UK United Kingdom

Fake memory implanted in mice with a beam of light

If you’ve ever been frustrated by erratic memories, spare a thought for the mice involved in a study published in the journal Science. Researchers have been able to consistently create a “false memory…

Hazy recollection: I’m sure I buried some cheese here. Paul.J.Hurtado

If you’ve ever been frustrated by erratic memories, spare a thought for the mice involved in a study published in the journal Science. Researchers have been able to consistently create a “false memory”, making a mouse fearful of a place it has no reason to fear. The memory was implanted by shining blue light into the mouse’s brain, which triggered a carefully chosen group of neurons.

The researchers used optogenetics, a technique that allows precise control of brain circuits. The control is achieved by expressing proteins that act as switches in particular types of brain cell. These switches are channels that, when struck by a particular colour of light, allow charged particles into or out of the neurons, which will either activate or silence them.

Susumu Tonegawa of the Massachusetts Institute of Technology and his colleagues wanted to find out whether they could create a new, negative association by flipping the switch on an old, neutral memory while giving the mouse a negative experience. Would this lead the mouse to be scared of the old memory?

To find out, Tonegawa’s group needed to identify the scattered set of neurons storing the first memory and install an optogenetic on-switch. They figured out how to do that last year.

The neurons that record new information are located in a particular corner of the hippocampus, the coiled brain structure that we know is crucial for memory formation. That area, the dentate gyrus, can be targeted with a virus, which acts as a courier that delivers genes that encode the protein switch. The difficult part is to deliver the virus only to cells that are storing the memory of choice. Tonegawa and his colleagues found that they could target those neurons that are busier than usual.

Armed with this discovery, they installed the optogenetic trigger in the neurons that were especially busy while a mouse got to know a new environment (we’ll call that Place A). The next day, in a different environment, they gave the mouse small electric shocks while triggering the memory of Place A using light. After that, even though it never had a negative experience in Place A itself, the mouse froze when it was returned there.

In another experiment, mice were given the same memory-shock treatment and then offered a choice between Place A and somewhere else. The mice avoided Place A. A group of mice that had the same virus inserted into a different part of the hippocampus was unaffected and just as happy in Place A as anywhere else. The artificial fear specifically required an alteration of the dentate gyrus.

Admittedly, “Uh-oh, Place A!” is not on the same level as the elaborate, special effects-laden falsehoods that featured in Christopher Nolan’s Inception. We cannot script new memories. The best recent fictional analogy for these experiments is one for fans of the Hunger Games trilogy: in the “hijacking” process, the Capitol authorities render existing memories traumatic by pairing them with doses of hallucinogenic poison.

Of course, we will never know exactly what the mouse remembers. The artificial firing of those selected cells is unlikely to conjure the full experience of being back in Place A, though it is certainly enough to influence behaviour. The mice returned to Place A after the “hijacking” spent about a third of their time frozen; other mice only froze for about 10% of the time.

Hapless laboratory animals have been enduring this type of behavioural testing for nearly a century. It is no surprise that we can train a mouse to dislike a room. But to do so while the mouse is in another room entirely, by triggering a memory with light? Science like this was impossible a decade ago.

When Francis Crick, famous for revealing the structure of DNA, floated the notion of light-activated neurons in 1999 he called it “far-fetched”. Now, just over a decade later, we see new experiments every week that drive or silence different circuits with light, affecting cognition and mimicking or mending the processes underlying mental illness.

This paper offers a relatively modest advance on previous work from Tonegawa’s own team, and other studies in mice and fruit flies. But it illustrates the power of a technique that has the global neuroscience community enthralled.

Optogenetics has brought surprising technical advances and changed the way many neuroscientists work. For me, however, there remains inherent wonder in the fact that we can control brain cells - if not create memories - with light.

Sign in to Favourite

Join the conversation

6 Comments sorted by

  1. John Crest

    logged in via email

    Well, that explains all the alien abductions and anal probing then.

  2. Paul Weldon

    Research Fellow

    so, how long before I can have a fake memory of a trip to mars and the discovery of an alien power plant to create breathable atmosphere?

    Aside from that, what applications are envisaged for this kind of ability?

    1. Jonathan Webb

      DPhil candidate at University of Oxford

      In reply to Paul Weldon

      I don't think these findings have an immediate application, Paul. Human memories are a more complicated phenomenon! In fact, calling the above a 'false memory' is probably stretching the point a little; it's more of a negative association. ("I've got a bad feeling about this," as they say in the classics.)

      This sort of work is aimed at teasing apart how those associations are formed by the brain's cells and networks. Understanding how that process works would be fascinating, and it also might help us deal with memory related problems like PTSD.

    1. Rob Buttrose

      University of Melbourne

      In reply to Rob Buttrose

      There was a typo in my previous comment. I meant to write "Needless to say" or was it "Needlessly"? Both are true.

      It should be noted that , in the U.S., where this research was conducted, mice, along with rats and birds, are not even classified as animals for the purposes of the Animal Welfare Act. (the Helms Amendment). The ethical issues of harmful experimentation on these animals for any purpose scientists can dream up has been in effect legislated away.

    2. Luke Weston

      Physicist / electronic engineer

      In reply to Rob Buttrose

      Animals used in scientific research - even scientific research with a more obvious, direct, benefit for human and animal welfare that might be more apparent to you than this case - are almost invariably killed humanely at the end of such research even though it is not necessary to kill them because killing the animal is considered the most humane thing to to - to absolutely prevent any possibility of pain or suffering.

      Not killing the animals would be considered unacceptable, as this is not minimising the possibility of further pain or suffering.