Initial intuition
CBASS follows a two-step signaling logic. A CD-NTase synthesizes a cyclic nucleotide (cGAMP, c-di-AMP, c-tri-AMP, c-AAA…) in response to phage infection. This signal activates a downstream effector (Cap4 for DNA cleavage, NucC for membrane complexes, CapV for phospholipases). The eukaryotic cGAS-STING parallel uses the exact same architecture.
Pharmacological study of STING has shown it is possible to design cGAMP analogues that bind to the recognition domain without triggering the downstream cascade. These decoys act as competitive antagonists. This strategy has been productive in immuno-oncology (agonists) and interferonopathies (antagonists).
Hypothesis
On bacterial CBASS effectors, phosphorothioate substitutions ([S]-cGAMP, [S₂]-cGAMP) and 2’-O-methylations on the ribose should:
- Preserve binding affinity to the Cap4/NucC allosteric pocket
- Block the conformational isomerization responsible for downstream enzymatic activation
- Resist degradation by endogenous bacterial phosphodiesterases (prolonged duration of action)
The ultimate goal is not an antibacterial therapy (CBASS protects the bacterium, not the other way around). It is identifying chemical scaffolds transferable to human STING—where selective competitive antagonists are sorely lacking in the pharmacopeia.
In silico plan (V1 campaign)
-
Pocket modeling: retrieve AlphaFold structures of selected CBASS effectors (Cap4 =
P0DXA6, CdnG =Q89Y83) + crystallographic PDB structures when available. Identify interaction residues with the native ligand via Mol* + alignment with STING-cGAMP complexes. -
Chemical enumeration: generate ~200 virtual analogues via RDKit (phosphorothioate substitutions, 2’-O modifications, base-cap variations). Filter by Lipinski’s rule adapted for cyclic nucleotides (Hattori 2020).
-
Docking and short MD (Bactaegion BYOK): affinity prediction by AutoDock Vina on the 200 analogues. Top 30 hits → 100 ns MD simulations to assess binding mode stability.
-
Go/no-go criteria: a hit must (a) have a Vina score < −7 kcal/mol on Cap4, (b) maintain a ligand RMSD < 3 Å over 100 ns, (c) not adopt an “activating” conformation compared to the native.
Limitations and risks
- CBASS effector allostery is not perfectly characterized crystallographically for all CBASS types. For Types I and III, we are likely extrapolating.
- Phosphorothioate analogues have a notoriously low bioavailability in eukaryotes (net charge, no dedicated transporter). Transposition to human STING will require significant prodrug work post-Bactaegion.
- The off-target risk on human innate immunity is serious. An accidental STING agonist would trigger an interferonopathy. The screen must include systematic anti-activation STING control.
Links to the Bactaegion scope
This hypothesis directly concerns the V1 CBASS target (high translational priority, small molecule modality). It follows the strategy listed in the family sheet: screen stable non-hydrolyzable mimetics against STING and bacterial effectors as probes.