Week 1: Principles & Practices

About Me

My name is Peter Olawumi, and I’m based in Ibadan, Nigeria. As a software developer with the handle @dev_roc, I’m passionate about bridging technology and biology to create innovative, accessible solutions for real-world problems, especially in the Global South. Joining HTGAA is an exciting opportunity to explore synthetic biology and apply it to challenges like waste management in our growing industrial sectors.

Proposed Biological Engineering Application or Tool

I propose developing microbial “Plastic Eater” pods for on-site industrial recycling. These are compact, factory-floor bioreactors using engineered bacteria to break down PET plastic waste into reusable monomers.

Why this? In bustling manufacturing plants in Lagos and Ibadan, discarded PET bottles and packaging pile up daily, leading to costly hauling, environmental pollution, and health risks from microplastics. Traditional recycling is energy-intensive and inefficient, with global rates at just 18%. In Nigeria, informal recycling dominates but lags in efficiency. My tool would be a lunchbox-sized pod that processes 500g-1kg of PET scraps per cycle at ambient temperatures, yielding 80-90% monomer recovery (terephthalic acid and ethylene glycol) for repolymerization or new chemicals. It’s low-energy, scalable, and deployable without shipping, inspired by natural degraders like Ideonella sakaiensis, supercharged with synthetic biology for faster action.

The core: Engineer Ideonella sakaiensis or a surrogate like Pseudomonas putida with optimized PETase and MHETase enzymes, fused to secretion signals and reporters for efficiency. This could cut waste transport emissions by 40%, create bio-recycling jobs, and align with UN SDG 12 for sustainable consumption.

Governance/Policy Goals

To ensure this tool contributes to an ethical future, I focus on non-malfeasance (preventing harm). I’ve adapted the synthetic genomics framework for safety/security and equity.

Goal 1: Biosafety Lockdown – Prevent Unintended Microbial Escapes and Toxicity
This goal contains recombinant strains to avoid ecological disruptions, like outcompeting native microbes or leaching toxins in biodiverse areas like Lagos lagoons.

• Sub-goal 1a: Engineered Containment Mechanisms – Integrate two orthogonal kill switches (e.g., mazEF toxin-antitoxin and light-inducible CRISPRi) in plasmids. Validate with in vitro escape assays (>99.99% die-off in 48 hours via qPCR).
• Sub-goal 1b: Risk-Stratified Environmental Release Testing – Implement tiered trials: lab (BSL-1), semi-contained (HEPA-barriered pods), pilot (metagenomic-monitored sites). Track HGT risks (<0.1% plasmid mobilization via 16S rRNA sequencing).
• Sub-goal 1c: Toxicity Profiling for Byproducts and Enzymes – Conduct assays on outputs (Ames test for genotoxicity <2x induction; yeast screen for endocrine disruption EC50 >100μM). Cap enzyme secretion to avoid risks.

Goal 2: Equitable Deployment – Ensure Broad Access Without Widening Industrial Divides
This prevents social harms like job displacement, promoting inclusive scaling inspired by the African Union’s biotech equity charter.

• Sub-goal 2a: Open-Source IP and Tech Transfer – Classify designs as Creative Commons (CC-BY-SA) for non-commercial use in developing economies. Host on iGEM registry with modular parts for local adaptations.
• Sub-goal 2b: Socio-Economic Impact Audits – Use agent-based modeling (NetLogo) to forecast job shifts (e.g., aim for Gini coefficient drop <0.1). Include community “right-to-reject” via town halls (>60% approval).
• Sub-goal 2c: Adaptive Monitoring for Long-Term Equity – Integrate IoT sensors into pods for blockchain-ledger yield tracking (70% monomer value back to operators). Cap market share (<30%) to avoid over-reliance.

Governance Actions

I’ve outlined three actions: a regulatory rule, an incentive program, and a technical strategy, involving different actors. Analogies draw from drones (certification), finance (buffers), and 3D printing (open designs).

Action 1: Mandatory Pre-Deployment “Escape-Proof” Certification (Regulatory Rule by Federal Agencies)
Analogy: FAA drone certification for safe airspace.

• Purpose: Current Nigerian biosafety (NBMA 2015 Act) is ad-hoc, risking spills. Propose standardized “synbio passport” with <0.01% escape risk proven via simulations, shifting to proactive approvals.
• Design: Amend Biosafety Regulations (2020) for dossiers (COPASI models, assays, audits). Actors: NBMA approves (6-month review); companies fund (₦500k-1M, offset by permits); academics validate. Use open API for data.
• Assumptions: Regulators have capacity (50+ assessors); models translate to real-world (e.g., floods); industry complies without loopholes.
• Risks of Failure & “Success”: Failure: Rigid certs stifle startups (80% rejection); corruption erodes trust. Success: Widespread adoption breeds complacency against evolving threats (like financial stress tests missing crises).

Action 2: “Green Pod” Subsidy Incentives with Equity Audits (Incentive Program by Industry-Academia Consortia)
Analogy: Basel III capital buffers for financial resilience.

• Purpose: Factories prioritize profits over equity; propose 40% tax credits for adopters passing audits (30% revenue shared with informal sectors), shifting to impact investing.
• Design: Co-designed by MAN/universities, funded by 1% levy (₦10B pot). Actors: Companies self-audit (NetLogo); consortia approve; NGOs monitor. Use blockchain for payouts; train 1k workers/year.
• Assumptions: Big firms lead (70% pilot adoption); audits capture nuances; economic stability holds.
• Risks of Failure & “Success”: Failure: Low opt-in (<20%); greenwashing erodes trust. Success: GDP boost (+5% recycling) floods markets, spurring overproduction (like drone supply chain jams).

Action 3: Open-Source “Watchdog” Microbial Sentinel Network (Technical Strategy by Academic Researchers)
Analogy: Thingiverse for 3D printing with safety mods.

• Purpose: Fragmented tracking leaves surveillance gaps; propose free platform with sentinel kits (qPCR for HGT) for crowdsourced monitoring, shifting to community-driven oversight.
• Design: Led by UNILAG/iGEM Africa with $500k grants. Actors: Researchers upload (CC-BY); factories deploy ($50/unit); NBMA integrates. Use Raspberry Pi/ML for alerts; beta in HTGAA, then 100-node pilot.
• Assumptions: Open-source thrives (1k contributors); low-tech adoption; data privacy holds.
• Risks of Failure & “Success”: Failure: Sparse coverage (<10%); false positives desensitize. Success: Panopticon erodes privacy (worker data misuse), amplifying biases (like financial algos).

Scoring Actions Against Goals

Using an adapted rubric (1 = best/strong positive, 3 = weak/neutral, n/a = not applicable):

Explanation: Action 1 excels in prevention but burdens innovation (higher costs). Action 2 boosts equity and feasibility via incentives but weaker on direct security. Action 3 is feasible and responsive but risks privacy issues.

Prioritization and Trade-offs

I prioritize a combination of Action 2 (incentives) and Action 3 (sentinel network), starting with academics and industry consortia, targeted at national audiences like Nigeria’s Ministry of Science & Technology and international like the African Union. Why? This balances proactive equity (Action 2’s audits prevent divides) with responsive monitoring (Action 3’s crowdsourcing flags harms early), scoring well on feasibility and constructive uses without heavy regulation that could slow adoption in resource-limited settings.

Trade-offs: Incentives may increase short-term costs (levy) but yield long-term savings (20% waste reduction); open-source risks IP theft but promotes access. Assumptions: Strong community buy-in (e.g., 70% SME uptake); uncertainties include enforcement in informal sectors and tech literacy. If unaddressed, fall back to Action 1 for high-risk deployments.

Reflection on Class Learnings

From lectures by David Kong, George Church, and Joe Jacobson, I learned about biotech’s rapid evolution and ethical imperatives like biosecurity and equity. A new concern for me: In the Global South, unequal access could exacerbate divides—e.g., advanced tools benefiting only elites. Another: Dual-use risks, where degraders might be misused for harmful polymers.

To address: Propose mandatory equity clauses in grants (e.g., 20% project budget for community training) and international standards for dual-use reviews (adapt WHO guidelines). This ties to my project, emphasizing open designs with built-in safeties.

Lecture 2 Preparation – Homework Answers

For Professor Jacobson Lecture

Error Rate of Polymerase

The error rate of nature’s DNA polymerase (specifically, error-correcting polymerase in biological synthesis) is approximately 1 error per 10⁹ (1 billion) base pairs added.

The human genome is roughly 3 × 10⁹ (3 billion) base pairs long. This means that, on average, DNA replication of the entire human genome would introduce about 3 errors per replication cycle if relying solely on this error rate.

Biology addresses this discrepancy through multiple layers of error correction and repair mechanisms beyond the base polymerase error rate. These include:

• Built-in proofreading via 3’–5’ exonuclease activity in the polymerase itself, which immediately detects and corrects mismatches during synthesis.
• Post-replication mismatch repair systems that scan for and fix errors shortly after replication.
• Additional DNA repair pathways (e.g., base excision repair, nucleotide excision repair, and double-strand break repair) that operate continuously to detect and correct damage from replication errors, environmental factors, or spontaneous mutations.

These combined mechanisms can reduce the effective mutation rate to as low as 10⁻¹⁰ per base pair in vivo, ensuring genome stability across cell divisions.

Number of Ways to Code for an Average Human Protein

An average human protein is encoded by approximately 1036 base pairs of DNA, corresponding to about 345 amino acids (since each amino acid is coded by a 3-base codon, or triplet).

The genetic code uses 64 possible codons (4³) to specify 20 amino acids and 3 stop signals. Excluding stop codons, there are 61 codons for the 20 amino acids, yielding an average degeneracy of about 3.05 codons per amino acid.

For a specific protein sequence of 345 amino acids, the total number of different DNA nucleotide sequences (coding sequences) that could translate to the exact same amino acid sequence is enormous — on the order of 3.05³⁴⁵ ≈ 10¹⁶⁷.

In practice, not all of these theoretically possible coding sequences work effectively to produce the protein of interest (especially in the context of gene synthesis and expression). Important limiting factors include:

• Codon usage bias — different organisms prefer certain synonymous codons due to tRNA abundance
• mRNA secondary structure and stability (hairpins, degradation signals)
• GC content and unwanted sequence motifs (restriction sites, splice sites, repeats)
• Synthesis errors — chemical DNA synthesis has higher error rates (~1:10² per base)
• Regulatory constraints (e.g., in recoded organisms with codon reassignment)
• Functional impacts of synonymous changes on folding, translation kinetics, and expression levels

For these reasons, synthetic genes are usually designed with a subset of “optimal” codons rather than exploring the full theoretical space.

For Dr. LeProust Lecture

Most Commonly Used Method for Oligo Synthesis Currently

The most commonly used method for oligonucleotide (oligo) synthesis is solid-phase phosphoramidite chemistry.

This involves a cyclic process on a solid support (controlled pore glass or silicon-based chips, as used by Twist Bioscience):

1. Coupling — DMT-protected phosphoramidite monomer is added to the growing chain
2. Capping — Unreacted sites are capped to prevent further extension
3. Oxidation — Phosphite linkage is oxidized to a stable phosphate
4. Deblocking — DMT group is removed to allow the next coupling

This method, developed in the early 1980s, remains the industry standard for automated, high-throughput oligo synthesis.

Why It Is Difficult to Make Oligos Longer Than 200 nt Via Direct Synthesis

Direct chemical synthesis of oligos longer than ~200 nucleotides is challenging primarily due to the limitations of coupling efficiency in phosphoramidite chemistry (typically 98–99% per step).

For a 200 nt oligo, theoretical yield of full-length product is approximately (0.99)¹⁹⁹ ≈ 13%, but in practice it is significantly lower due to accumulating side reactions such as:

• Depurination (acid-induced base loss)
• Incomplete deprotection
• Branching and other side products

These issues cause exponential yield drop and increasing error accumulation (deletions, insertions, substitutions), making purification of full-length, error-free products very difficult beyond ~200 nt.

While advanced platforms (e.g. Twist Bioscience) have improved chemistry to routinely reach ~350 nt and demonstrated ~700 nt experimentally (with ~97% full-length material), these are not standard for direct synthesis beyond 200 nt.

Why You Can’t Make a 2000 bp Gene Via Direct Oligo Synthesis

A 2000 base pair gene cannot be made via direct oligo synthesis because current chemical methods are fundamentally limited in length (routine max ~350 nt, experimental ~700 nt).

Attempting 2000 bp directly would result in near-zero yield due to:

• Extremely low coupling efficiency over thousands of steps → theoretical yield (0.99)¹⁹⁹⁹ ≈ 10⁻⁹ (practically nonexistent)
• Massive accumulation of chemical errors (depurination, oxidation byproducts, etc.)
• Impractical purification at that scale

Instead, genes of this length are constructed by assembling multiple shorter oligos (typically 50–300 nt) using enzymatic methods such as:

• Gibson assembly
• Enzymatic assembly platforms (e.g. Twist HELIX2)
• Followed by cloning, error correction, and verification via long-read sequencing

This modular approach overcomes the direct synthesis length barrier.

For George Church Lecture

Suggested Code for AA:AA Interactions

For AA:AA (amino acid–amino acid) interactions in proteins — which enable folding, oligomerization, and interfaces (analogous to NA:NA basepairing or AA:NA ribosomal translation) — I suggest a Side Chain Complementarity Code based on physicochemical properties of amino acid side chains.

This probabilistic code categorizes preferred pairings:

• Hydrophobic–Hydrophobic — van der Waals forces (e.g. Leu ↔ Ile, Val ↔ Phe) → core stabilization, coiled-coils, β-sheets
• Charged opposites — electrostatic attraction / salt bridges (e.g. Lys/Arg ↔ Asp/Glu)
• Polar–Polar — hydrogen bonding between uncharged polar groups (e.g. Ser/Thr ↔ Asn/Gln) → surface interactions
• Aromatic stacking — π–π interactions (e.g. Phe ↔ Tyr/Trp) → ring stabilization
• Special / covalent — disulfide bonds (Cys ↔ Cys), metal coordination (e.g. His ↔ His via Zn²⁺)

This framework aligns with natural protein interaction rules and could be extended for synthetic biology applications, e.g. incorporating non-standard amino acids to create novel interaction pairs.