tractable with effort? Can specific actions be taken to implement it?
  3. Novelty and Neglectedness: Is the idea innovative? Does it explore an area that is currently underexplored by others in the field?
  4. Clarity and Simplicity: Is the idea well articulated and easy to understand?

Following this procedure, these five ideas were selected to receive bounties – congratulations to the winners!

Anonymous

Fermentation of Gaseous Carbon Feedstocks

Fermentation of gaseous carbon feedstocks could be the catalyst to sustainability that has so far eluded sugar-based fermentation. In particular, the environmental thesis of microbial synthetic biology has revolved largely around the up-conversion of carbon-sequestering plant starch. However, if the goal is meaningful displacement of petroleum, the continued focus on crop-based sugar may be misdirected.

Learn More

Anonymous

Fermentation of Gaseous Carbon Feedstocks

Fermentation of gaseous carbon feedstocks could be the catalyst to sustainability that has so far eluded sugar-based fermentation. In particular, the environmental thesis of microbial synthetic biology has revolved largely around the up-conversion of carbon-sequestering plant starch. However, if the goal is meaningful displacement of petroleum, the continued focus on crop-based sugar may be misdirected. Its vastly lower abundance vs. petroleum, inherent conflict with the food supply, and questionable benefit (once petroleum-derived fertilizers and total energy use are considered) are among the well-documented critiques. An alternative that, in fact, exists on the scale of fossil carbon and directly mitigates atmospheric carbon is that of CO2, CO, and CH4 emissions themselves. Capture and upgrading of otherwise released emissions from anaerobic digestion (e.g., municipal solid waste) and industrial production (e.g., steel, cement, bioethanol) would decarbonize over 20% of the US carbon footprint. As important, the ability to utilize gaseous feedstocks would provide a long-sought, and easily scalable, pathway for mobilizing the 1–1.5B tons of potentially available lignocellulosic biomass. Despite the prospects of displacing almost 50% of US petroleum consumption, lignocellulose (e.g., energy crops, forestry and agricultural residues) remains largely untapped due to deconstruction challenges. Pretreatment and hydrolysis into monosaccharides are still sufficiently complex such that agricultural sugar continues to be the more competitive option. In contrast, gasification of biomass into syngas (CO2/CO/H2) is a mature technology that could, for example, be added to existing thermal plants at a reasonable cost of $400–600/kW. However, other than the biotech LanzaTech (syngas-to-ethanol), few others have invested in tools and methods for commercial-scale gas fermentation. Moreover, their technology remains proprietary. Therefore, greater — and publicly available — resources for engineering autotrophic and methanotrophic organisms (e.g., Clostridium, Methylocystis), as well as efforts to advance continuous gas bioreactor systems, are critical to enable the next generation of highly carbon negative chemicals.

Piyush Nanda

SELECTION PHASE FERMENTATION STRATEGY

Mitigating climate change requires a rapid switch to biomanufacturing. Scalable biomanufacturing poses a dual challenge: i) effectively driving metabolic fluxes towards the formation of the product of interest (a metabolic engineering challenge) and ii) maintaining the engineered strains in their optimal bioproduction state during fermentation (a bioprocessing challenge).

Learn More

Piyush Nanda

SELECTION PHASE FERMENTATION STRATEGY

“The dream of every cell is to become two cells”
-François Jacob

Problem
Mitigating climate change requires a rapid switch to biomanufacturing. Scalable biomanufacturing poses a dual challenge: i) effectively driving metabolic fluxes towards the formation of the product of interest (a metabolic engineering challenge) and ii) maintaining the engineered strains in their optimal bioproduction state during fermentation (a bioprocessing challenge). Natural selection complicates an engineer’s bioproduction objectives by favoring strains that grow faster throughout fermentation. Such growth dynamics lead to the enrichment of non-producing strains-fast dividing cheaters in the bioreactor, especially in a continuous reactor system (Rugbjerg et al. 2019, Nat. Biotech.). Addressing this challenge requires a new, innovative strategy.

Several strategies are proposed, including deep DNA sequencing for actively monitoring the rise of such mutants. Moreover, recent studies have proposed coaxing strains towards bioproduction by driving an essential gene’s expression proportionately to the product generated – a strategy that will penalize non-producing strains (Rugbjerg et al. 2020). Despite years of research, we have yet to develop an easier-to-implement but scalable strategy. I propose a simple approach to counter-select fast-growing mutants by transiently changing from a ‘production mode’ to a ‘selection mode’.

A risk-benefit tradeoff shapes growth rates. A cell can either allocate resources to reproduction-filling its cytosol with ribosomes and metabolic machinery to grow fast, or allocate a part of the resources towards accumulating proteins it might need during periods of starvation (Basan et al. 2020, Nature; Wu et al. 2022, PNAS). Studying the behavior of microbes in fluctuating environments has established tradeoffs between growth during feast and preparation for famine. Naturally occurring wild isolates of model yeast, Saccharomyces cerevisiae, show significant variations in the growth rate ( Chiara et al, 2022, Nat. Eco. and Evol.). Different strains were likely subjected to varying forms of environmental fluctuations. The presence of abundant glucose, like in a bioreactor, preferentially selects for populations with high growth rates. Meanwhile, periods of feasts followed by prolonged famines select strains that grow slowly but can survive well during famine.

Solution
I propose a new kind of fermentation strategy: Fluctuostat. This involves switching between a production phase (fermentation) and a selection phase (artificial famine induced by stressors). Fast-growing mutants will rise during the fermentation phase but will face a survival challenge during the selection phase, where slower-growing cell factories will get an evolutionary advantage. Of course, the selection phase won’t generate any bioproducts. Therefore, it is essential to estimate the duration of the selection phase so that effective productivity is maximized.

Current studies (Rugbjerg et al. 2019, Nat. Biotech.; Figure 1) reveal a pressing issue: within just 60-70 generations of growth from a single colony to a 200 m3 bioreactor, a significant fraction, i.e. 40% of the population, is overtaken by non-producing strains-a substantial productivity cost to the industry. Determining which stressors will be best suited for the existing biomanufacturing infrastructure is crucial. This strategy presents an unexplored approach to improving productivity by subjecting a population of microbial cell factories to a fluctuating environment.

Kweku Opoku-Agyemang

Modular Bioreactor Systems

Current fermentation systems are often designed for large-scale, single-product processes, which limits flexibility and adaptability. This one-size-fits-all approach hinders the scalability of synthetic biology applications that require customized conditions for different organisms and products.

Learn More

Kweku Opoku-Agyemang

Modular Bioreactor Systems

Concept
Modular Bioreactor Systems for Scalable Synthetic Biology.

The Problem
Current fermentation systems are often designed for large-scale, single-product processes, which limits flexibility and adaptability. This one-size-fits-all approach hinders the scalability of synthetic biology applications that require customized conditions for different organisms and products.

A Proposed Solution
Develop a modular bioreactor system that can be easily reconfigured for different synthetic biology applications. Each module would be designed to cater to the specific needs of different organisms, allowing for optimal growth conditions and product yields. The system would include:

Interchangeable Vessels: Different vessel designs that can be swapped out depending on the organism and product, ranging from micro-bioreactors for small-scale experiments to larger vessels for commercial production.
Adaptive Control Systems: Smart control systems that can adjust parameters such as temperature, pH, and nutrient flow in real-time based on feedback from sensors monitoring the organisms’ growth and product formation.
Scalable Architecture: A design that allows for the addition of more modules to increase production without the need for entirely new systems, thus reducing costs and increasing efficiency.
Integrated Data Analysis: Software that can analyze data from the bioreactors to optimize conditions and predict outcomes, leading to faster iteration and improvement of synbio processes.

This modular approach would allow manufacturers to scale their synthetic biology solutions more effectively, adapting quickly to new organisms and products without significant downtime or capital expenditure. It would also facilitate a more rapid transition from R&D to production, accelerating the path to market for new synbio innovations.

Potential for Lifting All Boats:
By providing a flexible and adaptable platform, this idea could serve a wide range of synthetic biology applications, from pharmaceuticals to biofuels, making it easier for the entire field to scale and meet the growing demand for sustainable and eco-friendly products.

This concept aligns with Astera’s objectives by offering a tangible innovation that could be further explored in workshops and residencies, potentially attracting interest and investment from various stakeholders in the synthetic biology community.

Kenza Samlali

Integrating Microcarriers

Scaling mammalian cell culture for cell therapies is challenging for many reasons, including the cells’ preference for the physiological conditions of their natural environment, leading to low throughput and high costs.

Learn More

Kenza Samlali

Integrating Microcarriers

Scaling mammalian cell culture for cell therapies is challenging for many reasons, including the cells’ preference for the physiological conditions of their natural environment, leading to low throughput and high costs. To address this, microcarriers—100-300 um beads—are becoming a popular solution to increase production capacity by enhancing the surface area of bioreactors. Similarly, in large scale fermentation of micro-organisms, solid state fermentation is an often used method. However, transitioning from screening into scaling still remains challenging as current high throughput screening methods are not capable of screening strains or cell lines for their ability to thrive in such environments. A proposed solution is to integrate microcarriers into screening steps using droplet microfluidics.

TL;DR – Use microcarriers in screening steps early on, offering a means to increase yield and reduce strain failure during scaling from screening to large-scale fermentation.

Eye’s Hand

Open-Source Library of Cell Lines

Creation of an open-source library of metabolomic knock-out CHO cell lines would eliminate one bottleneck in the scale-up of therapeutic protein production that is currently used to treat patients who suffer from cancer and various immunological diseases.

Learn More

Eye’s Hand

Open-Source Library of Cell Lines

My general idea is to create an open-source library of cell lines with protein expression or other revolutionizing capabilities. The library could be a relatively modest collection of a specific cell type e.g. Chinese Hamster Ovary (CHO) cell lines, which are commonly used to produce therapeutic antibodies in the oncology / immunology space. In the case of CHO cell production, one of the industry bottlenecks is the use of antibiotics at large-scale. The over-use of antibiotics across the fields of human health, agriculture and livestock is of grave concern, with the potential to generate antibiotic-resistant pathogens. Therefore, the use of antibiotics in large-scale bioreactors is limited, if not prohibited. A library of metabolic knock-out (auxotrophic) CHO cell lines would allow selective growth of protein-producing cells by simple addition of the necessary metabolite to the growth media, with the advantage over today’s technology of complete elimination of the use of antibiotics in the process.

On the other hand, the library of cell lines could be more extensive to include other mammalian cell types e.g. Human Embryonic Kidney (HEK) cells, which are often used for vaccine production among other industry uses. The library could go even further to include various metabolomic knock-out strains of Saccharomyces cerevisiae (baker’s yeast), which are much more amendable to genetic manipulation than the previously mentioned mammalian cell lines. Baker’s yeast could theoretically be engineered to metabolize environmental hazards (such as oil spills by utilizing petroleum as a carbon source) or produce an industrial product that is otherwise too costly or complex to make. The library could take one step further by incorporating various strains of (non-pathogenic) E. coli or other bacteria that are commonly used for protein production, with much quicker turnaround times due to their faster growth.

Overall, I would love to see someone take this idea and run with it. Currently, the vast majority of CHO and other cell lines are only available through cost-prohibitive licensing / royalty agreements, such that only large corporations have access to the best protein-producing cells. Even then, production of the protein of interest is often linked to the use of antibiotics in the growth media, creating scale-up challenges as well as contributing to the environmental crisis that the global community is beginning to realize. Creation of an open-source library of metabolomic knock-out CHO cell lines would eliminate one bottleneck in the scale-up of therapeutic protein production that is currently used to treat patients who suffer from cancer and various immunological diseases.

Additional Inspirational Ideas

In addition to our winning entries, we are delighted to showcase the diverse and creative ideas submitted by all our participants, which we believe have the potential to inspire further discussion, ideation, and iteration within the community. Consistent with the Call, we are reproducing all of these ideas in their entirety, save for a few that were off-topic. 

Thank you to everyone who contributed their vision and creativity to this initiative!

Karson Chrispens

a generalized platform to engineer export systems that target recombinant products

To avoid the requirement for lysis, one could imagine a system where export machinery is engineered into a cell that targets the desired product, meaning that lysis is no longer required. Native and viral export machinery already works in specialized areas (e.g. engineered extracellular vesicles or RNA export), and developing a generalized platform to engineer export systems that target recombinant products would enable continuous production and harvest of the desired product by collecting the waste media in perfusion culture.

Learn More

Karson Chrispens

a generalized platform to engineer export systems that target recombinant products

One challenge with scaling synthetic biology is the requirement that, if the product is produced intracellularly, you must lyse an entire bioreactor of cells and then purify the product out of the resulting fermentation broth/cell debris mixture. While bacteria and yeast are cheap, in most manufacturing processes, you want to avoid having to destroy the machinery you have built up! Designed proteins, biopolymers, and some chemical products are most easily produced intracellularly, meaning that the chassis being used for growth must be destroyed to gather the resource needed. To avoid the requirement for lysis, one could imagine a system where export machinery is engineered into a cell that targets the desired product, meaning that lysis is no longer required. Native and viral export machinery already works in specialized areas (e.g. engineered extracellular vesicles or RNA export), and developing a generalized platform to engineer export systems that target recombinant products would enable continuous production and harvest of the desired product by collecting the waste media in perfusion culture. Continuous production enables a tighter form of closed-loop optimization, where the various parameters of the system can be tuned in real time to increase yield. One could also add different chassis organisms that act symbiotically in a synthetic consortium and tune the populations in response to production feedback.

Engineered export systems could also enable direct measurement of parameters that are currently unobservable at the scale of a bioreactor, such as the buildup of cytotoxic metabolites within the organisms being grown up.

What will be required (high-level, at my current level of understanding):
1. High-throughput screening methods to identify and characterize effective engineered export systems
1. An engineered export platform could be based on a variety of different mechanisms, including by hacking viral packaging or surface display
2. Novel chassis organisms with engineerable export capacity
3. Integrated hardware and software solutions for real-time monitoring and control of the production process
4. Effective software tools to identify and optimize relevant variables while scaling production

Devon Stork

research on enhancing bacterial natural competence

DNA assembly, amplification and transformation is a bottleneck in modern DNA manipulation, especially for things like library construction that’s necessary for high-throughput science. There’s been a lot of work with using microbial systems to effectively assemble large DNA constructs (yeast & bacterial artificial chromosomes), but such approaches require good transformation and are unsuited to assembling DNA right out of chemical synthesis.

Learn More

Devon Stork

research on enhancing bacterial natural competence

DNA assembly, amplification and transformation is a bottleneck in modern DNA manipulation, especially for things like library construction that’s necessary for high-throughput science. There’s been a lot of work with using microbial systems to effectively assemble large DNA constructs (yeast & bacterial artificial chromosomes), but such approaches require good transformation and are unsuited to assembling DNA right out of chemical synthesis.

However, new research on enhancing bacterial natural competence shows a lot of promise for DNA assembly in-vivo with low-cost. I’m talking about this work: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10863642/ and this one https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4685060/.

I could imagine some kind of system where you dump DNA straight off the synthesis chip (or maybe partially assembled) into media with hyper-competent cells that have yeast-like DNA recombination abilities. Those cells uptake the DNA, finish assembling it and bam, ready-to-go library, cutting out the most laborious and failure-prone steps of library preparation.

Nathan Helm-Burger

genetically modified kelp

Plants growing in water is an efficient way to harvest energy from the sun. Algae farms haven’t worked out so great, in part because it’s easy to get contamination of bred or gene-modded cultures by wild strains. An easier solution to control would be macroscopic plants. Kelp grows crazy fast. Kelp farms already exist as a commercial product. Genetically modified kelp raised in kelp farms would be easier to control than algae.

Learn More

Nathan Helm-Burger

genetically modified kelp

Plants growing in water is an efficient way to harvest energy from the sun. Algae farms haven’t worked out so great, in part because it’s easy to get contamination of bred or gene-modded cultures by wild strains. An easier solution to control would be macroscopic plants. Kelp grows crazy fast. Kelp farms already exist as a commercial product. Genetically modified kelp raised in kelp farms would be easier to control than algae.

Downside: Your genetically modified kelp will definitely escape into the environment if you are using ocean-based kelp farms. And possibly even if you are using artificial ponds with no outlets. You will need to do careful environmental impact analyses beforehand to make sure you won’t out-compete native ecologies.

Some specific things you may want to do with this, that would likely be safe (because it would lower the overall fitness of the engineered kelp, and thus make specimen escape acceptable) include:

1. Kelp engineered to produce a substantial quantity of oils suitable for biodiesel
2. Kelp engineered to be extra digestible, tasty, and healthy for humans or livestock
3. Kelp engineered to produce long fibers which could be cheaply processed into textiles (e.g. like bamboo silk, but cheaper)
4. Kelp engineered to be extra efficient at uptaking nutrients from water. This could be used to surround or intersperse with ocean-based fish farms to reduce environmental impacts of the concentrated fish waste. Also could be used at the mouths of polluted rivers to prevent oceanic algal blooms which are bad for fishing industry.
5. Ammonia-saturated charcoal (biochar) is a useful soil amenity for agriculture, which is also a form of carbon sequestration. Engineered kelp could be a useful source of biomass for producing this. Turn dried kelp into charcoal, saturate charcoal with ammonia, sell as environmentally-friendly fertilizer and claim carbon credits.
6. Engineer kelp to be a particularly useful feedstock for a specific industrial fermentation process. In particular, you could have it produce an unusual protein or sugar or whatever which was broken down efficiently only by your engineered fermentation organism (e.g. yeast or bacteria). This would make it easier to feed you fermentation stock without worrying as much about contamination, since wild-type yeasts and bacteria wouldn’t be able to compete when digesting the engineered kelp.
7. Other places where algae has been considered, you could instead use engineered kelp. For instance, as part of a system for recycling CO2 and biowaste into oxygen, food, and clean water in the context of a space habitat. Kelp is easier for humans to manage, control, and selectively breed because it is a coherent macroscopic organism. It’s easier to keep it contained within a mechanical system and filter it out of the processed water without mechanical fouling or expensive filtration methods.

Adam Marblestone

Assurance against mutations: Fail-safe genetic codes or genetic checksums

Fail-safe genetic codes and/or genetic checksum mechanisms. Basically, genes that don’t or can’t mutate from a starting configuration, or that mutate much more slowly, or that result in self-destruction of the host if they mutate at all. Surprisingly little work has been done on this.

Learn More

Adam Marblestone

Assurance against mutations: Fail-safe genetic codes or genetic checksums


Problem
What we want an organism to produce may create a fitness cost or growth burden, relative to accessible mutations that no longer produce the desired product. Imagine you put an organism on another planet with an engineered enzyme added to its genome — by the time your next probe comes back to the planet, the bugs will have deleted out or inactivated the enzyme and evolved on their own. As a recent empirical paper (https://www.biorxiv.org/content/10.1101/2024.04.08.588465v1) stated: “Engineered DNA will slow the growth of a host cell if it redirects limiting resources or otherwise interferes with homeostasis. Populations of engineered cells can rapidly become dominated by “escape mutants” that evolve to alleviate this burden by inactivating the intended function.”

Solution
Fail-safe genetic codes and/or genetic checksum mechanisms. Basically, genes that don’t or can’t mutate from a starting configuration, or that mutate much more slowly, or that result in self-destruction of the host if they mutate at all. Surprisingly little work has been done on this, e.g., https://academic.oup.com/nar/article/47/19/10439/5568210, which has only 32 citations. Dedicated work on this could uncover other parts of the design space. In 2011 or so, Kevin Esvelt and I worked on a different genetic checksum concept, but never made that much experimental progress nor published it. There are probably much better versions possible now. Positive and negative biosecurity and biosafety implications and mitigations should be considered.

Anonymous

designing viral vectors

ML for designing viral vectors/plasmids specific to certain cells and biological environments. This seems to not exist. Seems like one of those problems we can throw a lot of data and compute and high frequency screening at and things will just work out. There are a couple of software programs for manual viral vector design but automation is ideal.

Learn More

Anonymous

designing viral vectors

ML for designing viral vectors/plasmids specific to certain cells and biological environments. This seems to not exist. Seems like one of those problems we can throw a lot of data and compute and high frequency screening at and things will just work out. There are a couple of software programs for manual viral vector design but automation is ideal.

Patrick Astarita

Flexible hybrid electronic over-pipe differential ultrasonic flowmeter

Developing continuous flow CFPS & Fermentation processes can require a lot of process instrumentation units compared to traditional chemical processes. The baseline cost of an over-pipe differential ultrasonic flow meter is ~$350. If one wants to develop quality proactive controls for their process, they’re going to need a lot of flow meters.

Learn More

Patrick Astarita

flexible hybrid electronic over-pipe differential ultrasonic flowmeter

Developing continuous flow CFPS & Fermentation processes can require a lot of process instrumentation units compared to traditional chemical processes. The baseline cost of an over-pipe differential ultrasonic flow meter is ~$350. If one wants to develop quality proactive controls for their process, they’re going to need a lot of flow meters. Desired is funding and orchestration of a project to develop an open source design of a flexible hybrid electronic over-pipe differential ultrasonic flowmeter, perhaps with thermocouple. Future development could entail the programmatic inference of flow-characteristics and rheological state using ultrasonic data.

Buzz

super dense mycelium applied to a road surface

Road Construction & Maintainence is an incredibly inefficent and waseful process and is arguably detrimental to global climate. I have imagined a super dense mycelium that can be applied to a road surface which will consume the existing road contruction materials as it spreads throughout the road network. The result is a transformed road surface, a living, self perpetuating, and self healing road surface. Lots of scale with less hardware!?

Learn More

Buzz

super dense mycelium applied to a road surface

Road Construction & Maintainence is an incredibly inefficent and waseful process and is arguably detrimental to global climate. I have imagined a super dense mycelium that can be applied to a road surface which will consume the existing road contruction materials as it spreads throughout the road network. The result is a transformed road surface, a living, self perpetuating, and self healing road surface. Lots of scale with less hardware!?

Anonymous

cheap autopipetting

Pipetting isn’t a hard robotics problem. Pipetting is a thing that labs spend a lot of researcher time on. Guaranteeing n 9s of error + making a cheap platform + making it really easy to use puts is in a great position.

Learn More

Anonymous

cheap autopipetting

Pipetting isn’t a hard robotics problem. Pipetting is a thing that labs spend a lot of researcher time on. Guaranteeing n 9s of error + making a cheap platform + making it really easy to use puts is in a great position.

Simon Dürr

the limitations of the PDB format

In order to scale to newer and larger proteins, researchers need to be educated better on the limitations of the PDB format and new software needs to be created and existing one adapted to help the transition to MMCIF.

Learn More

Simon Dürr

the limitations of the PDB format

The MMCIF/PDB problem: Synthetic biology and computation go hand in hand, however most tools are currently making use of the outdated PDB format (including e.g. AlphaFold). This format cannot handle large assemblies (i.e. to build synthetic protein materials) or ligands with the new 5 character IDs since we’ve exhausted the 3 character IDs.

In order to scale to newer and larger proteins, researchers need to be educated better on the limitations of the PDB format and new software needs to be created and existing one adapted to help the transition to MMCIF.

Nathan Helm-Burger

ECONOMIC EFFICIENCY VIA coppicing/pollarding

Engineering plants to produce unusual products (e.g. drugs) can be a difficult endeavor. A potential solution could be to use coppicing or pollarding. Coppicing is where you grow a tree that is good at regrowing from a stump, cut the tree down, and then repeatedly harvest the new growth that sprouts up. Pollarding is similar, but performed at a higher point on the tree, so that more original trunk remains and the regrowth occurs further above ground.

Learn More

Nathan Helm-Burger

ECONOMIC EFFICIENCY VIA coppicing/pollarding

Engineering plants to produce unusual products (e.g. drugs) can be a difficult endeavor. A potential solution could be to use coppicing or pollarding. Coppicing is where you grow a tree that is good at regrowing from a stump, cut the tree down, and then repeatedly harvest the new growth that sprouts up. Pollarding is similar, but performed at a higher point on the tree, so that more original trunk remains and the regrowth occurs further above ground.

Why is this valuable?

Commercial industry is already good at managing orchards. You can easily keep weeds down with physical methods (e.g. mowing). Because of this the engineered trees don’t need to compete with other plants, so growing less competitive engineered trees is easier.

Your crop won’t easily be contaminated with other plants during harvest. Harvesting is relatively cheap and easy, requiring very little human labor (no hand-picking). Plant matter can easily be washed before harvest and/or after harvest.

The trees can be engineered to be sterile, and cloned via grafting, so there is much less risk of escape into the environment. There is also no risk of gene-line contamination from related wild-type plants. The regrown shoots are relatively physically less robust than full trunks, and it is thus mechanically easy to grind them up to extract the compounds. You don’t need to replant for many years, you can keep re-harvesting the same trees. It’s easier to do experiments in the lab on the relevant plant tissues (leaves/branches/sap) than to do experiments with fruit or seeds. You don’t need to wait for the plant to produce fruit/seeds to test how much of the target product the plant can be expected to produce.

A related idea is to harvest the sap (as with rubber trees, maple trees, birch trees), or the leaves (e.g. tea). Or the bark, as with cork trees. My guess is that coppicing/pollarding would be more economically efficient at scale, but it depends on the specifics of the synthetic product you are trying to create and how that interfaces with the cells and macrostructure of the tree. Coppicing/pollarding would work best if the product will be stably embedded in the maturing/dying internal wood of the growing branches. If the product breaks down rapidly, then you probably need to instead harvest more frequently from the living tissues, such as leaves or sap.

Noah Weber

text prompts to create genetic circuits

Develop a novel approach to generate genetic circuits with specific properties by employing conditional stable diffusion on ontologies of Synthetic Biology Open Language (SBOL) guided by text prompts. I posit that this methodology will result in fully correct genetic circuits that satisfy certain biological constraints while providing an intuitive and human-readable interface for the design process.

Learn More

Noah Weber

text prompts to create genetic circuits

What if you can create genetic circuits that exhibit certain phenotypes and fulfill certain functions by only using natural language?

Prompts of a kind: “Design a genetic circuit that acts as a biosensor to detect the presence of a specific molecule and produce a fluorescent output.” Could be used to design genetic circuits in a standardized, machine readable way. This would allow non-biologists to create genetic circuits, the same way midjourney allowed anybody to create images.

Develop a novel approach to generate genetic circuits with specific properties by employing conditional stable diffusion on ontologies of Synthetic Biology Open Language (SBOL) guided by text prompts. I posit that this methodology will result in fully correct genetic circuits that satisfy certain biological constraints while providing an intuitive and human-readable interface for the design process. Additionally, we assume that conditioning the generation on certain reward functions might further enhance the generative process.

There is a longer 10 page document where I outline this research direction in detail.

Will DeLoache

golden gate-esque DNA assembly

I want someone to develop a staggered cutting CRISPR nuclease fused to a DNA ligase that enables scarless, one-pot, in vivo DNA assembly + genome editing of large payloads in any organism, especially those with low endogenous rates of homologous recombination. Basically I want to be able to perform golden gate-esque DNA assembly without needing to remove restriction sites from my constructs while simultaneous integrating those payloads into the genome of a target cell in a precise location.

Learn More

Will DeLoache

golden gate-esque DNA assembly

I want someone to develop a staggered cutting CRISPR nuclease fused to a DNA ligase that enables scarless, one-pot, in vivo DNA assembly + genome editing of large payloads in any organism, especially those with low endogenous rates of homologous recombination. Basically I want to be able to perform golden gate-esque DNA assembly without needing to remove restriction sites from my constructs while simultaneous integrating those payloads into the genome of a target cell in a precise location.

Nathan Helm-Burger

efficient producTION of more robust building material

Structural wood is a major economic product. Being able to efficiently produce large quantities of more robust building material would be very valuable. Bamboo produces a wood-like product, but is related to grass and regrows faster. If timber bamboo could be engineered to have more robust structural properties without loosing too much growth rate, this could make it a more valuable building product.

Learn More

Nathan Helm-Burger

efficient producTION of more robust building material

Structural wood is a major economic product. Being able to efficiently produce large quantities of more robust building material would be very valuable. Bamboo produces a wood-like product, but is related to grass and regrows faster. If timber bamboo could be engineered to have more robust structural properties without loosing too much growth rate, this could make it a more valuable building product.

A downside is that there would be high potential for escape into the environment. Bamboo reproduces prolifically from rhizomes. You could engineer the modified bamboo to be sterile and never produce seed, and then just clone it from rhizomes and/or stalk cuttings. Even still, the potential for prolific spread via rhizome should be kept in mind. Plantations would need to be surrounded by sturdy physical barriers (e.g. deeply buried concrete or plastic walls) to prevent escape.

The tightly packed cellulose of highly valued trees like teak also make them quite slow growing. Rather than aiming for improving the structural qualities of timber bamboo by having it produce more cellulose or related wood structural molecules, it might be feasible to engineer it to produce synthetic hydrocarbon polymers. The resulting plant matter would then be a mix of cellulose and plastic. If the properties of the plastic were chosen correctly, it could be somewhat thermoformable (sticky and malleable at temperatures lower than the combustion point of cellulose). Thus, the timber bamboo harvest could be dried and sawn, then heated under pressure to turn into self-gluing plywood. This would make it cheap and easy to turn into a valuable commercial product.

Likely having indigestible hydrocarbon polymer mixed in with the cellulose would also substantially improve the insect resistance of the crop.
Fire damage would perhaps be extra dangerous, as the burning partially-plastic trees could produce particularly toxic smoke. So the modified timber bamboo should only be grown in wet regions with negligible wildfire risk.

Kweku Opoku-Agyemang

kickstarting the terraforming process

Fertile ground remains a scarce resource. The idea behind microbial soil factories for terraforming is to engineer microorganisms to act as soil factories, converting barren planetary surfaces into fertile ground by producing essential nutrients and organic matter, thus kickstarting the terraforming process.

Learn More

Kweku Opoku-Agyemang

kickstarting the terraforming process

Fertile ground remains a scarce resource. The idea behind microbial soil factories for terraforming is to engineer microorganisms to act as soil factories, converting barren planetary surfaces into fertile ground by producing essential nutrients and organic matter, thus kickstarting the terraforming process.

Anonymous

lack of relevant animal model DATASETS

One of the things limiting the scalability of synthetic biology solutions is the lack of high-quality, comprehensive datasets from relevant animal models, which hinders the development of advanced AI models for understanding cellular and tissue-specific regulation.

Learn More

Anonymous

lack of relevant animal model DATASETS

One of the things limiting the scalability of synthetic biology solutions is the lack of high-quality, comprehensive datasets from relevant animal models, which hinders the development of advanced AI models for understanding cellular and tissue-specific regulation. To address this challenge, I propose generating extensive multi-omic datasets, including transcriptomic and proteomic data (FRO vibes), from carefully selected animal models to enable the fine-tuning of large language models (LLMs) for learning the complex regulatory mechanisms across different cell and tissue types. By supporting the creation of such datasets and the development of AI models, Astera can empower both the research community and the drug development industry to unlock new insights and applications in synthetic biology.

Anonymous

Interfacing biology directly with semiconductors

Much of the progress in synthetic biology can be attributed to its increasing ability to benefit from humanity’s most impressive exponentially increasing technology: semiconductors. The precipitous drop in the cost of DNA sequencing and DNA synthesis can be traced to innovations allowing them to take advantage of capabilities associated with semiconductors (e.g. patterned flow cells, photodiodes, and CCDs for sequencers and patterned silicon and piezo print heads).

Learn More

Anonymous

Interfacing biology directly with semiconductors

Much of the progress in synthetic biology can be attributed to its increasing ability to benefit from humanity’s most impressive exponentially increasing technology: semiconductors. The precipitous drop in the cost of DNA sequencing and DNA synthesis can be traced to innovations allowing them to take advantage of capabilities associated with semiconductors (e.g. patterned flow cells, photodiodes, and CCDs for sequencers and patterned silicon and piezo print heads). The tools used to characterize biology are powered and enabled by semiconductors and progress with machine learning and AI models can again be attributed to semiconductors through compute. Finally, many of the original concepts for synthetic biology were even borrowed from electrical engineering and circuits, which were revolutionized by semiconductor processes. To point this out isn’t to downplay the many other necessary and essential contributions, but to take heed where we might look for other opportunities.

One area that I think synthetic biology has yet to reap all the benefits is with more direct interfacing of biology with semiconductors. In particular, I think three areas are promising:

1) Microfluidics, MEMS, and other on-chip processing capabilities seem under-rated as areas to invest in, to permit them to interface with biology. Once the interface can be cracked, we could likewise expect faster than Moore’s law improvements (as sequencing had) as a field is able to quickly take advantage and ‘catch up’ on the capability offered by semiconductors and associated platforms. These capabilities could both miniaturize what is currently down with robots (and thus kept to a handful of large labs and companies) and could dramatically speed of the design, build, test loop, allowing for more rapid progress in many areas of exploration.
2) The application of compute can also be powerful and seems under-explored. With the advent of GPUs we saw radical progress in graphics, then we saw TPUs for tensor processing, and now we see customized application specific circuits. Why not Sequencing Processing Units (SPUs) for sequencers or Protein Processing Units (PPUs) for protein prediction and design? Investing in application specific architectures can allow for substantial speed ups advancing the state of the art and could have impressive down-stream benefits.
3) Interfacing biology directly with semiconductors. Semiconductor Research Corporation (SRC) attempted something related with SemiSynBio, but largely settled for incremental approaches. Exploring truly interfacing together humanities most impressive achievement in the top-down engineering of state-of-the-art chips with millennia of bottom-up evolutionary design to create the only known molecular machines, seems truly under-rated as a research avenue. For example, there is emerging evidence that many engineered proteins can be functional circuit elements – what if we could directly interface the two and read out single-molecule biological interactions. This could speed up discovery and characterization and lead to several new applications, such as direct DNA sequencing using a polymerase, ultra-low concentration antibody-target binding, etc…

Anonymous

Zymborgs

The innovative concept of Cyborg Enzymes, or Zymborgs, seeks to transcend these limitations by incorporating synthetic and inorganic elements into enzyme structures. This hybrid design maintains the enzyme’s native function while bestowing it with remarkable stability and the ability to utilize electricity as an energy source.

Learn More

Anonymous

Zymborgs

Technological bottleneck limiting innovation in synthetic biology
Enzyme stability and re-usability.

What is the most important class of proteins that enables microorganisms to function and are the targets of synthetic biology projects? Enzymes.
Microorganisms utilize enzymes for metabolic functions and to create vital metabolites. In synthetic biology, scientists manipulate these pathways by adding, removing, or over-expressing enzymes, or by introducing new pathways to achieve specific outcomes. However, this manipulation can cause a metabolic burden, making scaling up challenging and often resulting in high final product costs. This is because microorganisms use most of the input energy for growth and other pathways, not just the desired product pathway. Therefore, the question arises: why not use purified enzymes ex-vivo? This would involve using microorganisms to produce the necessary enzymes, then providing a substrate to directly yield the desired product without the metabolic burden of the microorganism’s other pathways.
Enzymes, being proteins, are limited by their physical and chemical properties, functioning only under mild conditions and having a short lifespan. Protein engineering has enhanced their stability and activity, but they still degrade over time, which hinders their use ex-vivo on a large scale. For example, using enzymes ex-vivo for bioethanol production could theoretically convert glucose to ethanol with high efficiency, bypassing the need for yeast cell biomass, but enzyme instability remains a challenge to achieving this.

Immobilizing enzymes on nanomaterials and using unconventional amino acids are two methods being explored to improve enzyme stability. However, immobilization is costly and can affect enzyme function, while integrating unconventional amino acids into proteins is complex but retains their biological nature leading to instability.

The solution
Zymborgs

Enzymes, nature’s own catalysts, are limited by their organic makeup. The innovative concept of Cyborg Enzymes, or Zymborgs, seeks to transcend these limitations by incorporating synthetic and inorganic elements into enzyme structures. This hybrid design maintains the enzyme’s native function while bestowing it with remarkable stability and the ability to utilize electricity as an energy source. Zymborgs represent a fusion of biology and technology, where the integration of nanomaterials and nanowires into enzymes not only enhances their catalytic activity but also allows them to interface with electronic systems. This groundbreaking approach could revolutionize fields such as nanomedicine and environmental management, offering applications beyond the reach of traditional enzymes.

To actualize Zymborgs, a modular strategy is adopted, dissecting the enzyme peptide sequence into parts, each expressed with non-biological extensions. These segments are then chemically assembled into a Zymborg structure, mirroring the form of natural enzymes. Post-assembly, the non-biological components can be further engineered to tailor the Zymborg’s properties. This method preserves the enzyme’s core structure, ensuring its activity, while integrating additional substances for enhanced functionality and stability. The Zymborg concept thus presents a blend of opportunities and challenges in the quest to augment enzyme performance.

Katherine Baney

Accelerating plasmid production

Plasmids — short, circular pieces of DNA — are the underpinnings of modern biological R&D. They encode genes and other genetic elements used to manipulate the genome. Every new vaccine, antibody, CRISPR gene editing technique, or therapeutic biologic starts with the humble plasmid, or rather hundreds to thousands of plasmids. But despite this ubiquity, they take too long to make.

Learn More

Katherine Baney

Accelerating plasmid production

Problem
Plasmids — short, circular pieces of DNA — are the underpinnings of modern biological R&D. They encode genes and other genetic elements used to manipulate the genome. Every new vaccine, antibody, CRISPR gene editing technique, or therapeutic biologic starts with the humble plasmid, or rather hundreds to thousands of plasmids. But despite this ubiquity, they take too long to make.

From design to experiment, creating a new plasmid takes about two weeks. Accelerating plasmid production will accelerate every iterative design-build-test cycle in synthetic biology. It will not only make research go faster, but it will enable better research because the most critical component — human minds — are no longer dedicating their brain power, creative energy, manual labor, and time to plasmid production.

There are three broad steps of plasmid production, and many, many intermediate steps:

1. DNA Synthesis — stringing base pairs together into a DNA fragment, current tech is limited to ~1000 bp
2. DNA Assembly — stitching DNA fragments together into a circular plasmid
3. DNA Sequencing — reading the base pairs to verify the DNA assembled correctly

Solution
Develop a piece of capital equipment that combines these three processes — DNA Synthesis, Assembly, and Sequencing — so that plasmid synthesis can be done quickly and locally at universities, companies, and other research institutes.

Nathan Helm-Burger

LESS nitrogen fertilizer

It should be possible to engineer plant cells of many crops to have this organelle and thus be able to fix their own nitrogen from the air in their leaves without the need for symbiotic nitrogen-fixing bacteria. Thus, need less (or no) nitrogen fertilizer.

Learn More

Nathan Helm-Burger

LESS nitrogen fertilizer

Yet another specific concrete problem, and possible solution. The following paper describes the discovery of a nitrogen fixing organelle in certain algae: https://www.science.org/doi/10.1126/science.ado8571

“These findings show that UCYN-A has evolved from a symbiont to a eukaryotic organelle for nitrogen fixation—the nitroplast—thereby expanding a function that was thought to be exclusively carried out by prokaryotic cells to eukaryotes.”

It should be possible to engineer plant cells of many crops to have this organelle and thus be able to fix their own nitrogen from the air in their leaves without the need for symbiotic nitrogen-fixing bacteria. Thus, need less (or no) nitrogen fertilizer.

Anonymous

TANGIBLE BENEFITS IN Syn-BIO Products

My view is that some of the biggest problems for syn-bio are not technical but cultural and I believe we need to take a cultural approach to fixing them. My proposed solution is that instead of focusing on therapeutics and inputs for R&D, emerging syn-bio companies should instead transition to rethinking consumer products.

Learn More

Anonymous

TANGIBLE BENEFITS IN Syn-BIO Products

Regulatory bottlenecks prevent wide adoption and high scale of many commercially valuable biologics. The reason for this is because bio-products are typically marketed as therapeutics or need to meet high purity standards for application in research/diagnostics, requiring heavy scrutiny from regulatory bodies and limiting the applicable market for the product. Synthetic biology companies have historically excelled in producing large datasets and overcoming technical hurdles to genetic transformation but have failed in achieving high scale and wide adoption of their products. My view is that some of the biggest problems for syn-bio are not technical but cultural and I believe we need to take a cultural approach to fixing them. My proposed solution is that instead of focusing on therapeutics and inputs for R&D, emerging syn-bio companies should instead transition to rethinking consumer products. Recent successes such as K-18 and ZBiotics demonstrate there is consumer appetite for syn-bio products at a high price premium if a tangible benefit can be engineered into the products. These examples also demonstrate how removal of GMP guidelines and clinical regulatory barriers enable a bio-product to make it to market with a fraction of the cost and time required by a typical pharmaceutical company.

David Specht

natural competence

Synthetic biologists could create an organism in which natural competence is a perpetual state of being, rather than a panic button. Instead of conveying permanent genomic edits to cells, non-replicating DNA transcripts could be fed to cells to give them on-the-fly instructions of what to do.

Learn More

David Specht

natural competence

All living things jealously guard themselves from invading genetic material. This is for good reason: the world is a perilous place, full of dangers like viruses and other opportunistic parasites. Despite the enduring power of evolution to drive change, life is resistant to instruction.

This presents a problem: living things, as much as we might like them to be, are not computers. They are not indefinitely reprogrammable as a function of external inputs.

Consider instead a microbe which unconditionally takes up, reads, executes, and then discards any genetic material that it is ‘fed’. Such an organism is a punch card reader, reading the transcripts of life (nucleic acids DNA and RNA) rather than holes in a sheet. Such an organism would be indispensable for scientific research and malleable to biological engineers in ways presently unattainable.

This isn’t pure fantasy. Many microbes exhibit so-called ‘natural competence,’ taking up DNA from their environment under tightly-controlled conditions, often as a stay on death itself. They give a little on what makes them a species and open themselves to whatever genetic lifelines might exist in their surroundings to pull them from the brink.

Synthetic biologists could create an organism in which natural competence is a perpetual state of being, rather than a panic button. Instead of conveying permanent genomic edits to cells, non-replicating DNA transcripts could be fed to cells to give them on-the-fly instructions of what to do. This is akin to running a command in a computer terminal, rather than changing the computer’s behavior by opening it up and fiddling with the internal circuitry.

How could this be useful? Consider an extremely flexible system for production of biological compounds. Bioreactors full of cells making critical compounds like amino acids or fuels could be reprogrammed on-the-fly to meet market demand as needed. New discoveries improving the production process could be conveyed via an ‘over-the-media’ update fed to the microbes. Cells could be made to be entirely dependent on the transient DNA for survival, creating a powerful system for biocontainment. An interplanetary mission equipped with an enzymatic DNA printer could reprogram cells recycling CO2 into food to produce a medicine without ever having to crack the bioreactor open, risking contamination and mission failure.

Kenneth Bryan Hsu

a new model system for developing & producing synthetic products

The key difference between existing GMOs and this modified plant strain is that rather than optimizing a specific species for enhanced characteristics, this new strain would be a general purpose workhorse production platform suitable for use producing product or performing a function, analogous to how E. coli is routinely used for cloning or protein production in the lab today.

Learn More

Kenneth Bryan Hsu

a new model system for developing & producing synthetic products

Problem
Most synthetic biology approaches use organisms that are well understood in terms of their biology and methods of manipulation. Using systems like E. coli, yeast, cell cultures, etc. are convenient to prototype but have specialized sterility and nutrient requirements (media, selection markers) that make it cost-prohibitive and logistically difficult to scale to the level of conventional industries like commodity chemicals. The scale of the output of the synthetic biology product is limited to that of specialty pharmaceuticals, which, to the end user, makes little difference in terms of the method used to produce or engineer the product they are using and discourages the use of synthetic biology approaches.

Solution
To address these challenges in feeding and maintaining synthetic systems beyond the laboratory scale, I propose the use of using an engineered plant species as a new model system for developing and producing synthetic products. This would eliminate the need for specialized production processes, and takes advantage of the existing infrastructure and techniques developed for farming of genetically modified plants. The key difference between existing GMOs and this modified plant strain is that rather than optimizing a specific species for enhanced characteristics, this new strain would be a general purpose workhorse production platform suitable for use producing product or performing a function, analogous to how E. coli is routinely used for cloning or protein production in the lab today.

Scaling production would be achieved through simply planting more fields of the desired strain. Farmlands are open to the wider biosphere, and so this platform will need to be engineered to not be able to release its genetic material unless directed to do so in a controlled environment in order to continue to propagate the strain. A particularly interesting example of controlled reproduction is in the pothos Epipremnum aureum, which has a defective gene responsible for producing gibberellins essential for flowering and will not flower naturally, but will readily flower when sprayed with artificial gibberellins.

This system will scale as quickly and as large as necessary with minimal modifications to a robust and well-developed agricultural industry.