In 2010 we published a paper on the synthesis and evaluation of pyrazolone compounds as SARS-coronavirus 3C-like protease inhibitors. The following year we published a review of 3C and 3CL protease inhibitors for anti-coronavirus and anti-picornavirus drug discovery.
At the time, coronavirus antivirals were an unfashionable subject. The 2003 SARS outbreak had ended, no coronavirus drug was on the market, and funding and attention had moved elsewhere. The two papers have since accumulated 147 and 119 citations respectively, and most of that accumulation happened after early 2020.
I want to describe why the work was done, because the reasoning is more useful than the result.
The target chose itself
The 3C-like protease is essential to coronavirus replication: the virus translates large polyproteins that must be cleaved at specific sites before functional replication machinery exists, and 3CLpro performs those cleavages. Inhibit it and replication stops. The enzyme is also well conserved across the coronavirus family, and it has no close human homologue — which means an inhibitor has a plausible path to selectivity.
Those properties were understood in 2010. What was absent was urgency. The judgement we made was that a conserved, essential, druggable target in a virus family that had already crossed into humans once was worth working on regardless of whether anyone was currently frightened of it.
What the chemistry showed
The pyrazolone scaffold was attractive because it is synthetically tractable, tolerant of substitution around the ring, and gave us a series in which structure–activity relationships could be read cleanly rather than inferred from a handful of scattered compounds. Systematic variation let us establish which substitution patterns mattered to inhibition, and docking studies gave a structural account consistent with the observed activity.
What the series showed was that activity tracked substitution pattern rather than any single functional group: the pyrazolines carrying a 1,3,5-triaryl arrangement were the ones that inhibited SARS-CoV 3CLpro. We then put those compounds against a panel — CoV-229E 3CLpro, and the 3C proteases of coxsackievirus B3, enterovirus 71 and rhinovirus 14 — and they were active there too. That cross-reactivity was the more interesting result. It said the scaffold was reading something the picornaviral and coronaviral proteases have in common, which is exactly the property you want if the goal is a starting point for a virus family rather than a single strain.
The compounds were early-stage. They were not drug candidates and we did not claim they were. What the series produced was a characterised scaffold and a structural rationale, published where others could use it.
The point about choosing problems
When COVID-19 arrived, groups working on SARS-CoV-2 main protease inhibitors did not begin from nothing. They began from a decade of accumulated work on a conserved enzyme, done largely by small groups during a period when nobody was watching. That body of work is why a protease inhibitor could be developed as quickly as it was.
I draw a specific conclusion from this, and it shapes how I advise research students and how I would allocate research funding as an academic leader. The problems worth working on are frequently not the ones currently attracting attention. Fundamental questions about conserved biological machinery reward patience, and the value of that work is often realised by other people, years later, under circumstances nobody predicted.
That is not an argument against applied research. It is an argument for institutions keeping some capacity pointed at problems that are important rather than urgent — because the urgent ones arrive without notice, and by then it is too late to start.