Harvard University - Department of Molecular & Cellular Biology


by Hayan Yoon and Catherine Dulac

November 18th, 2005

To ensure reproductive success, animals have evolved sensory and behavioral strategies to identify suitable mating partners. A particularly widepread strategy among mammals is the release and the detection of chemical cues called pheromones. Pheromones carry species- and gender-specific information and elicit innate and stereotyped changes in behavior and in hormone release among conspecifics: a male mouse detecting female cues will initiate courtship behaviors, while it will attack a male intruder. Neuroscientists and ethologists have long tried to identify the brain circuits that govern genetically preprogrammed behaviors. In a set of experiments published in the November 17 issue of Cell (Yoon et al., 2005), we have identified a complex neuronal network controlling reproduction and fertility in mice. Remarkably, in contrast to established notions on the nature of sensory inputs modulating reproduction, our study uncovered a major olfactory component of these circuits.

To study the chemosensory modulation of reproductive behaviors, we have traced afferent pathways to neurons in the hypothalamus and basal forebrain synthesizing luteinizing hormone releasing hormone (LHRH, also known as gonadotropin hormone releasing hormone, GnRH), a key neurohormone of reproduction. According to all texbooks, and based on classical tracing studies with lipohilic dyes, the pheromonal modulation of reproductive behavior is thought to rely on the neuronal connections between LHRH neurons and the vomeronasal organ (VNO). The VNO is an olfactory structure of the nasal cavity specialized in the detection of pheromones and distinct from the main olfactory epithelium (MOE), which detects odorants.

Surprisingly, however, we identified major projections from the main olfactory epithelium and the primary olfactory cortex to LHRH neurons but did not detect any connection from vomeronasal structures. In contrast to earlier studies, our tracing specifically targeted LHRH neurons and their inputs thanks to the use of genetically modified pseudorabies viruses, which are dependent on CRE-recombinase expression to replicate and express the green fluorescent protein (GFP). Pseudorabies viruses have the ability to jump across synapses in a retrograde manner (from postsynaptic to presynaptic neurons), thus revealing interconnected sets of neurons from the initial site of infection. We generated a transgenic mouse line expressing CRE-recombinase under the control of the LHRH promoter and injected the conditional pseudovirus into the hypothalamus of transgenic animals. We then performed a systematic survey of GFP-positive neurons across the brain resulting from the transynaptic spread of the virus from the original replication in LHRH neurons. This led us to discover a major neural pathway connecting a very circumscribed population of olfactory receptor neurons in the MOE to LHRH neurons in the hypothalamus controlling reproduction and fertility. This major sensory input from olfactory neurons had never been described or even suspected.

This unexpected finding was further supported by the functional and behavioral analysis of the connection between olfactory neurons and mating centers in the brain. Male mice with defects in odorant detection by either genetic mutation or chemical ablation of olfactory neurons showed little interest in females and displayed greatly reduced investigatory or mating behavior. In contrast, mice lacking a functional vomeronasal organ displayed normal mating behavior.

Our results support the idea that the pheromones required for the control of mating behavior in mice are recognized by the main olfactory system and not by the vomeronasal system. This research requires us to rethink how mammals detect pheromones, and to significantly revise the classical but simplistic division of olfaction, according to which the main olfactory system governs cognitively based behavior while the vomeronasal system controls hardwired and genetically preprogrammed pheromone-triggered behaviors.

Our results are particularly interesting to consider in the context of animal species thought to be devoid of a functional vomeronasal system – such as higher primates or humans – but displaying intriguing pheromonal-like responses that could thus be sustained through the main sense of smell.

Further reading: Harvard Gazette, HHMI Newsletter

View Catherine Dulac's Faculty Profile