Week 1 HW: Principles and Practices

1. Introduction

With a rather limited background in the field of synthetic biology and bioengineering, I sketched out my initial scope of interest in closed-loop controllers, in which they are autonomous and adjust to the environment around.

While I’m also interested in the bidirectional communication via the gut-brain axis. I want to explore the idea of engineering a gut bacterium with a synthetic genetic circuit that could detect biomarkers in the gut and conditionally produce neuroactive compounds that modulate brain activity via the GBA.

The circuit should ideally consist of a sensor module, processing module, and a response module. The logic is elucidated as following:

Inflammation detected → threshold exceeded → produce calming molecules → inflammation decreases → production shuts off.

This idea draws distinction from those open-loop, stress-relieving gummies and pills in that, this is a self-regulating therapeutic that produces compounds at the site where the gut-brain signaling infrastructure exists, and only produces upon conditional activation when the stress/inflammation biomarker exceeds a certain threshold.

2. Governance Goals

The overarching goal is Non-Malfeasance (preventing harm)

The nature of the technology involves releasing a genetically engineered organism into the human body, and potentially into the broader environment, making harm prevention and the Dual Use Research Concern (DUrC) indispensable presences and should be carried out at multiple scales.

SubGoal 1A: Preventing Uncontrolled Spread and Ecological Contamination

The engineered microbe must not exist beyond its therapeutic window, which means it should by no means spread to unintended hosts, or transfer its synthetic genes to wild microbial populations via the following possible routes:

• Horizontal gene transfer (HGT): Synthetic circuit components (especially antibiotic resistance markers used in cloning) could transfer to pathogenic gut bacteria.
• Environmental shedding: Engineered bacteria will be excreted and enter wastewater and soil ecosystems.
• Mutation: The organism could evolve and mutate overtime to the point where the original means of control no longer works, or it can gain unintended functions.

SubGoal 1B: Preventing Negative Neurological/Immunological Effects

The closed-loop circuit must not overproduce compounds that trigger immune reactions within the body or interferes with the existing microbiome in unintended ways, such as:

• Overproduction toxicity: A sensor that is too sensitive or a failed threshold filter could flood the gut with GABA/serotonin precursors.
• Immune overactivation: The engineered organism might trigger inflammatory responses, paradoxically worsening the target condition.
• Microbiome disruption: The engineered organism at therapeutic densities could outcompete native beneficial bacteria.

SubGoal 1C: Informed Consent

Governance must address who gets access and whether patients can meaningfully consent to hosting a living engineered organism, as the commitment is larger than taking in a single pill.

3. Potential Actions

Three potential governance actions are considered below, incorporating 1) Purpose, 2) Design, 3) Assumptions, and 4) Risk of Failure and “Success”.

Governance Action 1: Comprehensive policy framework and clear assignment on roles played by different actors

Purpose: The work conducted with living organisms in making them biotherapeutic product usually fall under FDA’s established framework of CBER, but due to the closed-loop nature of the synthetic circuit, there are no detailed requirements/regulations revolving around how to exert controllable influence that distinguishes from the treatment of those open-looped projects.

Design: Given the participation of various actors, when FDA issues the guidance, academic labs should design/provide corresponding biocontainment tools. While biotech companies comply and absorb testing costs. Research agencies should then standardize biocontainment toolkits to lower barriers for smaller labs. Cross-agency coordination with environmental protection agencies (e.g. EPA) may be needed.

Assumptions

• Effective switches can be engineered over time to keep the microbiome in check
• FDA has sufficient synbio experts in evaluating the circuit design
• In vitro stability testing predicts in vivo behavior

Risks

• Failure: IF the standards were set too high making the project difficult to perform, it could lead to the decline in industry as small labs and startups may choose to opt out.
• Success: A standard designed too well could lead to underestimation of risks.

Governance Action 2: Long Term Monitoring and Clinical Trials

Purpose: Given the closed-loop nature and the potential changes that could occur in living therapeutics, clincal trial framework should establish different tiers that occurs over a designated timescale for constant surveillance.

Design: The clinical trials should develop at least three tiers, with

• Tier 1 (1-3 yr): Standard testing phase
• Tier 2 (5 yr): Mandatory microbiome monitoring and tracking of genomic sequences
• Tier 3: Constant survillance of wastewater disposal in experimenting/trial regions

Assumptions

• Patient will remain in 5 year follow up
• The engineered organism can be effectively tracked within gut environment

Risks

• Failure: Unforseen development of organism is sighted after widespread distribution.
• Success: Over institutionalized framework could slow development of future iterations.

Governance Action 3: Transparency and International Oversee

Purpose: In considering the potential widespread use of such ideation, the public should gain transparency to the fundamental logic/codes. Simultaneously, international harmonization groups like WHO should develop and align the set of harmonized minimum standards for testing and monitoring.

Design: National governments in coordinating and aligning regulations under international organizations and synbio industry leaders. Commited collaboration between public and private sectors in a foreseeable timescale.

Assumptions

• Committed support among decision maker exists despite current issue in international relations.
• Applicable universal standard despite different cultural practice
• Development of technology be in pace with international harmonization.

Risks

• Failure: No actual efforts of enforcement made.
• Success: Rigorous standards that further stabilize the advantage of developed countries, and enlarge the medical development and accessibilities between countries.

4. Scoring Framework

The following rubric evaluates the governance options presented above on a 1–3 scale (1=week/limited, 2=moderate, 3=strong) across the span of biosecurity, lab safety, environmental protection, and practical considerations.

5. Prioritized Option

Given the overall scoring, Governance Action 2 yields the highest total amongst the three, because the design in stages of trial over a timescale monitors the progress of experiment closely and allows for early detection of incidents. The gradual development also allows brings the market into consideration, making the idea of wide application possible.

However, it also contain weakness that needs to be accompanied by complementary actions. Specifically on prevention, Action 1 scores higher in that it implants kill switches in the initial engineering phase.

Action 3 touches a little bit of everything, but it should be of a later consideration when the technology and domestic standards became more mature, as implementing regulations on an international level generates huge costs and often require longer time for reconciliation/negotiation.

Assignment:

Questions from Professor Jacobson

Nature’s machinery for copying DNA is called polymerase. What is the error rate of polymerase? How does this compare to the length of the human genome. How does biology deal with that discrepancy?

The error rate, according to slide 8, is 1:10^6. The human genome as noted is 3.2 billion base pairs (gbp), and hence if we were to do the calculation there would be around three thousand new mutations/cell division. The biology deals with the discrepancy through error correction like MutS Repair System, that detects the mismatched base pairs and resynthesize it correctly, therefore bringing down the error rate and enabling the copying to proceed with very few/zero errors.

How many different ways are there to code (DNA nucleotide code) for an average human protein? In practice what are some of the reasons that all of these different codes don’t work to code for the protein of interest?

An average human protein is encoded by around 1036 base pairs of DNA (slide 6), and divided by three (codon) will get roughly around 345 amino acids/protein. So given the number, there’s around 10^150 possible DNA sequences that result in the same primary chain of amino acids. But the majority are redundant, and in some situations a sequence of amino acid would create mRNA structures like hairpin that blocks the ribosome from binding and the forming of right protein.

Questions from Professor LeProust

What’s the most commonly used method for oligo synthesis currently?

The most used method is the phosphoramidite method, which is a 4 step chemical cycle that repeats for N times, specifically including coupling (with phosphoramidite), capping (unreacted sites), oxidation, and deblocking.

Why is it difficult to make oligos longer than 200nt via direct synthesis?

It is difficult mainly due to the inefficiency of the coupling steps and the accumulation of errors, given the exponentially decaying yield, as the error rate accumlates, the majority would be of failure sequence by the time it reaches 200.

Why can’t you make a 2000bp gene via direct oligo synthesis?

Because the direct oligo synthesis is performed via phosphoramidite, and due to the multiplicative nature of the success rate and the final yield follows an exponential decay curve, as the number of nucleotides increases, the accuracy will go down. By the time it reaches 2000, it would be hardly possible to extract the correct sequence among all disturbances and noises. Hence bioengineers synthesize smaller oligos and stitch them together to ensure the correct sequence.

Question from Professor Church

What are the 10 essential amino acids in all animals and how does this affect your view of the “Lysine Contingency”?

The 10 essential amino acid (from the slide and with the aid of google) are listed below:

• Arginine (Arg)
• Histidine (His)
• Isoleucine (Ile)
• Leucine (Leu)
• Lysine (Lys)
• Methionine (Met)
• Phenylalanine (Phe)
• Threonine (Thr)
• Tryptophan (Trp)
• Valine (Val)

The Lysine Contingency (according to Google) refers to the genetic alteration performed in the movie Jurassic Park, that made dinosaurs unable to produce lysine, therefore relying on human supplements to survive. But this idea does not stand as it is an essential amino acid within them that doesn’t need to be synthesized, and hence dinosaurs can gain lysine by eating other organisms. This idea sheds light on the biocontainment method of NSAA (non standard amino acid), which organisms cannot obtain in a natural setting, and hence is a more secure contingency.