Pfizer and Ligand had their application for Oporia (lasofoxifene) shot down by the FDA this week. Details are scarce about the grounds for this action, but lasofoxifene was supposed to be one of Pfizer's string of billion-dollar drugs - you know, the ones that they need every single one of to meet their growth targets. It's a selective estrogen receptor modulator (SERM) for osteoporosis, a useful therapeutic niche which at present is pretty much owned by Lilly.
And it's a hard area to make a living in. Nuclear receptors are wonderfully interesting things, and they're wonderfully complex if you're doing basic research on them. If you're doing applied research, on the other hand, they're hideously complex - same coin, different sides. There's a large family of NRs, including the PPARs that I was speaking about a few days ago, the steroid receptors, and many other odds and ends.
They all work at the gene transcription level, and they're one of the few ways that we can mess with that world through small drug-like molecules. Many nuclear receptors can bind small molecules (like a steroid or fatty acid), and that sets off a complicated chain of events. They then often pick up another nuclear receptor protein in a sort of face-to-face binding, and that other protein one can be another of its own type or a different one. (The various retinoic acid receptors are particularly known for heterodimerizing with other members of the family.)
Then that beast picks up a number of other proteins, forming a very large complex, and this is where we lose our ability to understand what's going on. These cofactors are involved in the binding of the whole complex to stretches of DNA (a closeup view) and its subsequent readout into messenger RNA. But some of them seem to inhibit this process, and some to enhance it. To make things more interesting, the nuclear receptor complex can bind to dozens (hundreds?) of different genes, and the cofactors can switch roles (promoter, inhibitor) depending on which one they're involved with. We don't know how many different cofactors there are - a bunch, that's for sure - and different types of cells can have very different suites of them available. (Those profiles change with time and the external environment, too.) Here's an extremely well-done Flash lecture from the comprehensive NURSA site if you want to become an instant nuclear receptor nerd. I took the step several years ago, myself, when it had to be done by hand, sonny.
All this variability makes it foolhardy to generalize about nuclear receptor actions. If you say that Cofactor X is a transcriptional inhibitor, someone can come up with a system where it does the opposite. If you say that Nuclear Receptor Y is responsible for the transcriptional regulation of Gene Z, someone can show you a cell line where it has nothing to do with it, or where it does the reverse of what it does in yours. That's how the Selective Estrogen Receptor Modulators (SERMs) work. The ones you want work in opposite ways in bone tissue versus breast tissue, and if the signaling pathways were simple, that probably wouldn't be possible. This particular split seems to work through the different cofactor profile mechanism.
The complexity of the field has made some companies just throw up their hands and decide to spend their money somewhere else. For the same reasons, it's made others decide to concentrate only on that area, and Ligand is the best example. It's been quite a rough ride for them over the years. They've had (and still have) deals with several major companies, and have had all sorts of things go into development, but there haven't been any home runs yet. Worse, they're in the middle of some accounting problems (to put the matter delicately) that have led to an SEC investigation and their stock being recently delisted from NASDAQ. No, tedium has not been a big problem.