On inherited memories

Sometimes there is such an ‘obvious’ way for a science story to be told that it more or less tells itself. Scientists themselves are (allegedly) taught, however, not to believe what they are told—nullius in verba, as the Royal Society puts it—and so in such cases it behoves us to look askance at the reporting.

One recent publication has provoked endless variations on a headline at the same time both tidy and astonishing: “We can inherit fears”. “Scared of spiders? It could be genetic”, the Daily Mail advised. “Phobias may not be simply down to unpleasant child [sic] experiences, but those of your parents or even grandparents.” The Telegraph agreed: “Phobias may be memories passed down in genes from ancestors.” But it’s not just the newspapers that were at it: the news site of Nature, under the umbrella of which the research was published (and to which I occasionally contribute), announced that “Fearful memories haunt mouse descendants.”

But isn’t this all a bit anti-Darwinian, specifically a bit Lamarckian—the inheritance of acquired characteristics? A bit like Jean-Baptiste Lamarck’s idea in the early nineteenth century that giraffes get their long necks from the stretching of their parents’? Well yes indeed, and that’s because the idea here is that these “memories” are being passed from parent to progeny by epigenetics: chemical alterations to the parents’ DNA, changing the way genes are activated, which are retained in the germ cells that produce the next generation. Epigenetic modifications to genomes are triggered partly by environmental factors, and are therefore ‘acquired’. There are many claims of this kind of non-Darwinian inheritance doing the rounds, although the extent to which epigenetics really does produce lasting genomic changes across several generations is hotly disputed.

It means that the latest results, reported in Nature Neuroscience by behavioural scientists Brian Dias and Kerry Ressler of the Emory University School of Medicine in Atlanta, Georgia, are not a priori absurd or impossible. The question, however, is whether they mean what they have been said to mean.

The researchers report that they conditioned mice to associate the smell of a particular volatile organic substance with fear, by applying mild electric shocks to their feet when the odour was present. They then found that the progeny of these mice showed “increased behavioural sensitivity” to the conditioning odour, relative to other mice whose parents and grandparents hadn’t had this conditioning.

This is interesting and surprising. What’s more, both the conditioned mice and the first-generation offspring proved to have an epigenetic chemical modification of the gene that encodes a particular odorant receptor: one of the creatures’ “smell proteins”. And the brain anatomy of the offspring was found to be altered, relative to other mice.

So something seems to be going on here—it’s an intriguing paper. What it doesn’t show is that memories are passed on through the genes. That is simply not the same as showing a possible epigenetic change in neural wiring and sensitivity to a smell. There is certainly evidence that neuropsychiatric disorders such as phobias, anxiety and post-traumatic stress disorder can be passed from one generation to the next, and indeed Kessler is interested in why such conditions, as well as behaviours such as drug addiction, recur in parents and children. Of course, distinguishing cause and effect here is fraught with difficulty: if you see your mother freaking out when a spider appears, you don’t need some epigenetic fiddling in your genes to start behaving that way yourself. The sole study of human “inherited behavioural conditioning” that Dias and Ressler cite is one that Ressler co-authored, in which the children of African American mothers from a deprived neighbourhood had a greater propensity to be startled in the dark if their mothers had suffered high levels of physical abuse. Are you surprised by that?

I have also seen questions raised about the small number of mice involved in Dias and Ressler’s experiments (and thus their statistical significance), and about the detailed sequence in which the ‘startle tests’ were conducted – for example, whether the responses of later generations to ‘startle’ stimuli (loud noises) with and without the ‘fear odour’ can be adequately compared. But in any event, what is neither being seen nor indeed claimed is the ‘inheritance of memories’. It’s possible that some increased sensitivity to certain types of smell – a change in the proportions of different “smell proteins” – might be passed on from one generation to the next. What isn’t known, as Dias and Ressler clearly say in their paper, is that the offspring mice experience the trigger smell fearfully. There is no evidence here that anyone has inherited a phobia.