scenario · intermediate · partly written
Potent on the enzyme, weak in cells: the compound or the cell?
A lead optimisation decision from a real coronavirus-protease programme: what to do when a tight enzyme binder looks far weaker in infected cells.
Situation
A ketone-based inhibitor binds the SARS-CoV-2 main protease with a Ki of 0.27 nM, yet in infected monkey Vero cells its EC50 is in the micromolar range (4.8 uM against SARS-CoV-1 in Vero 76 cells in the first paper), orders of magnitude weaker. The standard gut-permeability assay cannot rank the series because the compounds are beyond its sensitivity. You must choose where to spend the next round of chemistry. 12
Reported, not yet replicated
Ask
What do you do next?
Options
Redesign the molecule to cross membranes
Raise lipophilicity and cut polar surface area and hydrogen-bond donors, then re-test in the same cells. 1
Cost. A full design-make-test cycle; no cost was found in sources.
Tells you. Whether physical-property edits move the cell number.
Does not tell you. Whether the cell assay itself is the problem. In the real series, modest alterations failed to significantly decrease the high cell-to-enzyme ratios within the warhead class.
See: Caco-2 permeability assay
Reported, not yet replicated
Re-run the dose-response with a pump inhibitor
Add an efflux-pump inhibitor to the same cell assay and see whether the curve shifts left. 21
Cost. One extra arm in an assay you already run; no price found.
Tells you. Whether pumping in that cell line hides potency. The shift is the evidence, and a control in human cells that does not shift supports it.
Does not tell you. The intracellular concentration (not measured in any paper opened), or why the compound is poorly absorbed from the gut. The inhibitor is itself toxic at higher doses: it was used at 2 uM against an alone-EC50 of 23.4 uM.
See: Dose-response assay
Reported, not yet replicated
Move to human cells and a human-relevant virus
Replace the monkey line with human lines and re-measure. 1
Cost. New cell lines and biosafety work; no price found.
Tells you. The potency in the cell type that matters for the dose target.
Does not tell you. Anything about the cause when the virus also changes: in the first paper the monkey and human cells were infected with different viruses, which the authors say could contribute, so the comparison is weak alone.
See: Dose-response assay
Reported, not yet replicated
Measure efflux directly in a transporter-expressing cell layer
Run a cell-layer assay with the pump overexpressed and read an efflux ratio. 1
Cost. No price or timing was found in a source.
Tells you. Whether the compound is a pump substrate, separately from the virus assay.
Does not tell you. Why it is poorly absorbed overall. No efflux ratio for this compound was found published, so what it would have shown is unknown.
See: Caco-2 permeability assay
Our inference
What really happened
The pump inhibitor control
In the first paper, adding a pump inhibitor at 0.5 uM moved the monkey-cell EC50 from 4.8 uM to 0.11 uM (the authors' figures, about 44-fold) while a human cell line showed no shift. A later paper found, for SARS-CoV-2 in two monkey lines, 39.7 and 88.9 uM falling to 0.23 and 0.76 uM with 2 uM inhibitor, and no statistically significant difference in human cells. A VeroE6 line with the pump gene knocked out later supported the effect for the successor nirmatrelvir. 123
Reported, not yet replicated
Where it led
The dose target was set from a human-cell EC90 and not the pump-masked monkey value. The oral series replaced the warhead and went forward as nirmatrelvir with ritonavir, which cut hospitalisation or death in a randomised trial. Nearly all of this evidence is from the maker's own authors. 24
Reported, not yet replicated
Still open here
- The permeability and potency numbers for the oral edits were not opened at source and are not given.
Linked from
Sources
- Hoffman, Robert L., Kania, Robert S., Brothers, Mary A., et al.. Discovery of Ketone-Based Covalent Inhibitors of Coronavirus 3CL Proteases for the Potential Therapeutic Treatment of COVID-19. Journal of Medicinal Chemistry (2020). paper; opened 2026-11-17 all records citing this
- Boras, Britton, Jones, Rhys M., Anson, Brandon J., et al.. Preclinical characterization of an intravenous coronavirus 3CL protease inhibitor for the potential treatment of COVID19. Nature Communications (2021). paper; opened 2026-11-17 all records citing this
- Zhu, Yuao, Binder, Joe, Yurgelonis, Irina, et al.. Generation of a VeroE6 Pgp gene knock out cell line and its use in SARS-CoV-2 antiviral study. Antiviral Research (2022). paper; opened 2026-11-17 all records citing this
- Hammond, Jennifer, Leister-Tebbe, Heidi, Gardner, Annie, et al.. Oral Nirmatrelvir for High-Risk, Nonhospitalized Adults with Covid-19. New England Journal of Medicine (2022). paper; opened 2026-11-17 all records citing this
Where it sits on the map
History of this record (1)
This record has been changed 1 time, on the dates below. The note kept against each change is written for the people who maintain the atlas, in their own shorthand, so it is not shown here; it is in the record itself.
- Status
- partly written
- Last updated
- 2026-11-17
- Record
- content/scenarios/potent-enzyme-weak-in-cells-efflux.json