News & Views | Published:


Up, down, flying around

Nature volume 458, pages 156157 (12 March 2009) | Download Citation

The Johnston's hearing organ of the fruitfly has newly discovered sensitivities to gravity and wind. As in our inner ear, different sensory signals from this organ travel in parallel to separate zones in the brain.

The hearing organ in the antennae of mosquitoes and flies — first described by Johnston1 — detects sounds from nearby sources such as the vibrating wing of the courting male fruitfly. In this issue, Kamikouchi et al.2 and Yorozu et al.3 dramatically expand the number of known effective stimuli for Johnston's organ to include gravity and wind. These results elevate Johnston's organ from a near-field hearing sensor to a complex ear in which specialized clusters of neurons have distinct mechanosensory responsibilities.

Johnston's organ is the largest of the chordotonal organs that are located beneath the epidermal layer of the insect skin. Chordotonal organs are assemblies of units each of which consists of a ciliated neuron and several associated support cells4; Johnston's organ in the fruitfly contains hundreds of such units. Nearby sounds vibrate feathery structures (aristae) on the antennae, stretching cilia on Johnston's organ neurons (Fig. 1) and thereby exciting them to make sensory signals. The molecular mechanism of transduction is obscure, but sensory-neuron signals can be visualized by measuring increased levels of intracellular calcium ions, which universally accompany neuronal excitation. Kamikouchi et al. and Yorozu et al. therefore genetically engineered Johnston's organ neurons to express molecules that fluoresce when calcium-ion concentration rises. They then compared the calcium-ion signals evoked by sound, wind, gravity or movement of the aristae, looking at neurons either in Johnston's organ1 or in the brain's antennal mechanosensory centre2.

Figure 1: Johnston's organ.
Figure 1

a, In the fruitfly, each antenna on the head is made of three segments (1–3). b, Johnston's organ (JO) — which consists of hundreds of arrayed neurons, each with a sensory cilium that receives stimuli — is located beneath the epidermis covering segment 2 (A2) and functions by detecting rotation of segment 3 (A3) relative to A2. Near-field sound produces minute high-frequency deflections of the feathery arista connected to A3. The resulting signals project through a distinct subset of Johnston's organ neurons to zones A and B of the brain's antennal mechanosensory centre. It emerges2,3 that gravity and wind stimuli, which produce somewhat larger, low-frequency or sustained arista motions, are also detected by Johnston's organ neurons, albeit by different subsets of these cells, which project to zones C and E of the antennal mechanosensory centre. (Graphic modified from ref. 13.) Image: J. SCOTT & R. BHATNAGAR, DEPT BIOL. SCI., UNIV. ALBERTA

The need to know which way is up is considered so pressing that it has been assumed that flies are sensitive to gravity. These insects lack an obvious statolith — crystal aggregates that weight gravity-sensing receptors in organisms as diverse as plants, octopuses and humans. In the absence of statoliths, chordotonal organs on the flies' legs5 and Johnston's organ6 were considered likely candidates for gravity sensing.

Kamikouchi et al.2 (page 165) show that it is Johnston's organ that is responsible for gravity sensing in flies. They find that removing aristae severely decreases the tendency of flies to walk upwards (the direction opposite to the gravity vector). Their test was not designed to rule out minor contributions from neurons outside Johnston's organ to the overall sensing of gravity. Nevertheless, it clearly shows that flies detect a change in their orientation relative to gravity through the effect of gravity on the position of the aristae. The authors' calculations, taking into account an arista's apparent mass and stiffness, suggest that gravitational changes maximally deflect an arista by about one micrometre — a movement 100 times greater than that evoked by sound at the threshold of hearing7.

A previous study8, and now that of Yorozu and colleagues3 (page 201), also used simple antennal manipulations to investigate the role of Johnston's organ in sensitivity to wind. The earlier paper8 showed that gentle air currents such as those experienced in flight modify flight behaviour, an effect that is abolished by removal of the aristae. Yorozu et al. find that, in response to air currents moving at about 2 metres per second, fruitflies 'freeze' in place — a behaviour whimsically dubbed WISL, for wind-induced suppression of locomotion, and which also depends on the presence of the arista. The benefits of hunkering down in the wind may range from a short-term individual advantage of staying close to familiar terrain and food sources to a longer-term group benefit of controlling the dispersal of fly populations.

Whereas courtship song vibrates the aristae at frequencies of tens to hundreds of hertz, wind and gravity stimuli change more slowly and may be sustained. Both teams2,3 tested the importance of this difference for stimulus discrimination within Johnston's organ by using probes to displace the arista at different frequencies or durations. They find that fast and slow stimuli differentially activate neuronal groups that originate from distinct zones in the Johnston's organ array and target specific zones (A–E) in the antennal mechanosensory centre. Sound-sensitive neurons of zones A and B responded briskly to arista vibrations or to the onset of position changes; when new positions were sustained, however, their response adapted by declining. Wind- and gravity-sensitive neurons of zones C and E, by contrast, responded poorly to vibrations and relatively slowly to position changes, but showed robust, sustained responses to long-lasting stimuli.

These observations indicate that the main difference between sound-sensitive and wind- or gravity-sensitive neurons might be in the speed at which their transduction cascades respond and adapt. This difference could reflect quantitative or qualitative differences in the transduction pathways of the neuronal groups involved. Indeed, Kamikouchi et al.2 identify a qualitative difference, in that only the sound-sensing neurons express a particular stretch-activated ion channel called NompC (ref. 5).

But how do fruitflies distinguish wind from gravity? These two kinds of input are also likely to differ intrinsically from each other in duration and size — although the differences are probably more subtle than those from sound stimuli — and so may activate different subsets of neurons projecting to zones C and E. For example, gravity sensing may require the most sensitive neurons. Alternatively, wind and gravity may segregate perceptually through association with distinctive visual and proprioceptive cues arising from, say, upward versus forward locomotion.

Compartmentalization of Johnston's organ neurons into groups with distinct mechanical sensitivities is reminiscent of the vertebrate inner ear. In the ear, hearing receptor cells also show clear adaptations for fast responses relative to gravity-sensitive receptor cells9,10. This similarity extends the list of recently discovered parallels between fruitfly and vertebrate ears, including related genes for specifying ear development11, the involvement of particular classes of ion channels3, and evidence for mechanisms that amplify the mechanical input to the sensor7,12. Such similarities excite the speculation that hearing organs in fruitflies and vertebrates arose from a sensory structure present in a common ancestor11, rather than independently as was long thought.


  1. 1.

    J. Microsc. Sci. 3, 97–102 (1855).

  2. 2.

    et al. Nature 458, 165–171 (2009).

  3. 3.

    et al. Nature 458, 201–205 (2009).

  4. 4.

    , & Microsc. Res. Tech. 63, 388–399 (2004).

  5. 5.

    Pflügers Arch. 454, 703–720 (2007).

  6. 6.

    , & J. Comp. Neurol. 501, 756–764 (2007).

  7. 7.

    & Proc. R. Soc. Lond. B 268, 333–339 (2001).

  8. 8.

    , & J. Exp. Biol. 210, 4092–4103 (2007).

  9. 9.

    et al. J. Neurosci. 25, 7831–7839 (2005).

  10. 10.

    & J. Neurophysiol. 90, 2676–2689 (2003).

  11. 11.

    Dev. Dyn. 232, 550–558 (2005).

  12. 12.

    Neuron 59, 530–545 (2008).

  13. 13.

    & Nature Rev. Neurosci. 8, 510–521 (2007).

Download references

Author information


  1. Ruth Anne Eatock is at the Eaton-Peabody Laboratory, Department of Otology and Laryngology, Harvard Medical School and Massachusetts Eye and Ear Infirmary, Boston, Massachusetts 02114, USA.

    • Ruth Anne Eatock


  1. Search for Ruth Anne Eatock in:

About this article

Publication history




By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing