Olfaction in the malaria mosquito Anopheles gambiae : electrophysiology and identification of kairomones

J. Meijerink

Research output: Thesisinternal PhD, WU


<p>Female mosquitoes of the species <em>Anopheles gambiae</em> Giles <em>sensu stricto</em> are important vectors of human malaria in Africa. It is generally assumed that they locate their human host by odours. These odours are detected by olfactory receptor neurons situated within cuticular extensions on the antenna. These cuticular extensions, called sensilla, contain numerous pores through which the odours can enter the sensillum and reach the olfactory receptor neuron membrane. Despite the fact that these mosquitoes are so important for the transmission of malaria, hardly any sensory studies have been performed to date. Therefore, the goal of this study was to analyze the response spectra and characteristics of the olfactory receptor neurons encoding human-derived odours in female <em>An. gambiae</em> . Another goal of this study was the identification of human odours which guide female <em>An. gambiae</em> to its host. This was accomplished by making chemical analyses of the odour profile of human sweat.<p>Firstly, a scanning electron microscopic (SEM) study was undertaken to identify the different types of sensilla exhibited on the antennae of female <em>An. gambiae</em> . Chapter 2 shows SEM photographs of the six different types of antennal sensilla: large and small sensilla coeloconica, grooved peg sensilla, sensilla trichodea, sensilla ampullacea and sensilla chaetica. The distribution of the different sensilla on the thirteen segments is tentatively described. Odours present on the human skin or identified in the headspace of human sweat evoked responses of grooved peg- and sensilla trichodea receptor neurons (chapter 4 and 6). Although the grooved peg sensillum is easily recognized during SEM studies and light microscopic observations, the different subtypes of sensilla trichodea are hard to distinguish in <em>An. gambiae</em> . SEM photographs of two different types of sensilla trichodea housing receptor neurons responsive to sweat-borne components are shown in chapter 2.<p>The antennal olfactory responses of female <em>An. gambiae</em> were studied by means of electroantennography (chapter 3). The electroantennogram (EAG) is considered to be the summed activity of all (or a part of the) responsive sensory receptor neurons on the antenna. Initially, the technical aspects of the EAG recording technique were closely examined. It was found that when using tungsten electrodes, artefactual electrode potentials were generated by the carboxylic acids, propionic acid, butyric acid and hexanoic acid. No artefactual electrode potentials were obtained with glass electrodes. A blend of carboxylic acids has been reported to be attractive for female <em>An. gambiae</em> mosquitoes. These are present on the human skin where they display an enormous diversity in chemical structure. To quantify the antennal olfactory sensitivity to carboxylic acids, EAG studies were conducted with glass electrodes. Carboxylic acids with carbon chain lengths of 5-8 (C <sub>5</sub> -C <sub>8</sub> )elicited high EAG amplitudes, while lower responses were evoked by the less volatile acids, C <sub>9</sub> -C <sub>14</sub> . Hexanoic acid evoked the highest EAG response.<p>Single sensillum studies were undertaken to reveal antennal olfactory receptor neurons responsive to carboxylic acids (chapter 4). Neurons innervating one or two of the morphologically different subtypes sensilla trichodea were found to respond to the short chain carboxylic acids: acetic acid (C <sub>2</sub> ), propionic acid (C <sub>3</sub> ), butyric acid (C <sub>4</sub> ), iso-butyric acid (iC <sub>4</sub> ) and iso-valeric acid (iC <sub>5</sub> ). Usually the receptor neurons responded by inhibition, but receptor neurons were also found responding by excitation to the short chain carboxylic acids. Occasionally receptor neurons were found which responded by excitation to 1-octen-3-ol. Dose-response characteristics were assessed for the carboxylic acid-inhibited cell type. It was demonstrated that the carboxylic acid-inhibited neuron was equally sensitive to the short chain acids tested. This was revealed by making corrections for the differences in volatility of the different short chain acids. It is suggested that in this case an inhibitory response might function to block the response of a specialised cell normally responding by excitation to other stimuli (chapter 7).<p>Because behavioural studies indicated that in addition to carboxylic acids other components are involved in the host-seeking behaviour of female <em>An. gambiae</em> , studies were undertaken to identify new putative attractants (chapter 5). Rather than searching for minor differences in chemical odour profiles between different individuals, we focussed on components which are generally produced by every human host. Freshly collected pooled sweat samples obtained after physical exercise from a group of volunteers neither attracted female <em>An. gambiae</em> in a windtunnel bioassay, nor evoked a detectable EAG response. Incubation of the sweat samples, however, resulted in a behaviourally attractive source of volatiles which evoked reproducible EAG responses. Sweat obtained during physical exercise is most likely to originate from the eccrine and sebaceous glands. Several observations strongly indicated that during incubation attractants are produced by microbial activity. Although the fresh sweat did not elicit a behavioural or EAG response, it was possible that it may contain components which acted as synergists at the behavioural level together with components produced during incubation. Headspace analysis of the fresh and incubated sweat revealed that geranyl acetone (5%-6%) and 6-methyl-5-hepten-2-one (1.8%-1.9%) were relatively abundant in both the fresh and incubated headspace samples. Headspace samples of the incubated sweat comprised large amounts of indole (27.9%), 1-dodecanol (22.4%) and 3-methyl-1-butanol (10%). These components were absent or only present in minor amounts in the headspace samples of the fresh sweat. Indole, geranyl acetone and 6-methyl-5-hepten-2-one evoked an EAG response, while 1-dodecanol did not elicit any response. 3-Methyl-1-butanol was only tested at the single cell level (chapter 6).<p>In order to reveal olfactory receptor neurons responsive to the identified sweat-borne components, studies were conducted at the single cell level (chapter 6). For other mosquito species, such as <em>Aedes aegypti</em> , the grooved peg sensillum is considered to house receptor neurons sensitive to host odours. Indeed, incubated sweat elicited excitation of a receptor neuron innervating a subpopulation of the grooved peg sensilla in <em>An. gambiae</em> . The same receptor neuron was excited by ammonia, which was found to be present in large amounts in the incubated sweat. This strongly implies that ammonia is causing (most of) the attractiveness of the incubated sweat.<p>However, chemically identified components from the headspace of the fresh and incubated sweat (chapter 5) did not elicit responses of grooved peg receptor neurons. They evoked excitation of receptor neurons associated with the sensilla trichodea. Two different subpopulations were found, one was innervated by receptor neurons sensitive to indole, while the other subpopulation housed receptor neurons sensitive to geranyl acetone. 3-Methyl-1-butanol and 6-methyl-5-hepten-2-one elicited responses of receptor neurons associated with both subpopulations. Receptor neurons displayed lower sensitivity to 3-methyl-1-butanol and 6-methyl-5-hepten-2-one. Sensilla trichodea receptor neurons only occasionally responded to the source of the components, the incubated sweat.<p>It is suggested that not only the grooved peg receptor neurons encode host-odours but also the sensilla trichodea receptor neurons fulfill a function. The incubated sweat may evoke an increase in the spike frequency of only a few spikes per second and responses like these are very hard to detect. Another explanation might be that sensilla trichodea receptor neurons encode other behavioural activities, such as nectar feeding or location of oviposition sites. Identified sweat-borne components are very likely of microbial origin. Microbial products are very common in nature and therefore not restricted to human emanations (chapter 6).<p>Further studies on the behavioural level might elucidate the role of the identified sweat-borne components for the behaviour of <em>An. gambiae</em> . The function of the different olfactory receptor neurons in host-seeking <em>An. gambiae</em> is further discussed in chapter 7.<br/>
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • van Lenteren, Joop, Promotor
  • van Loon, Joop, Promotor
Award date26 Oct 1999
Place of PublicationS.l.
Print ISBNs9789058081179
Publication statusPublished - 1999


  • anopheles gambiae
  • smell
  • kairomones
  • electrophysiology
  • identification


Dive into the research topics of 'Olfaction in the malaria mosquito Anopheles gambiae : electrophysiology and identification of kairomones'. Together they form a unique fingerprint.

Cite this