New study reveals how the brain learns to associate smells with behaviors

Houston, TX - Jan. 17, 2026 - A new study by investigators at the Jan and Dan Duncan Neurological Research Institute (Duncan NRI) at Texas Children's Hospital and and Baylor College of Medicine (Baylor) and their colleagues, sheds light on how sensory inputs are translated by the brain into meaningful behavioral cues.  

The study, led by Benjamin Arenkiel, Principal Investigator at the Duncan NRI and Professor in the departments of Molecular and Human Genetics and Neuroscience at Baylor College of Medicine, focused on glutamatergic (vGlut2BF) projection neurons in mice from the basal forebrain (BF), a region known to play essential roles in motivation, arousal and sensory regulation, and found a previously unidentified function for these neurons.  Lead author Pey-Shyuan Chin, a graduate student at the Duncan NRI and Baylor and her colleagues discovered that though they initially respond weakly to odors before learning, responsiveness of vGlut2BF neurons markedly increases after the animals undergo associative learning (e.g. learning that bad smells should be avoided).  After a period of time, the collective activity of the neurons begins to differentiate between conditioned orders, displaying the emergence of value-based sensory representations.  

Taking their experiment one step further, the researchers then either stimulated or inhibited vGlut2BF neurons while exposing mice to either "good" or "bad" odors and they found that they could alter the animals' odor preferences.  Stimulation paired with an odor could make that smell preferred whereas inhibition of the neurons when paired with an odor would make the animals avoid it.  

The investigators posit that the BF acts as a central hub linking sensory processing, learning and behavior as this exciting work offers new insights into learning, memory and motivational states and expands our understanding of how experience shapes perception.  These findings also open new avenues of investigation for treating disorders that involve maladaptive associative learning like anxiety or addiction.  

####

Other contributors to this work include:  Zhuokun Ding, Mikhail Kochukov, Snigdha Srivastava, Elizabeth H. Moss, and Qingchun Tong.

This work was supported by DoD award PR22135 to BRA and QT, R01DK109934 to BRA and QT, R01DK138518 to BRA, USDA project # 58-3092-5-008 to BRA, and P50HD103555.