Monday, July 9, 2012

The Man Who Voled the World

Last Monday, Dr.Hasse Walum gave a talk titled "Genetic and Hormonal Influences on Pair Bonding Related Behavior in Humans" at the Center for Translational Social Neuroscience at Emory. I hadn't heard of Walum's work before I saw the e-mail announcement for his talk, but a little googling got me interested. Here's the most titillating version of his findings: Walum found the gene that makes men cheat.

Okay, that is most definitely not what he found, and I got the sense from talking with him briefly that he would be the first one to tell you that. So why am I misrepresenting his results?

Dr. Hasse Walum: hard-hitting Wired reporter  David Ewing Duncan compares him to Kurt Cobain, but my science and rock star senses detect a David Bowie influence


I want to make some points about how science interacts with the media. Like I said, it was reading news reports that made me want to go hear Walum talk. As a graduate student, in his first published study, Walum reported that different forms of a certain gene are associated with the strength of relationships that men form. As he told me after his talk, he spent the week after the study was published just answering phone calls from the press. On Wired’s site I found an article about the paper that struck me as fairly cautious, even if the author did point out that "Walum did not report if he carries the tell-tale gene". Then again, on the BBC website I found a piece that opened up with a pic of actor George Clooney, pointing confidently at someone off camera, along with a caption that asks if he's "afraid of commitment?" Clearly, the press sensationalized this story. In a recent post on this blog, Emory neuroscience grad student Jordan Kohn put a lot of blame on journalists for the sensationalizing of research results. Unfortunately, there's plenty of blame to go around for the way that science gets represented in the media, and I think some of it should fall squarely on the shoulders of the science establishment.

Let's face it; Dr. Walum's study wouldn't have been published if the title was: "We can't find any evidence that this gene has anything to do with dudes that suck at monogamy". He told me as much after his talk. I don't mean to use Walum's words against him--in fact, I think anyone who's studying anything remotely interesting about the brain faces the same ethical dilemma. I'll say more about that after I sum up the seminar.

Ratty (actually a water vole)
What I didn't realize when I got the e-mail announcement was that Dr. Walum was here giving a job talk. The reason for Dr.Walum's visit became obvious when he got to the slide in his Powerpoint presentation featuring a picture of Ratty from Wind in the Willows. Ratty, for the uninformed, is a water vole, and Dr.Walum's childhood love of Ratty establishes that he's wanted to work with voles ever since he was a kid. If that didn't convince us, he also informed us that he wrote a literature review of filial mate bonding for his degree project as an undergrad, and that the inspiration for his work was Larry Young's studies of the vasopressin receptor gene in voles.

As you are no doubt aware if you have spent more than five minutes on Emory's campus doing something besides playing beer pong, a significant component of Emory's neuroscience research revolves around voles. Voles provide a convenient model for understanding how monogamy works in the brain. Species like montane and meadow voles are promiscuous, but the prairie vole is monogamous. Young's group has shown that this difference is due in large part to a 482 base-pair long snippet of DNA which the prairie voles carry. This length of DNA lies in the regulatory region of the vasopressin receptor gene—that is, the DNA near the gene itself that affects how the cell’s machinery churns out vasopressin receptors. As the name implies, these receptors bind vasopressin, a hormone known to play a role in bonding (and in other more mundane physiological tasks, like water retention). By inserting the same length of DNA in the regulatory region of the vasopressin receptor gene in mice, which are easier to manipulate genetically, you can increase affiliative behaviors in males. Male mice carrying the insert show increased levels of olfactory exploration and grooming of females. Now all we have to do is figure out how that change in regulatory regions translates into a change in behavior, and...viola! We've figured out monogamy. And cured autism. Maybe.

Not everyone feels the love for the Young lab's research agenda. Last week this blog also featured an interview with feminist science studies scholar Angela Willey. While at Emory, Dr. Willey problematized the heck out of the Young lab's research on monogamy. I'm still processing that post. The neuroscientist in me wants to defend the Young group's work with adjectives like "elegant". On the other hand, I'm sympathetic to feminist and queer critiques of neuroscience research, and I think it's important for neuroscientists to engage with those critiques. To their credit, the Young lab did just that. I could echo Willey's points about how their definition of monogamy seems to reduce a very complex human behavior to a single number which they assign to vole behavior--a simplification I'm sure the lab is aware of--but let me stay focused on the seminar.

Walum’s began his with that first paper on the human vasopressin receptor gene that got him all the press. I'll state what he and his colleagues reported in scientific terms. Then I'll explain those terms, just in case anyone finds them as opaque as I once did, before undergoing years of training that allow me to speak complete gibberish to strangers. Their results show a correlation in humans between scores on a Partner Bonding Scale and variations in a microsatellite upstream of the human gene for the vasopressin receptor. Dr. Walum created the Partner Bonding Scale, as he proudly told us, basing it on the behaviors that are measured to put a number to the strength of pair bonds between great apes and other nonhuman primates. By sequencing microsatellites in roughly two thousand Swedes, Walum et al. were able to show an association between scores on his scale and the version of the microsatellite that men carried. Microsatellites are short regions of repeats in DNA. When they occur in the regulatory regions outside of a gene, these microsatellites can affect gene transcription, and by extension the protein that the gene encodes. Case in point: the vasopressin receptor has a different distribution in the brains of the monogamous prairie voles and the promiscuous montane voles. As outlined above, the Young lab has shown that a 428 base pair insert in the genome of the prairie vole, right next to one of the vasopressin receptor genes, causes this change in receptor distribution. Similarly, Walum et al found that men who carried the so-called "334 allele" were significantly more likely to receive lower scores on the Partner Bonding Scale. You might know that alleles are different versions of a gene, or in this case microsatellite. There’s not 334 versions—the number refers to the length of the repeat. The effect they found was “dose-dependent”: men that carried one or two copies of the 334 allele were much more likely to have lower Partner Bonding Scale scores. It’s worth emphasizing, though, that the 334 allele does not show any sequence similarity to the 428 base pairs of prairie vole DNA that supposedly make that species more monogamous, and neither does any other part of the regulatory region flanking any of our vasopressin receptor genes. Different microsatellites, related effects. At least that’s what Walum et al. argue.

Cyberball: image taken from a study
of "social exclusion" (Bolling et al. 2010)
Dr.Walum went on to talk about the other studies he carried out as a grad student. One set of experiments looked at the effects of oxytocin, another hormone that some argue increases trust. To try and measure the bond that female subjects formed with males they were partnered with during the experiments, Walum and company used a videogame of sorts they called “cyberball”. In cyberball, the female subjects could choose to “throw” a ball to either an icon of their partner’s face or to an icon of an unfamiliar male. If female subjects that recevied intranasal oxytocin more readily trusted the male partners, it should show up in the cyberball score. After the talk was over, I went up talk with Dr. Walum. I had to know, did the data from cyberball tell them anything? At first he said he couldn’t remember. Then he said something about “trending in the same direction”, only to interrupt himself and say that the scores had only been significant when lumped together with other metrics they’d used. I replied sympathetically. “It’s hard to find a way to measure human behavior.” Then I told him that the Cyberball game reminded me of the Partner Preference Test, a behavioral assay designed by the Young lab to measure monogamy. In the test, voles are put in a three-part cage, with their partner on one side and a stranger on the other. Prairie voles tend to spend more time with their partner. I asked Dr. Walum if he was trying to find a way to do the Partner Preference Test with humans. “Yes, exactly,” he said, “but you can't do that with humans. You can't do the experiments you would like to do." I agreed. "I don't think that an experiment proposing that you put humans in a large plastic tank would make it past the Institutional Review Board." He went on: “With human subjects, it's all about variation. You can do some pharmacological manipulations, like with the intranasal oxytocin, but with our first study [on the vasopressin receptor allele] we had to think more about our story than about our results”, he said. “In a way we published the story that everybody wanted to hear.”

Again, I’m not trying to use Dr. Walum’s words against him. While I might not have the expertise to critique his study’s use of Generalized Linear Mixed Effect Models, I also don’t have any reason to doubt his results. As Dr.Walum made perfectly clear in the Karolinksa Institute press release that the Wired and BBC alike were so quick to quote, “There are, of course, many reasons why a person might have relationship problems.” What am I trying to say is that we should keep the economic realities of science in mind when we talk about how research filters through to the media. Every neuroscientist who complains about how the only things the press can do is cut and paste pretty pictures from fMRI studies should remember that, right now, those are the studies that get published, and by extension, the studies that get media attention. Like it or not, the axiom of “publish or perish” still applies. All these papers that only trot out sexy results “provide strong evidence”, as us science types like to say,  that studies should be registered and accepted by journals before the experiments are carried out. Scientists already write their grants this way—they sell the research they’re going to do—so why shouldn’t that be what journals are buying? This is the system that Neuroskeptic has advocated on his blog (here’s a hyperlink by way of citation). If the experiments don’t provide any evidence, then publish that lack of evidence, and save everyone else the effort of pointlessly repeating the same study.

Under this system, if neuroscientists find themselves angry about how the media represents their results, at least they can tell themselves it’s not because they're only publishing what gets them funding. I don't think that system would put an end to press releases that tout tantalizing findings, though. After all, someone has to sell the science so more science can get done. I'm guessing that's the explanation I'd get if I talked with the Man who Voled the World.



Want to cite this post?
Nicholson, D. (2012). The Man Who Voled the World. The Neuroethics Blog. Retrieved on , from http://www.theneuroethicsblog.com/2012/07/man-who-voled-world.html

9 comments:

Laura E. Mariani said...

Nice post, David. I'm going to ignore all of your valid points about the interaction between science and the media and how it could be improved upon and focus on nit-picky details about the experiments discussed at the seminar.

Firstly, I have to say that I actually find it comforting that the V1aR microsatellites differ between humans and rodents. The rodent studies really do seem to imply that the prairie vole microsatellite acts like an on/off switch for social monogamy -- you can drastically alter the animals' behavior using this single bit of DNA. It seems obvious to me that human pair bonds are much more complex and don't have the same all-or-none quality as the voles', and maybe this is partly due to changes in the vasopressin system that occurred during evolution. (Along with lots of other factors affecting human behavior that voles don't have -- stratified social structures, culture, Facebook, etc.)

Secondly, can we talk more about cyberball? I get that you can't recapitulate the partner preference test in humans, what with the large plastic tank and leashes that it would require (two out of the three voles in that paradigm are physically restrained, so they can't escape when their partner wants to huddle with them!). But, was there really no better alternative than a video game that looks like it was designed in the '90s for an episode of "Clarissa Explains it All," where you don't even interact with your partner in physical space? I guess you have to go with whichever assays have been previously accepted in the social behavior field, but this one seems so strange to me.

Lanikea King said...

David,

Just a nitpick here.

What’s up with the “faithful fathers” bit in your summary of the microsatellite research? If it’s meant to be an ironic usage, I don’t see any language clearly indicating that you did mean it that way. If you are trying to say this phrase would be the way the media portrays some of the vole work or even how the researcher would portray the work, then you should probably state that more clearly. Otherwise, it sort of looks like you’ve taken license to simplify and hype the results into a good read, which is exactly the type of behavior you seem to be criticizing here.

Kristina Gupta said...

Great post David.

One question and one comment:

Question: Do you know any more about this Partner Bonding Scale? I think a close reading of how Walum is measuring pair bonding (especially if he is basing it on great apes?) could be very interesting.

Comment: I appreciate your message that neuroscientists can't simply blame the media for the way neuroscience is reported (really, we all carry responsibility for this). For example, I know Larry Young feels that his research has been misrepresented by the media, because the media has confused monogamy with fidelity. But, he is the one who published a more “popular” article titled “Being human: love: neuroscience reveals all” (http://www.ncbi.nlm.nih.gov/pubmed/19129828) which is fairly sensationalized, so I think he is at least partially responsible for some of the media frenzy around oxytocin as the “love drug” (e.g. http://abcnews.go.com/Technology/love-drug-oxytocin-cuddle-chemical-scientists-makes-mothers/story?id=15330910).

Thanks!

Kristina

David Nicholson said...

@Laura: yeah, I'm not sure why the authors chose cyberball, except that it was used in previous studies, like the one I took that image from, which was an fMRI study--so maybe they thought cyberball would keep keeps in the scanner from moving their head and disrupting the scan? As far as I know, though, Dr.Walum's study did not use fMRI.

David Nicholson said...
This comment has been removed by the author.
David Nicholson said...

Hey Lani,

I don't think that's a nitpick--I think you're exactly right. The phrase "faithful fathers" bothered me when I was re-reading my drafts, but I didn't take the time to go look at the original study before I published the post. Now I have, and I edited that sentence to reflect it. (For anyone who reads the post after I've edited it, it used to say that "you can convert other male rodents--such as mice, which are easy to manipulate genetically--into faithful fathers if you insert into their genome a regulatory region from the gene in prairie voles that encodes the vasopressin receptor".) As a result the sentence reads like a textbook, which just goes to show you how hard it is to make a story about science interesting, and to faithfully report results at the same time. I hope I made it clear that I don't have a problem with hype as much as I have a problem with science in general focusing on results that lend themselves to sensationalizing. Maybe I just need to learn how to hype negative results.

David Nicholson said...

Hi Kristina,

Good question...what do I know about the Partner Bonding Scale? I know that, in the methods section of the original paper, the authors cite several sources on measuring pair bonds between nonhuman primates. The authors go on to say that "in accordance with the behavioral domains observed" to measure those bonds in other primates, they borrowed questions from self-report surveys like the Dyadic Assessment Survey, the Support Seeking and Giving assessment, and the Marital Instability Scale. In other words, they kept questions from those surveys that they felt would quantify the same things that primatologists measure, like "spatial closeness between two individuals". It's easy to cast doubt on how accurately a person answering a question like "How often do you kiss your mate?" measures "spatial closeness", but if I had to argue Dr.Walum's case, I'd point out that we really don't have many other ways to measure these things, short of shooting people with tranquilizer darts and then attaching radio tags to them. Actually, Sweden has come pretty close to doing just that. One thing I didn't mention in my verbose post is that Walum's studies rely on data from massive Swedish databases that track twins throughout life--in the case of the study I talk about here, they used data from TOSS (Twin and Offspring Study in Sweden). One criterion for inclusion in TOSS is that "adult individuals were part of a dyadic relationship that had persisted for at least five years". Is that a confound for the study? Dr.Walum wasn't shy about pointing out that it's a possible limitation, and they tried to control for it by including questions like "Have you discussed a divorce or separation with a close friend?".

Thanks for backing me up on my message, such as it is. I meant to focus on how the media can affect science that actually gets done. It was all too easy for me to imagine myself in Hasse Walum's shoes, (since we both own Converse) trying to pull off a nice clean study like the Nature paper where they change mouse behavior by inserting a piece of DNA upstream of the vasopressin receptor gene, and then finding myself taking phone calls from the press all day because I reported that I'd reproduced the vole results in a species we seem to be much more obsessed with. I hope that as a post-doc he can do the experiments he's been wanting to do since he was an undergrad.

I'm not sure how sensationalistic I think the Nature essay from Larry Young is, but after staring at the Walum, et al., article for a while, I'm surprised that he doesn't mention any of the caveats. After talking about the way scientists sell their work with other people, I think what really bothers me is that scientific literacy overall has dropped to the point that there's no room for fine distinctions. We can't just talk about a cool result that might teach us something about how our brains work, because that wouldn't be worth funding. People would ask why scientists are wasting their tax dollars on voles. The end result is there's no way researchers can talk about this work without making it sound like they're about to synthesize a love potion (and then write an essay to warn us about the ethical implications).

Cyd Cipolla said...

"I think what really bothers me is that scientific literacy overall has dropped to the point that there's no room for fine distinctions.We can't just talk about a cool result that might teach us something about how our brains work, because that wouldn't be worth funding. People would ask why scientists are wasting their tax dollars on voles."

Yes. 100%. It's also worth adding this isn't just an issue in lab sciences that have to use external grants: there was an uproar here in Georgia a few years ago about the use of tax money to fund classes on sexuality. (http://chronicle.com/article/Georgia-Legislators-Say-State/42375/)

Really great post, David, (as usual!)

David Nicholson said...

Thanks, Cyd. I guess you can't blame any academic for presenting their research in the most somber fashion when anything else leaves them open to attack. Let's not forget when Senator Tom Coburn attacked the NSF and accused scientists of spending half-a-mil on a shrimp treadmill .