I really enjoyed reading this article in J. Med. Chem. on curcumin. (Update: here’s the take over at Practical Fragments). That’s a well-known natural product, found in large quantities in turmeric root (which is where most of the yellow color comes from). It has, over the years, been a hit in many, many assays, and I would not want to guess how many derivatives of the structure have been prepared along the way. There is an entire web portal that attempts to bring together all the publications (thousands) and patents (hundreds) relevant to the field. More than that, a quick search will reveal that there are plenty of people ready to sell you curcumin tablets in every variety, for a list of ailments longer than most people’s legs. Everything from Alzheimer’s to upset stomach – curcumin is good for what ails you, apparently.
So what do you get when you look closely at the molecule and its activities? Well, for one thing, curcumin’s stability and pharmacokinetics are absolutely terrible. It’s less than 1% bioavailable, and its half-life under physiological conditions is measured in minutes. This makes a person wonder how it can be such a wonder drug. The authors of the current paper, indeed, state that “To our knowledge, (it) has never been shown to be conclusively effective in a randomized, placebo-controlled clinical trial for any indication“. How about as a lead compound, a hit from an assay? The situation is just as bad. The paper cites a long list of references demonstrating that curcumin participates in pretty much every undesirable behavior possible in an assay: it reacts with proteins, it’s a redox cycler, it coordinates metal ions, it aggregates, it disrupts membranes nonspecifically, it interferes with fluorescent readouts, and it decomposes. Other than that, it’s a perfectly good hit.
In cell, tissue, and animal assays, it falls into the category of “Invalid/Improbable Metabolic Panacea” (IMP), a term that’s been used for natural products that seem to have all sorts of weakish activity all over the place. Unfortunately, the literature is full of reports of curcuminoid compounds as actives for all sorts of specific molecular targets (this latest review goes into several of these). Taking the properties and activities all together, though, these efforts would seem to be largely – and almost certainly completely – wastes of time and money. What’s worse, it appears that from 2001 on there are at least 135 registered clinical trials using the stuff. The review goes into four examples for which details are available, all of them complete wipeouts, and those would seem to serve as good proxies for the rest given that no curcumin trial has ever reported any convincing positive results.
The paper ends with a very sensible set of recommendations for anyone doing natural product bioactivity research (and indeed, these are also high recommended for anyone screening any set of compounds. They are not new – everyone should know this stuff. Check your compound’s solubility and aggregation under the assay conditions. Check its stability and purity. See if the activities you’re getting make any sense given those numbers. Check the selectivity data, and not just against something that’s very close to your target. Use biophysical methods (several, preferably) to confirm that your compound is really hitting the target, or some target – don’t just rely on cellular readouts without digging further. And so on.
These things are a pain in the ass, of course. They take more time, more effort, and more money. They will slow down your publication plans, and often enough they will invalidate your entire line of inquiry and you may have to find something else to do or come up with a completely new idea. It’s no mystery why so many labs slide over many of these tasks, because they’re really annoying. You should only follow these recommendations if you care about your research being valid and reproducible, and if you’d like it to lead to something useful. But if you just want to bang out some publications, scoop up some money, and fill the scientific literature with more waste products, then carry on as before. There are lots of experiments you can do with curcumin.