There's an article in C&EN by Sarah Everts on a new paper by Eric Block's group at SUNY Albany that throws a rather large stone at the window constructed by Luca Turin's vibrational theory of smell. As readers interested in the topic might recall, it's a saga going on for more than a decade now, and as someone who gave a graduate school seminar on the topic I have a mild but longstanding interest in it. The two positions seems to be firmly staked out: one side declares that smell is because of shape while the other declares that vibration also plays a role.
The new paper does some clever detective work in which the authors express the human musk receptor in kidney cells and expose it to various deuterated and non-deuterated musk compounds (in a nutshell - carbon-deuterium bonds vibrate at a different frequency from carbon-hydrogen bonds, so the receptors should be able to tell that difference if the vibrational theory is right). They found that the receptor is unable to distinguish between these compounds at the molecular level. Since Turin's hypothesis is essentially one at a molecular level this is a serious challenge. In response Turin conjectures whether they might have used the wrong receptor or expressed it in the wrong kind of cells - and the complexity of such systems makes both of these arguments not entirely worth dismissing.
The chemical blogosphere's very own Derek Lowe has some thoughtful remarks in the C&EN article which he distills into a post. Curiously the article does not mention Turin and his co-authors' work on fruit flies clearly distinguishing between acetophenone and deuterated acetophenone. To me that was a rather compelling study, but again one performed on a much 'simpler' system.
Nevertheless, the current work definitely throws a gauntlet in front of Turin's theory, and since the shape theory is well-established with truckloads of evidence on its side, the burden is still on Turin's shoulders to demonstrate the validity of his assertions.
Nobody knows what the truth is in this case right now, which of course makes the issue even more interesting. My own recent take on the shape vs vibration dichotomy is that it might be much less of a dichotomy than we think. There is no reason (except that offered by Occam's razor) to discount the possibility that nature might be using shape some of the time and vibration some of the time. Perhaps it's 80% shape and 20% vibration, or perhaps receptors are actually able to tune the shape/vibration mix depending on the context.
It's worth recalling that smell is a very primitive sense and the primordial environment was filled with a vast, dazzling, confusing complexity of chemical compounds. In the absence of heuristic intelligence, evolution needed to muster an equivalent complexity of physicochemical responses to sample that chemical complexity. To that end we shouldn't really be surprised if it turns out that shape and vibration are not really discrete phenomena but lie on a continuum. Maybe it's like wave particle duality - that convoluted gem of a weird theory where what we see depends on the experiment we choose to perform. Doing those right experiments and nailing down the proportions is going to be the tricky part, and it seems that we are making headway in that direction all right.