Sensory Systems/Insects/Olfactory System

While the human sensory system offers us stunning ways of perceiving our movement and environment, the sensory systems of insects and spiders are not any less fascinating. To give just a few examples, spiders have up to eight eyes, and some see almost as sharply as humans; bees "feel the rhythm" when other bees dance in the bee-hive, and learn from this the location of food sources; mosquitoes hunt their victims by smell. In addition, studies in insects have many fewer ethical or methodological limitations than studies in mammals. And especially in flies, with molecular genetic tools any gene can be targeted (e.g. knocked out or overexpressed), and the system is much more manageable than in humans.

The insect olfactory system

This sensory systems book is mostly about human sensory systems and there is a chapter about the olfactory system, so why do we need a chapter on the insect olfactory system? The fruit fly (drosophila melanogaster), which we will focus on here, is a very important model animal in biology and a lot of research on sensory systems is done in the fruit fly. The visual as well as the olfactory system are studied intensively and there are less ethical or methodological limitations. With molecular genetic tools, any gene in a fly can be targeted (e.g. knocked out or overexpressed) and the system is much more manageable than in humans. While the olfactory system functions quite different from the human’s, it is possible to find common principles. Furthermore, the insect olfactory system inspires engineering in robotics, medicine and many other areas.

The nature of smell

Different antenna types.

To understand the specifics of odor sensing one has to be aware that smell is quite different from other stimuli. It differs from light and sound by the fact that it is not carried by waves but by diffusion, air flows and turbulences. Furthermore, while light and sound only have the two perceptually relevant characteristics of frequency composition and amplitude, smell has a variety of discrete odorants and even more possible mixtures in different concentrations.

Overview

Drosophila melanogaster antenna under a light microscope. The olfactory sensilla can be seen (like hairs on the antenna). The antenna is fixated with a glass capillary and on the top right a recording electrode can be seen. The diameter of the antenna is about 90 µm.

Simplified schematic of a basiconic sensillum, like it can be found for example in the fruit fly.

In insects (but also in most vertebrates) the sensory system is of importance for orientation and food foraging but has also social (nest mate recognition e.g. in ants) and sexual (mating partner search and selection by pheromones) significance. The main path of the odor information begins at the olfactory sensilla (insect’s sensory organs that contain the sensory neurons) that can in most insects be found on the antennae and look like small hairs in the fly (see Figure). There exists a huge variety of antenna types (that are not only used for olfaction) and many different sensillum types.

To understand the general principle the example of drosophila melanogaster basiconic sensilla should suffice. The odorant molecules go through slits or pores of the cuticle into the aqueous sensillum lymph, where some types of odorant molecules are bound to odorant binding proteins and carried towards the dendrites of the olfactory receptor neurons (ORN), others diffuse in the lymph towards the dendrites. On the membrane of the dendrites there are odorant receptors (OR) that bind the odorant molecules and are responsible for the conversion of the signal into a membrane current. This current propagates through the dendrite to the cell body where (at the axon hill), an action potential is generated. The action potential travels in the ORN axon to the antennal lobe (which is analog to the olfactory bulb in vertebrates), where ORN make synapses to local interneurons and projection neurons. The antennal lobe is organized in so called glomeruli. It is not fully understood how they are involved in pattern recognition, but the glomerular activation pattern can provide information about the odors presented to the fly.

Projection neurons project into the lateral horn (where probably innate odor responses are processed) and to the Kenyon cells in the mushroom bodies. The mushroom bodies are a neuropil in the insect brain and have their name from the similarity to mushrooms. There, odors are associated with other sensory modalities and behavior which is why the mushroom bodies are an important model system to study learning and memory.

Odorant reception in ORN

response profile of selected drosophila odorant receptors to two odors. Data source: DoOR Database [1]

An insect’s olfactory sensillum contains one or more olfactory receptor neurons that transform the odor information into an electrical signal (action potential). Most ORNs contain only one receptor type, but each receptor type reacts to many odors (see Figure). However there are some receptors that are more specific as their detected odors have either an important role in the insect’s behavior or are chemically unique. An example for a more specific receptor is the receptor for CO2. Other odorants activate it only weakly and it directly triggers an avoidance response in the fly . Specificity of ORs is due to different affinity of odorants to the receptor. The higher the affinity the more receptors are occupied when an odor is applied and the stronger the current response. However, ORNs do not react linearly on stimulation. It can be assumed that most of them respond logarithmic within their working range which increases their dynamic range. The logarithmic relation does not apply to stimuli below the respective detection threshold and in saturation. Most ORN show a phasic-tonic on-response, i.e. they react with a strong increase in firing rate when a stimulus is presented and then show rate adaptation if the stimulus persists.

Odorant receptors are membrane proteins that elicit an ionic current through the membrane, when an odorant molecule binds to them. There are two ways how this is accomplished in the olfactory system: Metabotropic and ionotropic receptors. In mammals olfactory receptors are known to be metabotropic g-protein coupled receptors that release a g-protein into the cell that activates an ion channel via a short intracellular signal cascade . In contrast most olfactory receptors in insects are ionotropic receptors that are ion channels which open when an odorant molecule binds. Saving time on the signal cascade, ionotropic receptors are much faster than metabotropic receptors .

Odor information processing

The odor that reaches the antenna contains different parts of information: On the one hand the odor identity, i.e. which odor or mixture it is, on the other hand the quantity of the components. Furthermore, the timing of the stimulus contains information. If two odors start for example at the same time, it is probable, that they belong to the same object. It has been shown, that insects indeed use the odorant timing information not only to detect the direction of an odor source but also to distinguish and track odor objects . The processing mechanism that enables this behavior is not clear, but it is amazing that already stimulus onset delays in the range of a few milliseconds can be useful. There are recent results that suggest that the speed and the temporal resolution of the insect’s olfactory system are remarkable and much higher than expected.

By means of calcium imaging, a method that visualizes cytoplasmic calcium by a fluorescent marker and therefore activity in neurons, it is possible to create a functional atlas of the antennal lobe. Basically all ORNs of one receptor neuron type, containing one OR, project into one glomerulus. So the (in the fruit fly about 54) glomeruli each unite the response of one receptor type and form a spatial pattern. However it has been shown that also the temporal dynamics of the glomeruli response is used to encode information . The odor information is therefore encoded in the olfactory system by spatiotemporal firing patterns .

So far it has not been possible to disentangle and understand the odor code in a way that is comparable to the knowledge of the visual and the auditory system. This might be due to the peculiarities of smell discussed above. There is no easily mappable topographic organization of neurons as the odor space is multidimensional and not continuous.

Glomerular activity patterns are linked in the mushroom bodies to behavior. Most insects show a high amount of plasticity there and e.g. bees are able to associate odors with a food reward after only a few presentations. Odor Information in the mushroom bodies is said to be represented by a sparse code, which means that only few kenyon cells respond with only few spikes. In contrast the above described code in the antennal lobe and in the ORNs it is a combinatorial code.

Odor perception and behavioral significance

For insects the olfactory system is of great behavioral significance. For example, as we all have probably experienced first-hand, mosquitos can track their victims by smell. Ants follow pheromone traces to food sources, but are also able to identify their nestmates by a colony specific hydrocarbon profile (and are therefore able to eliminate foes and thieves when they enter their territory). And many moths use sex pheromones to find mating partners.

Usually odors in nature are not pure chemical substances but mixtures. However, those mixtures are perceived as a unit and are very often directly linked to a behavior. The neuronal response in the antennal lobe of a mixture cannot always be predicted by the response to the components . It should therefore not be taken for granted that the olfactory system works like an e-nose that is designed to analyze the components of the presented odor. Furthermore, compared to vision, where the information has to be processed deeply until the relevance of the content becomes obvious, the olfactory system is more strongly and directly linked to behavior (and at least in higher animals emotions). These connections are sometimes innate, but often also learned and idiosyncratic.