A single protein may have helped shape the emergence of spoken language

The origins of human language remain mysterious. Are we the only animals truly capable of complex speech? Are Homo sapiens the only hominids who could give detailed directions to a far-off freshwater source or describe the nuanced purples and reds of a dramatic sunset?

Close relatives of ours such as the Neanderthals likely had anatomical features in the throat and ears that could have enabled the speaking and hearing of spoken language, and they share with us a variant of a gene linked to the ability to speak. And yet it is only in modern humans that we find expanded brain regions that are critical for language production and comprehension.

Now researchers from The Rockefeller University have unearthed intriguing genetic evidence: a protein variant found only in humans that may have helped shape the emergence of spoken language.

In a study published in Nature Communications, researchers in the lab of Rockefeller researcher Robert B. Darnell discovered that when they put this exclusively human variant of NOVA1—an RNA-binding protein in the brain known to be crucial to neural development—into mice, it altered their vocalizations as they called to each other. 

The study also confirmed that the variant is not found in either Neanderthals or Denisovans, archaic humans that our ancestors interbred with, as is evidenced by their genetic traces that remain in many human genomes today.

“This gene is part of a sweeping evolutionary change in early modern humans and hints at potential ancient origins of spoken language,” says Darnell, head of the Laboratory of Molecular Neuro-Oncology. “NOVA1 may be a bona fide human ‘language gene,’ though certainly it’s only one of many human-specific genetic changes.”

Three decades in the making

Anatomical adaptations of the vocal tract and intricate neural networks enable our language capabilities. But the genetics behind them isn’t well understood.

One theorized genetic language driver is FOXP2, which codes for a transcription factor involved in early brain development. People with mutations in this gene exhibit severe speech defects, including the inability to coordinate lip and mouth movements with sound. Humans have two amino acid substitutions in FOXP2 that aren’t found in other primates or mammals—but Neanderthals had them too, suggesting that the variant arose in an ancestor of both human lineages. But some findings on FOXP2 have been disputed, and its role in human language development remains unclear.

Now NOVA1 has arisen as a candidate. The gene produces a neuron-specific RNA binding protein key to brain development and neuromuscular control that was first cloned and characterized by Darnell in 1993. It’s found in virtually identical form across a wide swath of the biosphere, from mammals to birds—but not in humans. Instead, we have our own unique form characterized by a single change of an amino acid, from isoleucine to valine, at position 197 (I197V) in the protein chain.

I197V isn’t the only amino acid substitution that distinguishes modern humans from other organisms, points out first author Yoko Tajima, a postdoctoral associate in Darnell’s lab. Several of them may be integral to brain development. “Such changes may have played important roles in the acquisition of characteristics that have contributed to the emergence, expansion, and survival of Homo sapiens,” she says. 

A specialist in how RNA binding proteins modulate gene expression, Darnell has been researching NOVA1 since the early 1990s, when he and his colleagues first identified it as the trigger of a neurologic autoimmune disorder called POMA that can cause extreme motor dysfunction. Recently they have begun to identify cases in which NOVA1 genetic variants are associated with developmental language and motor difficulties. 

“Understanding NOVA1 has been a career-long effort for me,” he says.

The current study, led by Tajima, used CRISPR gene editing to replace the common NOVA1 protein found in mice with the human variant I197V. They then used advanced techniques such as cross-linking immunoprecipitation (CLIP) analysis, a method developed by Darnell, to identify the RNA binding sites of NOVA1 in the mouse midbrain.

The big reveal

The first notable discovery was that the human variant had no impact on RNA binding related to neural development or motor control. It operated exactly as the one it had replaced.

So what was it doing? The second significant finding gave them a hint: binding sites that were substantially affected by the human variant were located at genes that coded for RNAs related to vocalization.

“Moreover, many of these vocalization-related genes were also found to be binding targets of NOVA1, further suggesting the involvement of NOVA1 in vocalization,” says Tajima.

“We thought, wow. We did not expect that,” Darnell says. “It was one of those really surprising moments in science.”

Darnell’s lab then joined forces with Rockefeller’s Laboratory of Neurogenetics of Language, headed by Erich D. Jarvis, who studies the molecular and genetic mechanisms underlying vocal learning.

Altered communications

Over the next few years, the collaborators investigated the impact on vocalizations among mice of various ages in different contexts. They found altered vocal patterns among both pups of both sexes and adult males.

“All baby mice make ultrasonic squeaks to their moms, and language researchers categorize the varying squeaks as four ‘letters’—S, D, U, and M,” Darnell notes. “We found that when we ‘transliterated’ the squeaks made by mice with the human-specific I197V variant, they were different from those of the wild-type mice. Some of the ‘letters’ had changed.”

They found similar patterns when they studied the hopeful mating calls of male adult mice exposed to female adult mice in estrus. “They ‘talked’ differently to the female mice,” he says. “One can imagine how such changes in vocalization could have a profound impact on evolution.”

The human element

The potential influence of I197V on human evolution became their next focus. To confirm that it wasn’t found in our nearest human relatives—the Neanderthals, who largely lived in Europe, and the Denisovans, named after the central Asian cave where they were discovered—the researchers compared eight human genomes with three high-coverage Neanderthal genomes and one high-coverage Denisovan genome.

As expected, our archaic relatives—from whom we are thought to have split about 250,000-300,000 years ago—had the same NOVA1 protein as all non-human animals. 

They then combed through 650,058 modern human genomes in the dbSNP database, a catalog of short sequence variations drawn from people around the world. If an alternative to I197V existed, it would be found here.

Of those 650,058 people, all but six had the human variant. Those six had the archaic variant; because the samples are de-identified, details about them are unknown.

“Our data show that an ancestral population of modern humans in Africa evolved the human variant I197V, which then became dominant, perhaps because it conferred advantages related to vocal communication,” he suggests. “This population then left Africa and spread across the world.”

Disease and disorders

In the future, Darnell’s lab will investigate how NOVA1 regulates language function with an eye on language or developmental disorders.

“We believe that understanding these issues will provide important insights into how the brain operates during vocal communications—and how its misregulation leads to certain disorders,” says Tajima.

Its neural pathways may come into play, for example, when various disorders renders someone unable to speak. Perhaps it influences the development of nonverbal autism; NOVA1 is one of the many genes linked to autism spectrum disorder. And in 2023, the lab reported on a patient with a NOVA1 haploinsufficiency whose neurological symptoms included a speech delay.

Adds Darnell: “Our discovery could have clinical relevance in many ways, ranging from developmental disorders to neurodegenerative disease.”

---

---