BX LAB TECHNICAL ARTICLE
Cyclic Dinucleotides: Why the Backbone Is Sulfurized
What the phosphorothioate does for a STING agonist, and what the first-in-class clinical trial actually reported.
Key takeaways
Natural cyclic dinucleotides are degraded by phosphodiesterases; the dithio backbone is the fix.
ADU-S100 reached a ~24-minute terminal plasma half-life in phase I, with no MTD found.
The (Rp,Rp) configuration is a synthesis requirement, not a label.
The natural molecule does not survive
cGAMP, c-di-AMP and c-di-GMP are the native ligands of the STING pathway, and as drugs they have an obvious problem: phosphodiesterases degrade them. The synthetic answer was to replace the bridging phosphate oxygens with sulfur, which is why the STING agonists that reached the clinic are phosphorothioate dinucleotides rather than the natural compounds. The sulfur is not decoration. It is the reason the molecule lasts long enough to do anything.
What Corrales showed, and what it did not show
Corrales and colleagues generated synthetic cyclic dinucleotide derivatives that activate all human STING alleles as well as murine STING — a real obstacle, since human STING is polymorphic and mouse-active compounds do not automatically translate. Injected into established mouse tumours, they produced profound regression, systemic responses that rejected distant lesions, and durable immunologic memory. That paper is where the enthusiasm for this target came from, and it is worth reading alongside what happened next.
The phase I result is the honest part
MIW815 (ADU-S100) was tested in 47 patients with advanced solid tumours or lymphomas, dosed intratumorally from 50 to 6,400 µg. A maximum tolerated dose was never reached; the common treatment-related events were pyrexia, chills and injection-site pain. Lesion size was stable or decreased in 94% of evaluable injected lesions. But there was one confirmed partial response — a Merkel cell carcinoma — and two unconfirmed. And the terminal plasma half-life was roughly 24 minutes. A compound engineered specifically for stability still cleared from the injection site in about the time it takes to read this page. Whatever the next generation of STING chemistry solves, that is the number it has to beat.
(Rp,Rp) is a synthesis problem, not a label
Putting sulfur on both bridging phosphates creates two stereocentres, and the diastereomers are not interchangeable — they differ in STING binding and in nuclease resistance. So a dithio CDN specified as (Rp,Rp) is making a claim about configuration at both positions, which has to be built in during synthesis and demonstrated afterwards. This is the same chemistry problem as stereodefined antisense, in a smaller molecule: Knouse and colleagues developed their P(V) reagent platform explicitly for both antisense oligonucleotides and cyclic dinucleotides, because it is the same stereocentre.
What to pin down when sourcing
Ask which stereoisomer you are being sent, and how it was assigned. Ask about the counterion — sodium, ammonium and free acid forms of the same CDN are different materials with different handling, and the salt is often what arrives when the paper specified the acid. Ask how the linkage isomer is specified: 2′,3′ and 3′,3′ cGAMP are distinct compounds with different STING affinity, and the shorthand hides it. For assay work, an unresolved diastereomer mixture may be adequate; for anything comparative, it is not.
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.
- Corrales et al. (2015) Direct activation of STING in the tumor microenvironment leads to potent and systemic tumor regression and immunity. Cell Reports
- Meric-Bernstam et al. (2022) Phase I dose-escalation trial of MIW815 (ADU-S100), an intratumoral STING agonist, in patients with advanced/metastatic solid tumors or lymphomas. Clinical Cancer Research
- Garland et al. (2022) Chemical and biomolecular strategies for STING pathway activation in cancer immunotherapy. Chemical Reviews
- Knouse et al. (2018) Unlocking P(V): reagents for chiral phosphorothioate synthesis. Science