The study by scientists at Columbia University Medical Center (CUMC) found that variations in FTO gene indirectly affect the function of the primary cilium, a little-understood hair-like appendage on brain and other cells.
Specific abnormalities of cilium molecules, in turn, increase body weight, in some instances, by affecting the function of receptors for leptin, a hormone that suppresses appetite.
The findings, made in mice, suggest that it might be possible to modify obesity through interventions that alter the function of the cilium.
"If our findings are confirmed, they could explain how common genetic variants in the gene FTO affect human body weight and lead to obesity," said study leader Rudolph L Leibel, the Christopher J Murphy Memorial Professor of Diabetes Research, professor of pediatrics and medicine, and co-director of the Naomi Berrie Diabetes Center at CUMC.
Since 2007, researchers have known that common variants in the fat mass and obesity-associated protein gene, also known as FTO, are strongly associated with increased body weight in adults.
But it was not understood how alterations in FTO might contribute to obesity.
"Studies have shown that knocking out FTO in mice doesn't necessarily lead to obesity, and not all humans with FTO variants are obese," said Leibel.
"Something else is going on at this location that we were missing," Leibel added.
In experiments with mice, the CUMC team observed that as FTO expression increased or decreased, so did the expression of a nearby gene, RPGRIP1L.
RPGRIP1L is known to play a role in regulating the primary cilium.
"Aberrations in the cilium have been implicated in rare forms of obesity. But it wasn't clear how this structure might be involved in garden-variety obesity," said Leibel.
Leibel and his colleague, George Stratigopoulos, associate research scientist, hypothesised that common FTO variations in noncoding regions of the gene do not change its primary function, which is to produce an enzyme that modifies DNA and RNA.
Instead, they suspected that FTO variations indirectly affect the expression of RPGRIP1L.
Stratigopoulos created mice lacking one of their two RPGRIP1L genes, in effect, reducing but not eliminating the gene's function.
Mice with one copy of RPGRIP1L had a higher food intake, gained significantly more weight, and had a higher percentage of body fat than controls.
In a subsequent experiment, the CUMC team found that RPGRIP1L-deficient mice had impaired leptin signalling.
The study was published in the journal Cell Metabolism.