BX LAB TECHNICAL ARTICLE
Enzymatic DNA Synthesis: What It Threatens, and What It Does Not
The papers open by naming two real limits of phosphoramidite chemistry. Then look at what the enzymes actually made.
Key takeaways
Phosphoramidite chemistry caps out around 200-mers and generates hazardous solvent waste — its critics are right about that.
TdT-based synthesis attacks both limits, and has produced 144 bits of stored data.
It writes DNA. A gapmer is not DNA, and TdT does not install a 2′-MOE or a phosphorothioate.
The criticism is fair, and it is in the first sentence
Palluk and colleagues open their 2018 Nature Biotechnology paper by stating that oligonucleotides are almost exclusively made by the phosphoramidite method, "even though it is limited to the direct synthesis of ~200 mers and produces hazardous waste". Both halves of that are true. Coupling efficiency compounds: at 99% per step a 200-mer leaves you about 13% full-length product, and the arithmetic does not care how good your reagents are. And solid-phase synthesis consumes acetonitrile, dichloroacetic acid and capping reagents in quantities that make a plant manager wince. Anyone selling phosphoramidites who pretends otherwise is not worth listening to on the rest of it.
What the enzyme approach actually does
Terminal deoxynucleotidyl transferase adds nucleotides to a DNA 3′ end without a template, which is the property the field is trying to exploit. The problem is stopping it after one addition. Palluk solved this by tethering: each TdT molecule carries a single dNTP, and once it incorporates that nucleotide the primer 3′ end stays covalently attached to the enzyme, physically blocking any other TdT-dNTP from reaching it. Cleave the tether and the primer is free for the next cycle. Lee and colleagues took a different route entirely — no terminator at all, just kinetic control, with apyrase degrading the substrate to limit how far a homopolymer run extends, and information encoded in the transitions between non-identical nucleotides rather than in individual bases.
Now look at the scale
Lee and colleagues used their method to store 144 bits. Not 144 bases — 144 bits, including addressing. That is a real demonstration of a real mechanism, and it is roughly eighteen bytes. Hoose and colleagues, reviewing the field in Nature Reviews Chemistry in 2023, put the commercial position plainly: synthesis of sequences beyond about 200 bp remains unaffordable, which is why they describe a gene writing gap in the first place. Enzymatic synthesis is a serious attempt to close that gap. It has not closed it yet.
The part that matters for therapeutics
Here is the distinction that usually gets lost. TdT writes DNA. It incorporates deoxyribonucleoside triphosphates and builds a natural phosphodiester backbone. A therapeutic antisense oligonucleotide is not that. It is a 16- to 20-mer gapmer with 2′-MOE or LNA wings, a phosphorothioate backbone at every linkage, often a GalNAc conjugate, and increasingly a defined stereochemistry at each phosphorus. There is no enzyme that installs a 2′-O-methoxyethyl group. There is no polymerase that gives you an Rp linkage where you want one and an Sp linkage where you want the other. Kandasamy and colleagues are putting phosphoryl guanidine linkages into stereopure oligonucleotides — that is a chemist deciding where each atom goes, and it is not a reaction an enzyme performs.
So what would have to change
The honest answer is: engineered polymerases with much broader substrate tolerance, plus a way to control backbone stereochemistry enzymatically. Neither is impossible and both are being worked on. If they arrive, the therapeutic oligonucleotide supply chain changes shape and people who make phosphoramidites should be paying attention rather than reassuring themselves. What is not true is the version you sometimes hear at conferences, where enzymatic synthesis is described as about to replace solid-phase chemistry generally. It is closing in on long natural DNA — gene synthesis, data storage, assembly fragments. That is a large and valuable market. It is a different market from the one that consumes 2′-F amidites.
What to take from it
If you are buying DNA for gene assembly or storage, watch this field closely; the economics may move under you. If you are making a modified therapeutic oligonucleotide, the near-term impact is that some of your natural-DNA inputs may get cheaper, and nothing else. And the two criticisms in Palluk opening sentence still stand against the chemistry we all use. The 200-mer ceiling is arithmetic. The waste is a process problem that the industry has largely chosen to live with. Neither is a reason to dismiss the phosphoramidite method — it is a reason to be honest about what it costs.
Turn the product name into a decision-ready specification
A purchasing name is only the starting point for a research material. Build the specification around exact chemical identity, form and assay basis. Define how identity will be confirmed and distinguish chromatographic purity from assay, concentration or active-content measurements. Add the attributes that could alter use, such as water, residual solvent, counterion, residual metals, known related substances or storage sensitivity. Acceptance limits should reflect the intended function rather than copying every available analytical result into a specification. Where a method is critical, record the technique, column or detection principle and sample preparation assumptions. This gives chemistry, analytical and procurement teams one shared definition of the material and reduces the chance that a technically correct but operationally unsuitable lot enters the project.
Qualify the material in the workflow that matters
A CoA can establish lot results against an agreed specification, but it cannot by itself prove performance in the intended research workflow. Plan a small, controlled qualification using a representative substrate, sequence, enzyme, reaction or analytical method. Include a suitable reference or previously accepted lot when available, keep preparation conditions comparable and predefine the observations that will trigger acceptance or investigation. Useful endpoints can include conversion, coupling efficiency, impurity formation, recovery, transcript integrity, conjugation yield or downstream signal, depending on the material. Record both the chemical result and practical behavior such as dissolution time, foaming, color change, precipitation or instrument delivery. This separates a material-quality question from a process-compatibility question and creates evidence that can support later lot-to-lot comparison.
Review process and scale risks before increasing quantity
Scale changes can expose risks that are not visible in a small research batch. For a research material, review handling, purification, analysis and scale-up before committing to a larger campaign. Estimate concentration, heat and mass-transfer needs, hold times, mixing limits, filtration area and the stability window of intermediates or prepared solutions. Identify the operations that control impurity rejection rather than assuming reaction yield alone will predict an acceptable isolated product. Raw-material availability and variability should be included in the route review, especially when a protecting-group reagent, specialty precursor or purification medium has a long lead time. A staged plan—feasibility, confirmation batch and then larger manufacture—provides checkpoints for analytical comparability, safety review and yield reconciliation while there is still room to adjust the process.
Build an analytical package that answers buyer questions
The useful documentation package should connect identity, purity and intended use. At minimum, align the specification and lot-specific Certificate of Analysis; then identify which supporting data are appropriate for the chemistry, such as HPLC or UPLC, LC-MS, high-resolution MS, NMR, water, residual solvents or inorganic-ion analysis. Chromatograms and spectra are most helpful when the material form, sample preparation and acceptance logic are clear. For complex or highly polar materials, a single percentage may not describe all relevant attributes, so note whether the result represents area purity, assay, concentration or another basis. Change control should cover shifts in route, critical starting material, purification approach, analytical method or manufacturing site when those changes could affect comparability. This framework helps technical, QA and procurement reviewers reach the same conclusion from the same evidence.
Use a structured request to shorten technical review
A complete inquiry for a research material should include the structure or unambiguous identifier, desired form, target quantity, expected future scale, intended use, target date and required documentation. Add known constraints: sensitive functional groups, prohibited solvents, metal limits, water sensitivity, shipping temperature, packaging preference or compatibility with an established method. If the request supports a regulatory submission, state the development stage and the documents expected, but do not assume that research-use or non-GMP material carries a qualification it does not have. Ask the supplier to identify open assumptions in the quotation and to separate confirmed specification items from values that still require method development. The resulting technical exchange is more efficient, and the final quotation can be tied to a material definition that is meaningful to scientists, procurement and quality reviewers.
Build a fit-for-purpose request
Before sourcing or scale-up, align the exact structure, intended workflow, target purity and assay basis, required form, storage conditions, quantity, timeline and documentation. If a catalog item does not match those requirements, a related analog, impurity standard, reference material or research-use custom route can be reviewed.
References
Technical context is supported by the peer-reviewed literature below.
- Palluk et al. (2018) De novo DNA synthesis using polymerase-nucleotide conjugates. Nature Biotechnology
- Lee et al. (2019) Terminator-free template-independent enzymatic DNA synthesis for digital information storage. Nature Communications
- Hoose et al. (2023) DNA synthesis technologies to close the gene writing gap. Nature Reviews Chemistry
- Kandasamy et al. (2022) Impact of guanidine-containing backbone linkages on stereopure antisense oligonucleotides in the CNS. Nucleic Acids Research