Week 1: Logic-Gated CAR-NK

🧬 Logic-Gated CAR-NK for Solid Tumors

Project Proposal: Regulatory & Ethical Frameworks (2026)

Project Curiosity: I am interested in the development of Logic-Gated CAR-NK cells specifically for solid tumors. While CAR-T therapies have revolutionized the treatment of hematological cancer, they struggle in solid tumors due to the hostile tumor microenvironment, antigen heterogeneity, and manufacturing complexities.

1. The Engineering Concept

Using Boolean Logic Gates (such as AND, OR, and NOT gates):

• AND Gate: Ensures the cell attacks only if two specific antigens are present.
• NOT Gate: Stops the attack if a healthy tissue marker is detected.
• Manufacturing: Using allogenic sources such as cord blood or iPSCs allows these therapies to be manufactured off-the-shelf, potentially addressing the bottleneck of expensive, personalized autologous therapies.

2. Governance & Policy Goals

Core Goal: Ensure safety and accessibility in the deployment of advanced cellular therapies.

3. Governance Actions

Action A: Standardization of Logic-Gate Stress-Testing Protocols

Purpose: Current testing for CAR-T occurs in idealized laboratory conditions that do not reflect the complex, suppressive tumor microenvironment of a solid tumor. We must implement a regulatory “stress testing” logic gated under stimulated Tumor Microenvironment Conditions (e.g., hypoxia, high antigen density) before human trials.

• Design: During the Investigational New Drug review, developers would be required to submit data showing that the “NOT” or “AND” gates function correctly even when the CAR-NK are exhausted or exposed to immunosuppressive factors, such as TGF-beta. This ensures the off switch doesn’t fail under physiological stress.
• Assumption: This assumes that in vitro, the TME models (like 3D organoids or tumor-on-a-chip systems) are sufficiently advanced to accurately predict how these genetic circuits will behave inside a human body.
• Risks (Failure): The logic gates might work in the model but fail in the patient due to unforeseen biological interactions (patient-specific antigen shifts), leading to severe toxicity.
• Risks (Success): If testing standards are set too high, they could become a barrier to entry, preventing smaller academic labs or startups from advancing new therapies and centralizing control among a few large pharmaceutical companies.

Action B: Mandatory Long-Term Registry

Purpose: Gene therapies carry long-term risks, such as insertional mutagenesis or unexpected survival or engineered cells. We propose a mandatory, centralized patient registry designed to track and monitor requirements enforced by regulatory agencies like the FDA.

• Requirements: Manufacturers must track patients for over 5 years. This registry would aggregate data across different products to identify class-wide safety signals (e.g., the “NOT” gate consistently failing after 6 months). It will require cooperation between hospitals, developers, and regulators.
• Assumption: This assumes that patients will remain compliant with follow-up exams for years after their treatment, which is difficult to ensure. Also, it will assume that data can be shared without violating patient privacy or commercial trade secrets.
• Failure: If data are incomplete, safety signals may go unnoticed until a large number of patients have been treated.
• Success Consequence: Detecting a rare, delayed side effect could lead to a reactionary suspension of the entire class of therapies, potentially denying treatment to ill patients who would have accepted the risk.

4. Policy Scoring Matrix

title: “Week 2: Biology & DNA” weight: 20 description: “Analysis of polymerase error rates, oligo synthesis, and essential amino acids.”

🧬 Questions Week 2

Question 1: Polymerase Error Rates

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?

Answer: Polymerase error rate is approximately 1 error per 10^6 bases, alongside of 3′→5′ mismatch repair drives the final mutation rate down to roughly 10^{-8 to 10}-9 per bp per replication.

Human genome size: 3.2 Gbp for human. If fidelity were only 10^{-6, we would expect errors per genome copy = 3.2 × 10}9 × 10^-6 ≈ 3,200 errors.

Question 2: The Coding Space

How many 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?

Answer: Average human protein coding length ~ 1036 bp (~345 aa). Because of codon degeneracy, roughly (61/20)^{345 ≈ 10}16 possible synonymous coding sequences.

They will not work for:

• Sequence features (restriction sites, repeats)
• Problematic GC% secondary structure
• Synthesis/assembly failure models

Question 3: Synthesis Methods

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

Answer: Solid-phase phosphoramidite synthesis (cyclic detritylation / coupling / capping / oxidation).

Question 4: The 200nt Barrier

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

Answer: Each base addition is <100% efficient, so full-length yield drops exponentially with length. At 200 nt, even tiny per-step losses become huge.

More cycles also mean more side reactions + truncations, making purification of true full-length products increasingly painful.

Question 5: The Lysine Contingency

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

Answer: Common “10 essential” set used for many non-ruminant mammals (incl. pigs): Lys, Met, Trp, Thr, Val, Ile, Leu, Arg, His, Phe.

It’s a weak containment idea for animals because lysine is already essential, so removing lysine synthesis doesn’t uniquely constrain survival (since they already need to eat it to survive).