Engineering biology as a programmable system. From standardised parts and DNA synthesis to CRISPR genome editing and the first synthetic organisms — the field that proposes biology should be built, not just studied.
An engineering discipline applied to living systems — designing organisms or biological components to do specific tasks, with the same systematic approach electronics applies to circuits.
The field's central premise: biology can be made predictable through standardisation. Standardised DNA parts, characterised behaviour, hierarchical assembly, modelling-based design. The aspiration is that biology will become engineerable in the way silicon is.
The reality, in 2026, is partial. Some applications work routinely (insulin production in modified bacteria; CAR-T cancer therapy; CRISPR gene editing in plants). Others remain hard (designing complex genetic circuits that work reliably in living cells; predicting protein function from sequence). The field is somewhere between molecular biology and chemical engineering, leaning steadily toward the latter.
Recombinant DNA technology — splicing genes from one organism into another — was developed by Stanley Cohen (Stanford) and Herbert Boyer (UCSF) in 1973. Their experiments showed bacterial plasmids could carry foreign genes and replicate them.
The 1975 Asilomar Conference brought 140 molecular biologists together to debate biosafety. The voluntary moratorium that followed established the principle that scientists could organise to govern their own field's risks. The conference's containment guidelines became the basis of subsequent regulation.
Genentech (founded 1976 by Boyer and venture capitalist Robert Swanson) commercialised recombinant DNA. The first product: human insulin produced in E. coli (1982 FDA approval, marketed as Humulin).
The term "synthetic biology" was used occasionally in the 20th century but the contemporary field dates from the early 2000s. Three foundational projects:
Tom Knight at MIT proposed the BioBrick standard (2003) — a standardised format for DNA parts that snap together with consistent restriction sites. Standardisation was the central engineering insight: biology had not previously had standard interfaces.
Drew Endy co-founded the Registry of Standard Biological Parts (2003) at MIT — a public catalogue of characterised DNA components.
Jay Keasling at Berkeley demonstrated metabolic engineering at scale: producing the antimalarial precursor artemisinic acid in engineered yeast (2006). The Bill & Melinda Gates Foundation funded the project; semi-synthetic artemisinin entered production in 2013.
The International Genetically Engineered Machine competition began at MIT in 2003 with five teams. By 2024 it had grown to over 400 teams from 50+ countries. Undergraduate teams design, build, and test genetically engineered systems and present at an annual jamboree (now held in Paris, having moved from Boston).
iGEM's structural innovation: it forced standardisation. Teams use BioBricks (now SynBio Hub) parts and contribute new ones back to the registry. The competition trained the first generation of synthetic biologists and substantially expanded the field's part library.
Notable iGEM-launched companies: Ginkgo Bioworks (founded 2008 by MIT iGEM alumni), Synlogic, Asimov. The field's industrial pipeline runs in part through iGEM alumni networks.
Two technical curves define the field's economics. DNA reading (sequencing): Sanger sequencing in 1977; cost per genome dropped from $3 billion (Human Genome Project, 2003) to ~$200 (2024). Faster than Moore's law for two decades.
DNA writing (synthesis): synthesising arbitrary DNA sequences. The cost has dropped from ~$10/base in 1990 to under $0.05/base in 2025, mostly through column-based phosphoramidite synthesis (Caruthers, 1981) and the inkjet-printed array methods of Twist Bioscience and Agilent.
The synthesis cost matters because designing biology requires writing DNA. A 5,000-base-pair plasmid — a typical project — costs $250 in 2025; the same project cost $50,000 in 2005.
The current frontier is enzymatic synthesis (DNA Script, Ansa Biotechnologies, Molecular Assemblies) — using DNA polymerases instead of phosphoramidite chemistry. Promises lower error rates and longer products.
The 2012 paper by Jennifer Doudna and Emmanuelle Charpentier (with Martin Jinek as first author) showed that the bacterial CRISPR-Cas9 system could be programmed to cut any DNA sequence using a guide RNA. The technology made genome editing routine.
Feng Zhang at the Broad Institute demonstrated CRISPR editing in mammalian cells in early 2013. The patent priority dispute between Berkeley (Doudna) and the Broad ran through 2022.
Doudna and Charpentier won the 2020 Nobel in Chemistry — the first all-female chemistry Nobel.
Targeted gene knockout and replacement at scale. CRISPR-edited crops (Calyxt's high-oleic soybean, 2019). Casgevy (2023) — the first FDA-approved CRISPR therapy, for sickle cell disease. The first CRISPR-edited babies (He Jiankui, 2018, China — universally condemned; the researcher imprisoned).
The 2020s have produced CRISPR variants — base editing (Liu lab), prime editing (Liu lab again), CRISPR interference and activation (CRISPRi, CRISPRa). The toolkit is more precise than the original.
The most ambitious synthetic-biology projects build genomes from scratch.
JCVI-syn1.0 (Craig Venter Institute, 2010) — the first organism with a fully synthetic genome. Researchers chemically synthesised the 1.08-million-base-pair genome of Mycoplasma mycoides, transplanted it into a recipient cell, and showed the cell adopted the synthetic genome's properties.
JCVI-syn3.0 (2016) — minimal genome. Reduced to 473 essential genes (901,000 bp). About 30% of the genes had unknown function — a striking reminder that even minimal life retains substantial mystery.
Synthetic Yeast 2.0 (Sc2.0) consortium — synthesising the entire yeast genome chromosome by chromosome, with edits. The first complete synthetic yeast was reported in 2023.
The implication: the genome is increasingly a designed object. The remaining limit is interpretive, not synthetic — we can build genomes faster than we can design them.
The most commercially mature application of synthetic biology. Metabolic engineering reroutes a microbe's biochemistry to produce a desired molecule.
The pioneering case: Jay Keasling's artemisinic acid pathway in yeast (2006) — a precursor to the antimalarial drug. Sanofi commercialised it; semi-synthetic artemisinin entered production in 2013, providing a stable supply for malaria treatment.
Subsequent products: Impossible Foods's heme protein (soy leghemoglobin produced in engineered yeast) — the molecule that makes the Impossible Burger taste of meat. Perfect Day's animal-free dairy proteins. Geno (formerly Genomatica)'s 1,4-butanediol from sugar. Amyris's farnesene-based hydrocarbons.
The economics: bio-produced molecules compete with petrochemical ones on cost only at niche commodity prices, but win at high-value molecules where chemistry is hard. The 2020s commercial frontier is high-value pharmaceuticals, food ingredients, and specialty chemicals.
Cell-free protein synthesis (CFPS) extracts the protein-making machinery from cells and runs it in a tube. Originally a research tool (Nirenberg's 1961 work that cracked the genetic code used cell-free extracts). Increasingly a manufacturing platform.
Sutro Biopharma uses cell-free synthesis to produce antibody-drug conjugates. Greenlight Biosciences (now part of Fall Line Capital) developed cell-free RNA production for vaccines and crop protection.
The Pardee lab at Toronto demonstrated paper-based cell-free diagnostics (2014) — freeze-dried biological circuits that activate when rehydrated, capable of detecting Zika or Ebola without lab infrastructure. The technology proved the concept that synthetic biology could reach low-resource settings.
Cell-free systems also underpin most modern directed evolution workflows for protein engineering — the toolkit that produced the Nobel-winning work of Frances Arnold.
Until the 2010s, protein engineering meant editing existing proteins or evolving them under selection pressure. Designing entirely novel proteins from amino-acid sequence was an open problem.
The field changed in the 2020s. David Baker's lab at the University of Washington's Institute for Protein Design produced increasingly sophisticated computationally-designed proteins from 2003 onward. The combination of Rosetta (Baker's structure-prediction software), AlphaFold (DeepMind, 2020 — accurate structure prediction from sequence; 2024 Nobel in Chemistry), and diffusion-model-based design tools (RFdiffusion, ProteinMPNN, 2023) has made protein design routine.
Applications: novel binders to disease targets, new enzymes, vaccine immunogens, programmable biomaterials. The 2024 Nobel was shared by Baker (for protein design) and DeepMind's Demis Hassabis and John Jumper (for AlphaFold).
The pace is striking. In 2020, designing a novel high-affinity binder took years. By 2024, similar projects were taking weeks.
Messenger RNA — the molecule that delivers genetic instructions to ribosomes — became a therapeutic platform during the COVID-19 pandemic. The Pfizer-BioNTech and Moderna mRNA vaccines were the first widely-deployed mRNA medicines.
The foundational work was done by Katalin Karikó and Drew Weissman (University of Pennsylvania, 2005) — they demonstrated that nucleoside-modified mRNA could enter cells without triggering destructive immune responses. They won the 2023 Nobel in Physiology or Medicine.
The platform's appeal: rapid development (weeks rather than years for new vaccines), reusable infrastructure, and the cell does the protein production. Active development for cancer vaccines (Moderna, BioNTech), CRISPR delivery, gene therapy, and rare-disease enzyme replacement.
The synthetic-biology connection: mRNA medicines depend on engineered untranslated regions, codon-optimised coding sequences, and synthetic lipid nanoparticles for delivery — every component is a designed object.
Gene therapy — replacing or modifying genes to treat disease — had a difficult early history. The 1999 death of Jesse Gelsinger in a University of Pennsylvania trial set the field back a decade.
Recovery and acceleration came through better vectors. Adeno-associated virus (AAV) as a delivery vehicle has enabled approved therapies for spinal muscular atrophy (Zolgensma, 2019), retinal disease (Luxturna, 2017), and haemophilia.
The 2023 approval of Casgevy (Vertex/CRISPR Therapeutics) for sickle cell disease was the first commercial CRISPR-based gene therapy. Lyfgenia (bluebird bio) followed for the same indication. Pricing — $2.2 million per patient for Casgevy — has been the principal access barrier.
The 2020s pipeline includes treatments for hereditary blindness, muscular dystrophies, cystic fibrosis, and various rare metabolic disorders. The current technical challenge: in vivo delivery of CRISPR to specific tissues.
CAR-T cell therapy: take a cancer patient's own T cells, engineer them to express a chimeric antigen receptor (CAR) targeting tumour antigens, and infuse them back. The reprogrammed T cells hunt and kill the cancer.
The 2017 FDA approval of Kymriah (Novartis, for B-cell acute lymphoblastic leukaemia) was the first commercial CAR-T therapy. Yescarta, Tecartus, Breyanzi, Abecma, and Carvykti followed for various blood cancers.
CAR-T pioneers: Carl June (UPenn) — the early clinical work. Steven Rosenberg (NIH) — the foundational immunotherapy research. The first CAR-T patient, Emily Whitehead, was treated in 2012 at age 6 and remains in remission as of 2025.
The 2020s pipeline: solid tumours (much harder than blood cancers), allogeneic CAR-T (off-the-shelf instead of autologous), engineered NK cells, in vivo CAR generation. The field is reshaping cancer immunotherapy more than any other current platform.
The original synthetic-biology dream: build genetic circuits the way engineers build electronic ones. Two foundational papers, both 2000: Gardner, Cantor, Collins's genetic toggle switch in E. coli, and Elowitz, Leibler's repressilator (an oscillator made from three repressor genes in a feedback loop).
Subsequent work: cell-cell communication circuits (Weiss, Basu); RNA-based logic gates (Smolke); CRISPR-based circuits (Voigt). The 2010s field at MIT, Harvard, and Boston University extended the toolkit substantially.
The realised difficulty: cellular environments are noisy, heterogeneous, and resource-limited in ways electronic substrates aren't. A circuit that works in E. coli may fail in mammalian cells. Resource competition between circuits and host metabolism causes unexpected coupling.
Active 2020s research: cell-state-aware circuits (sensing host conditions and adjusting), distributed circuits across populations of cells, modular circuit-synthesis pipelines (Asimov's Kernel platform).
Engineered cells can detect specific molecules or conditions and report through fluorescence, colour change, or downstream gene expression. Applications:
Heavy-metal detection in groundwater (arsenic biosensors developed for use in Bangladesh). Inflammation detection in the gut (Synlogic's engineered probiotics). Cancer detection through circulating tumour DNA (multiple academic groups).
The 2014 Pardee lab paper-based diagnostics extended the concept: freeze-dried cell-free systems on paper that activate on rehydration. The technology demonstrated that synthetic biology could produce diagnostic infrastructure for low-resource settings.
Wearable biosensors are the 2020s frontier. The Fitbit Sense measures stress through skin electrochemical activity; experimental devices measure glucose, hydration, and biomarkers continuously. Most are not yet "synthetic biology" in the strict sense but use biological recognition elements.
Crop genetic modification predates contemporary synthetic biology — Bt corn (1996) and Roundup Ready soybeans (1996) used recombinant DNA techniques. The synthetic-biology era has expanded the toolkit and ambition.
CRISPR-edited crops: Calyxt's high-oleic soybean (2019, US-approved without GMO labelling), the GABA-elevated tomato (Sanatech Seed, Japan, 2021), browning-resistant Arctic apple, non-browning Innate potato.
The frontier: nitrogen-fixing crops (engineering cereal grains to fix their own nitrogen, eliminating fertiliser dependency — a longstanding goal, still unachieved); C4 rice (C4 photosynthesis is more efficient; rice naturally uses C3 — the C4 Rice Project has been working on the conversion since 2008); climate-resilient crops (drought, heat, salinity tolerance).
The regulatory landscape is fragmented. CRISPR-edited crops without inserted foreign DNA are unregulated in the US, Japan, Brazil; regulated as GMOs in the EU.
The transition from lab to industrial bioproduction is the field's hardest commercial step. Engineering a microbe to produce a molecule in a flask is one project; producing it economically at 10,000-litre scale is another.
Major US biomanufacturing companies: Ginkgo Bioworks (founded 2008, "platform for designing organisms"), Amyris (founded 2003, fragrance and beauty ingredients), Zymergen (founded 2013, bankrupt 2023 — case study in difficulty), Genomatica.
The economics are challenging. Biomanufacturing competes with petrochemicals; oil at $80/barrel sets a hard price floor. Most viable bio-products are high-value molecules or have specific functional advantages over chemical alternatives.
The 2022 US Bioeconomy Executive Order and the European Union's 2024 biomanufacturing investments suggest national-strategic interest. The geopolitical motive: reduce dependency on petrochemical supply chains; develop capability to produce critical medicines domestically.
The same techniques that enable beneficial synthetic biology can enable engineered pathogens. The 2002 reconstruction of poliovirus from synthetic DNA (Cello, Paul, Wimmer at Stony Brook) demonstrated that arbitrary viruses could be synthesised from sequence data alone. The 2005 reconstruction of the 1918 Spanish flu virus (Tumpey et al., CDC) extended the concern.
The infrastructure: DNA synthesis providers (Twist, IDT, GenScript) screen orders against the IGSC harmonized screening protocol — flagging requests that match dangerous pathogen sequences. Compliance is voluntary but widely adopted. The 2024 introduction of mandatory screening in the US (via Office of Science and Technology Policy guidance) is the first regulatory teeth.
Dual-use research of concern (DURC) policies require institutional review of research that might enhance pathogen transmissibility or virulence. The 2014–2017 US gain-of-function research moratorium followed several high-profile experiments.
The 2020s biosecurity discussion increasingly includes AI-enabled biothreat modelling — the concern that AI tools could accelerate dangerous protein/pathogen design.
In November 2018, He Jiankui (Chinese researcher at Southern University of Science and Technology) announced that twin girls Lulu and Nana had been born after CRISPR editing of their CCR5 genes, intended to confer HIV resistance. The announcement was made at the Second International Summit on Human Genome Editing in Hong Kong.
The condemnation was global and immediate. The scientific community broadly considered the work premature, ethically unjustified, and methodologically flawed (the actual edits did not produce the intended CCR5-Δ32 allele). He was sentenced to three years in prison by Chinese courts in 2019; released in 2022.
The case crystallised the prohibition on germline genome editing in humans. The 2020 WHO advisory committee, the 2023 International Bioethics Committee, and most national academies have endorsed the moratorium. Whether the line will hold against medical pressures (preventing inherited diseases) is uncertain.
The somatic-vs-germline distinction is the field's clearest ethical line. Editing somatic cells (the patient's body cells, not passed to offspring) is broadly accepted. Germline editing (changes that propagate to descendants) remains forbidden in human clinical practice in 2026.
The 2020s synthetic biology industry consists of several major segments.
Tools and instruments: Twist Bioscience and IDT (DNA synthesis), Illumina (sequencing), 10x Genomics (single-cell genomics), DNA Script (enzymatic DNA synthesis).
Therapeutics platforms: Vertex, CRISPR Therapeutics, Editas Medicine, Beam Therapeutics, Prime Medicine. Most large pharma has CRISPR programs.
Industrial biotech: Ginkgo Bioworks (~$300M revenue in 2024), Amyris (chapter 11 in 2023, restructured), Genomatica, Solugen.
Food and ingredients: Impossible Foods, Perfect Day, EVERY Co (animal-free egg protein), Geltor (protein design), Beyond Meat (plant-based, not strictly synthetic biology).
AgBio: Pivot Bio (nitrogen-fixing microbes for corn), Pairwise (CRISPR-edited produce), Inari Agriculture.
The funding environment shifted dramatically in 2022–2023 — the SPAC-era valuations collapsed (Ginkgo's stock fell 95% from 2021 peak; Zymergen bankrupted). The post-2024 recovery has been uneven; the field has had to demonstrate revenue-generating product fit.
Industrial-scale automated facilities for designing, building, and testing biological systems. The DARPA-funded 1000 Molecules programme (2019–) supported foundry development; Ginkgo Bioworks's Boston facility is the largest commercial example, with millions of square feet of automated lab capacity.
Academic foundries: The Edinburgh Genome Foundry, The London DNA Foundry, The Concordia Genome Foundry, The Singapore Bio-Foundry. The Global Biofoundry Alliance (founded 2019) coordinates standards and protocols across 35+ member institutions.
The promise: automation reduces the cost and time of build-test cycles by 100×; what currently takes a graduate student 6 months can be completed in days. The reality: automation is most useful for the standardisable middle of the pipeline (cloning, transformation, screening); the design and analysis ends still require human expertise.
Foundries also produce data at scales that enable machine-learning approaches to organism design. The 2024 Ginkgo + Google partnership and similar arrangements signal the convergence of bio-foundries with AI infrastructure.
The intersection of AI and synthetic biology is the field's fastest-moving frontier in 2026. Three major threads.
Protein and structure prediction: AlphaFold2 (2020), AlphaFold3 (2024 — extended to ligand and nucleic-acid binding), ESMFold (Meta), RoseTTAFold (Baker lab). Prediction quality is now sufficient for most design tasks.
Generative protein design: RFdiffusion (2023), ProteinMPNN, ESM-3, Chai-1. Generate novel protein sequences with specified functions or binding properties. The 2024 launch of Chroma (Generate Biomedicines) and similar commercial offerings has commoditised the technology.
Biological foundation models: Evolutionary Scale's ESM, Genentech and partners' DNA language models, single-cell foundation models. The aspiration is GPT-style models for biology — trained on enormous biological datasets, then prompted for specific tasks.
The field has not yet had its "ChatGPT moment" — biology lacks the same kind of clean unified text corpus and the experimental feedback loops are slower. But the rate of capability improvement is striking and shows no sign of plateauing.
Synthetic biology offers several potential climate applications.
Engineered photosynthesis: more efficient CO₂ fixation in crops or photosynthetic microbes. The C4 Rice Project (since 2008) attempts to convert rice from C3 to C4 photosynthesis. RIPE (Realizing Increased Photosynthetic Efficiency, 2018+) reports yield improvements of 15–20% in field trials.
Microbial carbon capture: engineered cyanobacteria and algae for CO₂ uptake and conversion to high-value products. Scale remains the bottleneck — competitive with chemical direct-air capture only at very specific conditions.
Methane reduction: DSM-Firmenich's Bovaer (a small-molecule, not strictly synthetic biology, but adjacent) reduces cattle methane emissions by 30%. Multiple companies are developing engineered rumen microbiomes.
Replacement of carbon-intensive products: bio-produced versions of petroleum-derived chemicals, animal-product replacements that reduce land use. The aggregate climate impact depends on scale of adoption.
Five things synthetic biology has not yet solved.
1. Predicting circuit behaviour from design. Genetic circuits in cells still behave in unexpected ways more often than electronic circuits do. Resource competition, host stress responses, and evolution within populations all interfere.
2. Scaling from microbe to higher organism. Most synthetic biology works in bacteria and yeast. Mammalian cell engineering is harder; whole-organism engineering harder still.
3. Long-term stability. Engineered organisms tend to lose their introduced functions over many generations as evolution selects against the metabolic burden. Maintaining function in production strains is an ongoing challenge.
4. Designing complex multi-cellular behaviour. Tissue and organ design — for organoids, transplantation, biocomputing — remain at the research frontier.
5. The protein-design-to-function bridge. Designing a protein with a desired structure is largely solved; designing one with a desired catalytic function is much harder and still mostly empirical.
↑ Jennifer Doudna · How CRISPR lets us edit our DNA · TED
Watch · George Church · genomics, biotech, synthetic biology
Watch · iGEM · The heart of synthetic biology
Online courses. MIT OCW 20.020 "Introduction to Biological Engineering Design" (free). Coursera's "Molecular Biology" specialisation (Eric Lander). edX's "Synthetic Biology" by MIT.
Research groups to follow. Doudna lab (UC Berkeley). Church lab (Harvard). Voigt lab (MIT). Keasling lab (Berkeley/JBEI). Liu lab (Broad). Endy lab (Stanford).
Conferences. SynBioBeta (annual, San Francisco — industry-focused). Synthetic Biology: Engineering, Evolution & Design (SEED, academic). The CRISPR Conference. iGEM Grand Jamboree (annual, Paris).
Magazines. Genetic Engineering & Biotechnology News (GEN). Endpoints News for biotech business. Nature Biotechnology, ACS Synthetic Biology for primary research.
Communities. The Engineering Biology Research Consortium (EBRC) for US academic community. iGEM alumni networks. The biosecurity-focused group at the Johns Hopkins Center for Health Security.
Synthetic biology in 2026 sits at an inflection. The technical enablers — DNA synthesis cost, CRISPR, AlphaFold, computational design — have come together. The commercial validation — Casgevy, mRNA vaccines, Impossible Foods — has happened. The economic returns to biological-engineering capability are now substantial.
What's still missing: the analogue of the integrated-circuit revolution that made electronics ubiquitous. Biology remains craft-intensive, project-specific, and slow compared to electronics. The aspiration that biology will become as predictable as silicon is partly true and partly perpetually receding.
The next decade will determine whether the field continues to deliver high-value but specialised products (the current pattern) or breaks through to mass-scale infrastructure (engineered crops at planetary scale, gene therapies as routine medicine, biomanufacturing as commodity). Both futures are plausible. Both involve rewriting the relationship between humans and biology in ways the discipline has not yet fully reckoned with.
Synthetic Biology — Volume X, Deck 11 of The Deck Catalog. Set in Inter and Tiempos Text. Lab-dark #0a1416; lime, cyan, and amber accents. Monospace metadata.
Twenty-eight leaves on the engineering of biology. The field is younger than the readers it has.
↑ Vol. X · Future · Deck 11