Advocate / skeptic debate

Debate Transcript

Advocate — round 1

I'll make the case for acceptance of this manuscript, recognizing that "acceptance" for In Silico means publication alongside a full, candid review record — not certification of a finished story.

The core discovery is real and independently corroborated. Every specialist panel converged on the same conclusion: IPO9 is a genuine, previously unknown target of SR-218 and a genuine regulator of cGAS activity. This isn't a single assay result — it rests on convergent, orthogonal evidence: shRNA knockdown eliminates SR-432 labeling of the correct molecular weight band (Fig. 2c), recombinant and overexpressed IPO9 are labeled and competed specifically by active but not inactive analogs (Fig. 2d-h, Extended Data Fig. 4a-c), SAR for binding tracks SAR for phenotype across an expanded compound series, and a structurally unrelated IPO9 ligand (dbk-032A) cross-competes (Extended Data Fig. 4g). The scientific validity reviewer, despite proposing further confirmatory experiments, explicitly states the target identification pipeline is "rigorous" with "strong SAR correlation."

The mechanistic distinction between H2A-H2B and NCP inhibition is a clean, well-controlled finding. Fig. 3d shows IPO9 rescues cGAS activity from H2A-H2B inhibition but not NCP inhibition — a within-experiment negative control that strengthens rather than weakens the claim, since it rules out a trivial explanation (e.g., generic IPO9 activity enhancement). The reporting/reproducibility reviewer calls this "cleanly separated" evidence for "a functionally distinct regulatory axis."

The cryo-EM structure, while acknowledged by the authors as resolution-limited, is not a weak contribution — it is an honest one. The manuscript explicitly states "resolution limits our ability to draw definitive conclusions about the interaction at an amino acid level" and frames the H18-19 loop mechanism as a plausible structural hypothesis consistent with BLI data, not a proven mechanism. This is precisely the kind of calibrated, non-overreaching claim this venue should reward — the contribution reviewer notes the paper "appropriately acknowledges mechanistic uncertainty."

The panel's substantive concerns (physiological IPO9/H2A-H2B concentrations, direct-displacement vs. sequestration, IACUC/IRB statements) are genuine and should be addressed in revision, but they are refinements to an already-standing, multiply-corroborated core finding, not refutations of it.

Skeptic — round 1

The load-bearing objection is that the panel converges on a single unresolved gap: the manuscript never establishes causal necessity of IPO9 for SR-218's phenotype. As the scientific_validity report states, "Figure 3b–c shows that IPO9 knockdown reduces phospho-STING to similar levels as SR-218 treatment, but does not directly show whether SR-218 retains activity in IPO9-depleted cells." Correlation between target-engagement SAR (Fig. 2g-h) and phenotypic SAR is consistent with, but does not prove, IPO9 as the operative target — SR-218 could hit IPO9 and something else, with the phenotype driven by the latter. This is FIXABLE: an epistasis/rescue experiment (SR-218 dose-response in IPO9-knockdown vs. WT cells) is a standard, achievable test that the authors could run and report.

Second, and more serious for the mechanistic claims of the title itself ("IPO9 modulates histone-dependent inhibition of cGAS"): three reports (contribution_context, reporting_reproducibility, scientific_validity) independently flag that the three proposed mechanisms — direct displacement, sequestration of free H2A-H2B, and sequestration from DNA — are not distinguished by any experiment presented. The authors themselves concede this in the Discussion ("our data supports three non-mutually exclusive mechanisms... we cannot exclude the possibility that multiple mechanisms contribute"). Given that this is a genuine authorial admission rather than a reviewer-invented flaw, it should lower confidence in, not the honesty of, the mechanistic model. This is FIXABLE — it requires additional biophysical dissection (soluble-phase BLI, SEC/AUC, sequential titration) that the panel specifies, but is not a fatal flaw given the paper's more modest, well-supported claim that IPO9 antagonizes H2A-H2B-mediated inhibition in some fashion.

Third: the ethics report flags two HARD gaps — no IACUC reference for the mouse work and no IRB/consent statement for human PBMCs. These are compliance omissions, not scientific flaws, and are trivially FIXABLE by adding protocol numbers, but as raised they are non-negotiable for publication and no other reviewer addressed them, so this is a genuine panel blind spot outside the science reports.

None of these amount to a fatal flaw in the core discovery (IPO9 binds SR-218, and IPO9 antagonizes H2A-H2B/cGAS in vitro is directly demonstrated biochemically), but the "regulator of cGAS-STING signaling" framing in the title and abstract currently outruns the cellular necessity evidence.

Skeptic — round 2

Conceding what's been established: I agree the IPO9-SR-218 target engagement chemistry is solid, and the H2A-H2B vs. NCP differential in Fig. 3d is a genuine, well-controlled result. I'm not disputing that IPO9 binds SR-218 or that IPO9 antagonizes H2A-H2B-mediated inhibition of cGAS in a test tube. My objection is narrower and, I think, decisive for how this paper is titled and framed.

The load-bearing gap is causal necessity in cells, and it is untested, not just under-tested. The title claims "IPO9 modulates histone-dependent inhibition of cGAS" as a cellular/biological fact. The only cellular evidence is Fig. 3b-c: IPO9 knockdown phenocopies SR-218's effect on dsDNA-dependent phospho-STING. But as the scientific_validity report notes, no experiment asks whether SR-218 still works in IPO9-depleted cells — the single epistasis test that would rule out an off-target driver of the phenotype. The data_analysis report adds that even the phenocopy itself is shaky: for SR-717 stimulation "both knockdown lines show p>0.05 (sh1 p=0.0951, sh2 p=0.0741)," and even for the dsDNA condition sh1 doesn't reach significance (p=0.0465 for sh2 only, borderline). This is not a nitpick about post-hoc tests — it's the paper's only in-cell evidence for the pathway model, and it's weaker than the narrative text ("phenocopy") implies.

This is FIXABLE, not fatal — a dose-response of SR-218 in IPO9-knockdown cells is a straightforward experiment, not a new discovery program. But it must be flagged clearly in any published review, because right now the abstract's central mechanistic claim rests on correlation (SAR tracking) plus an underpowered knockdown phenotype, not a rescue/epistasis result.

Panel blind spot: no reviewer tested whether the in vitro biochemistry (Fig. 3d, 3f) generalizes to physiological stoichiometry — contribution_context raises this but no one checked it against literature estimates of nuclear H2A-H2B pools, leaving the "extrachromosomal H2A-H2B" framing speculative.

The ethics HARD gaps (IACUC/IRB) are non-negotiable but trivially fixable and don't bear on scientific validity — I'd separate those from the scientific score entirely.

Advocate — round 2

I concede the two substantive points cleanly. The epistasis gap is real: no experiment tests whether SR-218 retains activity in IPO9-depleted cells, so causal necessity for the compound's phenotype is inferred, not shown. And the three-mechanism ambiguity (displacement vs. sequestration vs. DNA-competition) is genuinely undistinguished by any single experiment — though I'd stress this is the authors' own admission, not a reviewer's discovery, and per this venue's stated criteria ("reward candour rather than treating an admitted weakness as a new one"), that transparency should count in the paper's favor, not against it.

But neither gap undermines the finding that is directly demonstrated: IPO9 physically binds SR-218 (multi-orthogonal target-ID, uncontested by any reviewer), and IPO9 biochemically reverses H2A-H2B-mediated (not NCP-mediated) inhibition of purified cGAS in a clean, internally-controlled reconstitution (Fig. 3d). That is a load-bearing claim the evidence does support, independent of whether it's the sole mechanism operating in cells. The title's "regulator of cGAS-STING signaling" is arguably scoped too broadly relative to the necessity evidence — that's a fair, specific, revisable critique, not grounds for rejection.

On the IACUC/IRB gaps: these are compliance-documentation omissions, correctable by adding protocol numbers in revision — Scripps Research studies of this kind are routinely IACUC/IRB-governed, and the omission is a reporting lapse, not evidence the work was conducted improperly. In Silico's model is to publish the finding alongside the review noting the gap; that is exactly what this record should do, not a reason to withhold the listing entirely.