Robert B. Darnell (left) and Erich D. Jarvis (right). (Credit: Lori Chertoff/The Rockefeller University)
Whether humans are singular among animals in their use of complex language remains a hot topic among scientists in many fields. Rockefeller University researchers Robert B. Darnell and Erich D. Jarvis recently made a discovery that could change the terms of that debate: The alteration of a single amino acid in a single gene may have contributed to the evolution of the more complex vocal communication found in human spoken language.
That this variant (dubbed I197V) is unique to humans was known, as was its location in an RNA-binding protein in the brain called NOVA1-a protein essential for normal development in virtually all animals studied. Why humans have their own unique change in the gene, however, was not. To find out, the researchers placed I197V in mice using CRISPR-and surprisingly, the animals began to vocalize differently between themselves with more complex vocalizations than wild-type mice used.
Intriguingly, the researchers also roughly determined when the variant may have emerged. Deep dives into the genomes of our archaic human cousins, the Neanderthals of Europe and Denisovans of Central Asia, confirmed that they lacked the variant, and a review of more than 650,000 modern human genomes documented its presence in all but six people. (Because the rare six persons are de-identified, we don't know if their speech is affected.) This suggests that the human unique change appeared in Africa sometime after we evolutionarily split from our ancient cousins but before we left the continent-tantalizing clues that this NOVA1 change may be an unexpected human-specific, language-associated gene.
We spoke to Darnell and Jarvis about what makes NOVA1 different from other language-associated genes and the possible clinical implications of their discovery.
How did you come to study the NOVA1 human variant?
Darnell: In my lab, we've been researching NOVA1 for more than 30 years. It's one of the neuron-specific RNA binding proteins that we specialize in; we study the links between these proteins and cognitive function and disease. Decades ago, my colleagues and I discovered that NOVA1 was the target of a neurological autoimmune disorder called POMA that causes severe motor dysfunction. More recently, we described the first patient missing one copy of NOVA1-a child with language and motor dysfunctions. These studies inspired us to explore why humans have their own unique form of the protein.
Jarvis: Bob approached me with an interest in working together on the project because my lab has expertise on human-specific genetic changes associated with spoken language in and complex vocal learning among non-human species, such as songbirds. The project was in line with one of our long-term goals-the neuroengineering of more advanced vocal communication behavior and brain circuits in mice.
The gene FOXP2 has long been investigated for its function in spoken language and associated disorders. How does NOVA1 compare to it?
Jarvis: One difference is that the mutations we have in FOXP2 were once thought to be unique to humans, but they've since been found in other hominid species, like Neanderthals, and in some non-hominid species too, such as bats, dogs, and wolves. In contrast, the NOVA1 substitution we have is limited to Homo sapiens alone.
Darnell: Most researchers think Neanderthals communicated to some degree, but the jury is still out on how sophisticated their verbal communications were. So the fact that we have this singular variant unique to the earliest humans, and can link this to vocalization, suggests it might be connected to the evolution of complex language.
When you put this human variant into mice, how did it impact the animals?
Darnell: We were completely surprised to see that it affected binding sites in midbrain neurons that code for vocalization. That's when the first author-Yoko Tajima, from my lab-collaborated with César Vargas, from Erich's, to test the mice's behavior. They found that replacing the mouse gene with the human NOVA1 variant changed how the mice "spoke" to each other. Baby mice called to their mothers differently, and male mice looking to lure a female for mating tried to attract her attention with altered vocalizations.
Jarvis: We categorize the ultrasonic vocalizations the mice use to communicate with each other into four syllable types. Yoko and César found that when they analyzed these syllables, the human-variant mice produced different sequences than the wild-type mice did.
Would you call NOVA1 a "language gene," then?
Darnell: Potentially, but it's certainly not the only one. Many factors were involved in the emergence of language.
Jarvis: I would call the human variant a possible spoken language-associated genetic change. But the gene itself is not exclusively used for vocal communication or spoken language.
Where do you see this research going next?
Darnell: Our focus is understanding genetic mechanisms to help humans who have disease, so we're interested in the structure and function of neurons and their pathways contributing to that. For example, it's very common that if you have a dysfunction such as a stroke in Broca's area, which is the language generator in the frontal cortex, you can't get your words out-you can think them but can't say them clearly. So we're interested in trying to track down exactly what the pathways are in this process from thought to speech, in disorders ranging from developmental forms of autism to frontal dementia.
And then there's the potential to develop a better understanding of who we are and how we evolved. I'm not an expert in evolutionary biology, but Erich, of course, is well positioned to have great insights into the evolution of vocalization.
Jarvis: I think we need to do more behavioral testing on mice with genetic variants that are specific either to humans or to vocal learners to see if the mice gain a greater flexibility to learn new vocalizations. We could also try introducing the NOVA1 human variant into songbirds, with the prediction it might make them more advanced vocal learners. Although spoken language likely came about through a combination of genetic changes, our study shows that at least one genetic change uniquely found in humans can impact vocal communication in a non-human species.
