Low-serotonin mice less choosy about sex of partners

By Ed Yong | March 23, 2011 2:05 pm

Would you prefer to have sex with males or females? It’s such a simple question, but a loaded one. Biologically, we know very little about what goes on behind such sexual choices, in humans or other animals. Socially, it’s a question that can provoke fierce debate, social stigma, and psychological anxiety. Into this minefield steps a new study from Yan Liu and Yun’ai Jiang at Beijing’s National Institute of Biological Sciences.

Here’s the simple version. They found that male mice lose their normal preference for female mice if they have low levels of serotonin – a hormone that carries signals in the brain. Instead, they try to mate with individuals of both sexes, in equal measure. Inject them with more serotonin, and you can restore their preference for females. The obvious conclusion is that serotonin affects the sexual preferences of mice. The obvious question is whether it also affects the sexual preferences of humans. And, as is often the case, it’s a little more complicated than that.

Serotonin is no stranger to sexual behaviour. With lots of serotonin around, male mice lose interest in sex. Their libidos fall, they lose erections and they fail to ejaculate. If you kill the neurons that respond to serotonin, you can reverse these effects and restore the rodents’ sex drives. The same thing applies to humans. This is why people who take SSRIs – antidepressant drugs that raise levels of serotonin – often face a variety of sexual problems. Liu and Jiang wanted to extend this work to see if serotonin can also influence an animal’s choice of sexual partner.

What they did

They worked with male mice that lacked a gene called Lmxb1, which they normally need to make serotonin. When another male was shoved into their cages, the low-serotonin males almost always tired to mount the intruders; they started after 8 minutes, and kept on trying (see graphs). By contrast, males with working copies of Lmxb1 (and normal serotonin levels) never tried to mount the intruders, even after half an hour.

When Liu and Jiang gave the serotonin-deficient mice a choice between male and female partners, they tried to mount both animals around 80% of the time. For comparison, normal mice try to mount the female around 60-80% of the time, and with the male around 20-30% (see graphs).

This shift in sexual preference – from males to both sexes– became apparent in other ways. Male mice often serenade females with high-pitched calls to entice them into sex; they only direct these “songs” at other males around 10% of the time. But mice without Lmxb1 serenade males around 60% of the time (although they still prefer to sing at females).

Typical male mice also spend twice as long sniffing straw bedding if it has been rubbed with the smell of female genitals rather than male genitals. Those without Lmxb1 spent just as little time on the scent of either gender and were slightly more likely to choose the male-scented straw (see graphs). They had lost the preference for female pheromones, even though they could still detect these (and other) smells .

Liu and Jiang repeated the same experiments on mice with a different genetic fault, in a gene called Tph2 that helps to produce serotonin in the brain. These mice were also short of serotonin, they were equally likely to mount mice of both sexes, and they didn’t prefer either male or female smells (see graphs).

Finally, Liu and Jiang found that they could shift these preferences by altering serotonin levels (see graphs). They injected normal adult mice with a chemical called pCPA that depletes serotonin. After three days, these males were more likely to mount other males. The duo also gave a boost of serotonin to mice with faulty copies of Tph2, by injecting them with the hormone’s precursors. Half an hour later, and their serotonin levels were back to normal. Their tendency to mount other males fell to typical low levels, and they developed a preference for female-scented bedding.

What this means for mice

To Liu and Jiang, the conclusion is clear: serotonin “is crucial for male sexual preference in mice.” But there is another possibility: a lack of serotonin could have just made for hornier mice that were happy to indiscriminately mount anything within range. After all, low serotonin levels are already linked to high sexual behaviour.

Liu and Jiang tried to control for this but it’s debatable how strongly they’ve made their case. For example, they say that when males were presented with female mice, they didn’t make more advances. But it’s not clear why you would expect them to, when males already show strong sexual behaviour towards females. Indeed, the supplementary figures show that normal mice mounted females 90% of the time, and those without Lmxb1 mounted them all of the time (see graphs). That may not have been a significant difference, but it’s the largest possible one!

Liu and Jiang also found that the mice without Tph2 showed higher sex drives when they were tested in the presence of males and females. Again, the supplemental figures are telling (see graphs). They suggest that mice with one or more broken copies of this gene are more amorous towards females as well as males, when given the choice. With less serotonin around, they mounted mice of both genders more often, more quickly, and for longer. This fits the idea that they were more generally aroused.

To complicate things further, serotonin isn’t all about sex – this hormone has subtle influences on everything from aggressive behaviour to social interactions. For example, mice show a wide variety of unusual behaviours if they don’t have the gene for the serotonin transporter protein, which recycles the hormone after it’s been used. They’re less active, they interact with other mice in a more positive way, and they’re less aggressive. However, Liu and Jiang did show that their mice spent just as much time next to strangers, whether they had Lmx1b or not.

What this means for humans

Do these results apply to humans? Liu and Jiang call this an “unavoidable question” but it’s not one they can currently answer. No other studies have specifically looked at whether serotonin affects human sexual preferences. Elaine Hull, who has worked on serotonin and sexual behavior, says, “In terms of possible applications to humans, this may have implications for bisexual behavior.” But she adds, “Much more information is needed to specify the brain areas involved, and possible developmental regulation of serotonin in those areas, before we can jump to the conclusion that serotonin is the factor that inhibits male-to-male attraction.”

What about SSRIs? These drugs boost serotonin levels and they are widely used. And despite plenty of documented side effects, there’s no evidence that they can change sexual orientation.  Liu and Jiang cite a study by Milton Wainberg, which they say showed that “SSRIs inhibited compulsive sexual behaviours in homosexual and bisexual men”. But Wainberg isn’t happy with this description.

His trial tested the effects of SSRIs in gay and bisexual men with compulsive sexual behaviours. The drugs did lower their libido, as well as reducing the frequency of solo sex acts like masturbation. But contrary to what Liu and Jiang write, all of the volunteers, whether they took SSRIs or placebo, showed less compulsive sexual behaviour. More importantly, even though their serotonin levels had gone up, none of the trial’s volunteers started having more heterosexual sex.

Results like this make it clear that we must be cautious before extrapolating the results of rodent studies into humans. Serotonin may be a common player in animal nervous systems but its effects can vary from species to species. For example, drugs that affect serotonin levels have very different effects on the sexual behaviour of rabbits and rats.

At most, the results in these studies can tell us something about the biology of sexual preference. In that regard, there does seem to be something in Liu and Jiang’s results, and certainly intriguing hints that are worth following up. Problems will only arise if (or perhaps, when) people try to apply the results to cultural debates.

Even describing the mice with human labels like “straight” or “bisexual” would be unwise. As Liu and Jiang showed, when mice with normal levels of serotonin are given a choice between males and females, they will mount the male at least 20% of the time. This, and the widespread nature of homosexual behaviour in animals, supports the idea that sexual preference is more of a continuum. As Wainberg says, the assumption that there’s a clear line between heterosexual and homosexual behaviour “is out of step with the field of sexuality research. It is not evidence of nature at work; rather, it is evidence of the historical forces that continue to shape our concept of relationships.”

The biggest worry is that anti-gay groups will seize upon these results to suggest that homosexuality or bisexuality can be ‘cured’ by giving people more serotonin. Such “conversion therapies” have a sordid history and they’re still around – earlier this week, religious organisation Exodus International came under fire for launching an iPhone app that provides “freedom from homosexuality”. There is a clear moral argument against this – sexual orientation is part of a person’s identity, no less in need of change than the colour of their hair, or whether they’re right- or left-handed.

Wainberg adds, “Some countries prosecute and penalize homosexuality. Some cultures deny its existence or condemn those with same sexual behaviour. Making interpretations beyond the findings of this study could feed those who want to eliminate homosexuality, or may push those who are afraid of their own homosexual feelings to seek ‘cures’ that may damage them and will not address their reality.”


Footnote: What do the scientists themselves make of the social implications of their work? I emailed Yi Rao, who led the study, with a set of questions covering both the social and technical sides of the paper. He replied simply with, “I believe that the scientific question we are addressing is broader than what the news media is focusing on now and time will tell what are the implications of our research.” When asked further about the scientific question he is trying to address, he added only, “The previous email will be my standard on-the-record reply.”

To some extent, I appreciate that jobbing scientists may be reticent to become embroiled in politically charged debates – I can understand that attitude, but I cannot support it. Science is not a hermetically sealed box. It’s mostly funded by public money and its results ripple through society. When your research has potential social implications, and when it can very easily be misinterpreted, it does no one any good to fling data from the ivory tower, while locking yourself in and drawing the curtains.

Reference: Liu, Jiang, Si, Kim, Chen & Rao. 2011. Molecular regulation of sexual preference revealed by genetic studies of 5-HTin the brains of male mice. http://dx.doi.org/10.1038/nature09822 (Warning: apparently, this paper is not going to be published to coincide with the embargo lift. No word yet on when it will be available)

More on serotonin:


Comments (18)

  1. Hi Ed, I don’t disagree with your stance regarding the role scientists should play in the interpretation of their research, but in this particular case, there is not much more for the authors to say. “Only time will tell what the implications are…” is another way of saying “we have no idea what this means right now.” The only way Rao’s statement could have been more responsible was to add “…and I certainly hope all Earthlings don’t attempt to draw crazy conclusions based on these data, which are only a first step in a long research process.”

    With regards to the interpretation, although your point is taken regarding a general increase in arousal, it is still clear that sexual preference is lost, since other manipulations that induce a general increase in male rat sexual arousal are not necessarily associated with a loss of gender preference. It actually seems you really want to better-understand the mechanism. So the question, in one simplified sense, is whether serotonin may be “gating” sexual preference by providing inhibitory signals in response to inappropriate cues (like male olfactory/pheromone stimuli for males,) or is more directly driving the circuitry dictating social dominance and interaction, which if disrupted could merge sexual and dominance mounting into one big mess. There are some indicators here, but more details will have to come out in the next installment.

  2. There have been reams of hormonal, physiological, anatomical, and biochemical studies of sexual orientation. No consistent differences have ever been found and it would take more than one rat study to overturn that. Extraordinary claims require extraordinary proof.

  3. I’m confused by your last little bit there, Ed. You’re angry that this researcher didn’t try to extrapolate his data to say something about Human social behavior? As a model system, mice have proven many times to not mirror humans both from a molecular and behavioral perspective. I think you got the answer you deserved. This scientist understands the limitations of his data and knows more research must be done to determine if his results have any implications in humans. He likely saw your questions about methods as an easy way to get your foot in the door while thinking that the main points you wanted to report on were the juicy conclusions that could erroneously be connected to human behavior.

    Give him a break. He’s probably already gotten 15 emails from the Daily Mail asking if this means low serotonin makes dudes gay…

  4. Katharine

    The Beeb is already reporting this as ‘SEXUAL PREFERENCE CHEMICAL FOUND!’

    Bloody BBC.

  5. @ Brian – You seem to be trying to tell me that extrapolating to humans is dodgy ground, which is odd because that’s exactly what I think and what I said in my post. I expected that Rao would have voiced similar caution and given the likely misrepresentation in the media, I thought it was important to let the lead researcher talk about what the results really mean and the limitations of the data. Even the bit that you wrote would’ve been fine: “As a model system, mice have proven many times to not mirror humans both from a molecular and behavioral perspective… more research must be done to determine if his results have any implications in humans”

    In such cases where misinterpretation is easy, I feel that scientists should take opportunities to push out the right message. I don’t think the answer that he gave did that. You, of course, may feel differently.

  6. Ed, I think the case is that he’s had a bunch of reporters email him asking him for quotes on subjects that his data do not support. Maybe you were at the tail end of his e-mail pile and he was fed up with reporters trying to game him. I’m not saying you intended on doing this, but I don’t have any objection to his response and I don’t think it warranted your “ivory tower” comment.

  7. I think that it is precisely because his funding comes from public sources that Rao was unwilling to say anything one way or the other. He knows which side his bread is buttered on. And this is a systemic problem with research today. The goal of too many a scientist is a bit like the goal of a telephone psychic — to make the research sound generic and interesting enough that people will get sucked into thinking that it supports whatever conclusions they want to make. I suspect that Rao fears that making comments on the public record that would shatter the illusion that this research can be used to prove that you can cure gay would be a recipe for losing his funding, even though he knows that it shows nothing of the sort.

  8. May I suggest that the low serotonin mouse loses its discriminatory capability and ruts at any opportunity just as the severe alcoholic no longer seeks out his beverage of choice but employs Thunderbird, rotgut whiskey or even methyl alcohol to satiate the craving? One’s innate preference, be it for claret, bourbon or Budweiser is fixed by other currently unknown determinants as is homo or heterosexuality.

    The combination of a dopamine and serotonin agonist or the precursors of same, levodopa and 5-hydroxytryptophan (5-HTP), can halt alcohol and cocaine human cravings quite rapidly (in minutes) and successfully. Please use link at my name above if interested in details.

  9. In some species, the mating fixed-action patterns are also used for aggressive encounters. In some cases, when a male mounts another male utilizing a typical mating pattern, this is actually a display of social dominance. However, in other species, a male mounting a male is actually homosexual behavior, evolved not to determine social hierarchy, but quite the opposite, to increase social coherence within the group (“making friends”). So perhaps all of this is more about social order – either using mating to assert dominance, or making love to make friends – and has nothing to do with sex or sexual orientation at all?

  10. Noah, just rescued your comment. Great points – actually “What is the mechanism?” was the first question I asked Rao, which added to my irritation over the response.

  11. holly

    I get tired of reading what purports to be science articles that descend into moralizing about scientific research’s possible social ramifications, especially when the “journalist” or the journalist’s editor doesn’t make clear whose conclusions, whose moralizing is being presented.

    For example, this paragraph:

    “The biggest worry is that anti-gay groups will seize upon these results to suggest that homosexuality or bisexuality can be ‘cured’ by giving people more serotonin. Such ‘conversion therapies’ have a sordid history and they’re still around – earlier this week, religious organisation Exodus International came under fire for launching an iPhone app that provides ‘freedom from homosexuality’.There is a clear moral argument against this – sexual orientation is part of a person’s identity, no less in need of change than the colour of their hair, or whether they’re right- or left-handed.”

    Whose position is this: that of the scientist referred to in the previous paragraph? Or that of the journalist?

    If the reporter is ascribing this to the researcher, that isn’t made clear, is it?

    And, if it is the reseacher’s position, then I submit that the reporter has a journalistic duty to find someone to present the other side of the argument–that it’s no one’s damn business that a scientist should decide for me or for anyone else whether there’s a “clear moral argument” against any damn thing. For instance, what if an adult homosexual himself or herself wishes, given the chance, to change orientations? Is the idiot who said that above suggesting he or she ought not be given that choice?

    The scientific search for what determines heterosexuality and homosexuality is not a search for moral imperatives, and the minute a scientist/researcher starts thinking he or she can control what I or anyone else does with the results of any_ damn _study on any_damn_ thing is the minute we descend into the very world that scientist says he fears.

    Got that, Mr. Scientist?

  12. It’s my position. If I was ascribing it to the scientist, I would have quoted or made that clear, although I take your point that you could read that as a continuation of the previous paragraph.

  13. “…journalistic duty to find someone to present the other side of the argument…”? What ? He Said She Said journalism is now a “journalistic duty”? No surprise the public trust in media is so low if any journalists out there actually think that way….

  14. Well quite. If it’s false “balance” people are after, there are plenty of places to find it around the web. This isn’t going to be one of them.

  15. holly


    Thanks, then, for clarifying it was YOUR position.

    Researchers in the field of sexuality have to worry not only about sensationalistic headline writers (who always get it wrong) and science writers who rarely get it right, but also about interest groups that worry about findings. This affects not only their willingness to give interviews about their work but also their loquaciousness or lack of it regarding said work.

    The recent politics of sexual orientation has, unfortunately, had a deleterious effect on how many researchers chose to pursue grants for studies in this area. Ridiculous.

    Recall that whole PETA thing with the sheep? Ridiculous. That scientists might find out why rams or why men prefer females or males is all that counts. I don’t like the idea that others (you or a researcher) get to influence what science uncovers about *anything.* I will make choices for myself.

  16. they say>> Typical male mice also spend twice as long sniffing straw bedding if it has been rubbed with the smell of female genitals rather than male genitals. >>>
    ya know, I prefer clean bedding. and no, I really don’t want to know how they ran this particular test….

  17. It’s amazing the extrapolation made from very simple lab data: That sexuality is purely influenced by one neurotransmitter. The observation that low serotonin levels is related to sexual preference is interesting but only one piece of a much larger puzzle. I don’t think the author of the paper was implying anything more than that…but a lot of buttons were pressed.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Not Exactly Rocket Science

Dive into the awe-inspiring, beautiful and quirky world of science news with award-winning writer Ed Yong. No previous experience required.

See More

Collapse bottom bar