The first piece of the facial recognition puzzle

Humans have an incredible ability to recognize friends and family members from far away. This cognitive function helps us find the one familiar face hidden among hundreds of unfamiliar ones; it also informs us about their attention or emotional state of mind. Those with face blindness, however, must rely on alternative cues to recognize familiar individuals. In a new study, published in the June issue of GENETICS, Sun et al. apply classical genetic mapping strategies to identify the first gene involved in recognizing faces.

Face recognition relies on intricate neuronal processes that first identify individual elements of a face—blue eyes, heavy eyebrows, or small ears—before detecting the spatial relationship among these elements. Researchers have studied the neuronal circuits underlying facial recognition since the 1970s, and they have identified several brain regions that contain clusters of face-selective neurons. In contrast, our current understanding of the molecular players responsible for this complex cognitive behavior is non-existent.

Congenital prosopagnosia (CP) is a hereditary condition that causes face blindness. When neurobiologist Yi Rao, senior author on the study, learned of CP, he realized that identifying its genetic cause could help us begin to define the molecules that make up the facial recognition machinery.

Rao’s group started searching for individuals who struggle with identifying faces around 2007 but faced several hurdles along the way, the first of which was reliably identifying patients. “We thought it would be straightforward to check the phenotype versus the genotype. But it turns out [that the field] was using different questionnaires and behavioral assays. So, we had to come up with a standard ourselves,” Rao said. Along the way, his team learned that questionnaires were more reliable than most face recognition behavioral assays. “People have adopted different compensation strategies: they recognize people by their hair, by their voice, or their gestures. So, some of them can pass a face recognition assay. But if you really ask them, they will admit that they actually have problems recognizing faces in daily life,” Rao explains.

Once the researchers had developed a reliable patient phenotyping pipeline, they started screening individuals and were surprised to identify a large family with prosopagnosia among their own staff members on campus. They next performed linkage analysis, analyzing single nucleotide polymorphisms that segregate with CP to identify genomic regions responsible for the face blind phenotype. Combining the results of their linkage analysis with whole genome sequencing, Rao’s team identified a potential disease-causing mutation in MCTP2, a calcium-binding protein. They then screened an additional 3,000 people for face blindness, finding MCTP2 mutations in several other affected families.

Not all prosopagnosia families carry MCTP2 mutations, which means there are more genes involved in face recognition for scientists to discover. However, this study has now identified the first piece of the molecular machinery responsible for our ability to recognize familiar faces.

Some of the identified pathogenic mutations reside in a primate-specific isoform, causing Rao to suspect that this isoform acquired a novel function that unlocked face recognition. Identifying what precisely that function is will not be trivial. “The problem is that the mutation is not in the calcium-binding domain, so we don’t have a cellular or molecular assay to study [MCTP2] function,” says Rao. “We would have to rely on making knock-in mutations in monkeys [to study their consequence on gene function],” he adds.

Despite these difficulties, Rao is committed to continue this line of research, hoping to identify the precise nature and function of the molecular machinery that mediates this higher cognitive behavior. He plans on following the approach researchers took to understand the molecular machinery underlying circadian rhythm by piecing together the pathway one element at a time through in-depth genetic and molecular studies.

“We will have to keep working on it, and it is hard because you have to use other species (atypical animal models, like non-human primates), but I believe this is the first time we’ve been able to study [the genetics of] a higher cognitive function,” Rao concludes.