A boost for hearing in mosquitoes

It is hard to imagine ears more different from a human's than a mosquito's. Look at a mosquito head-on, and you'll see a pair of large, wraparound eyes, and projecting from the front of its face is a pair of long, hairy antennae—its ears. Each antenna arises from a large, spheroidal base, the Johnston's organ (JO). Function follows form. Whereas human ears are pressure detectors, a mosquito's detects the particle velocity component of a sound field, which is restricted to the immediate vicinity of the sound source in acoustic near field. The mosquito's ears are insensitive to pressure fluctuations in the acoustic far field. The antenna, with its fine, interlacing network of flagellar hairs, senses movements of air particles as they are swept to and fro by impinging acoustic waves. The antennal movements excite the sensory receptors contained within the JOs to generate action potentials that flood into the insect's brain. The bottom line of functional hearing is acuity—maximize sensitivity and sharpen the tuning of each receptor cell. This tuning goes on at the level of cell and molecular biophysics—and when it comes down to fundamentals, functional similarities are often more apparent than differences, even in animals as diverse as humans and mosquitoes. As Jackson and Robert (1) cleverly demonstrate in this issue of PNAS, nonlinearities dominate the biomechanics of hearing that enhance acuity. Similar nonlinearities have been known in mammalian ears for several decades, where they also enhance acuity.

The functional imperatives of any hearing organ—of mice and men or mosquito—are to detect nanoscale movements of air particles at threshold and retain sensitivity over a millionfold dynamic range of intensity. We know a lot about the macroscopic workings of hearing from more than a half-century of inquiry into the workings of the mammalian ear, as we delve down from a macroscopic view of hearing from the eardrums to the middle ear bones and cochlear mechanics. The microscopic workings of hearing occur within the cochlea at the hair cell receptors, and here we reach today's research frontiers. The hair cells of the mammalian cochlea are arrayed in three long, orderly columns along the basilar membrane. Georg von Bekesy (2) won a Nobel Prize in 1961 for demonstrating that traveling waves are propagated through the cochlear fluid, with the peak of the traveling wave arriving at a particular place on the basilar membrane that depends on the frequency of stimulating tone. Thus, frequency is place-coded along the length of the basilar membrane; specific frequencies excite hair cells corresponding to their location on the membrane. The basic idea of place-coding still stands, but the canonical story has been extensively revised and elucidated at the level of cellular mechanism. These recent findings have altered our view of hearing from that of a basically passive process to an active process involving hair cells acting like molecular micromachines to augment auditory sensitivity and tuning. Are such dynamic properties of sensation confined to the vertebrate ear, a testament to the needs of “higher” organisms? Apparently not. Jackson and Robert demonstrate that the adaptive imperatives of functional hearing have resulted in similarities in biomechanical processes that affect vertebrate and invertebrate animals alike.

Male mosquitoes detect the presence of females on the wing by ear.

The paradigm shift in our view of auditory processing is realizing that hearing is not simply a passive process but an active one. These active processes are manifested only in live animals. This peculiar caveat is important, because von Bekesy's founding studies of cochlear function were with organs from human cadavers. In recent decades, the ability to culture whole organs, cochlear tissue, and even isolated hair cells in vitro, mostly in vertebrate and mammalian animal models, has permitted penetrating molecular and biophysical studies of hair-cell function. These studies show that there are metabolically dependent processes within hair cells that alter their sensitivity to sound level and spectral acuity and that these processes are nonlinear in their action. These findings resonated with Jackson and Robert. What they accomplished is remarkable because they demonstrate (1) nonlinear auditory processes in intact live insects, without recourse to dissecting of the animal or macerating its tissues for molecular analysis. Instead, they made painstaking measurements of the mosquito's antennal/JO movements in response to stimulation by the sounds of real mosquitoes, or simulated mosquito sounds, through a loudspeaker. The investigators applied the technique of Doppler laser vibrometry, which, by means of measuring modulations in coherent light due to Doppler-shifted object motions in the light path, is completely noninvasive and perfectly applicable for intact, lively mosquitoes. By using this powerful technique, Jackson and Robert (1) make a strong case for the presence of nonlinear response properties in the auditory periphery of male mosquitoes. They have combined their instrumental analysis with some nice physical modeling in the classic tradition of fruitful collaborations between physicists and biologists in hearing science. As biologists, they interpret their findings in terms of adaptive or evolutionary rationales for the function of active hearing in male mosquitoes.

How does active acoustics facilitate mating behavior, which was known to rely on acoustic mate finding (3)? Male mosquitoes detect the presence of females on the wing by ear. Males hear a nearby female's flight tones, ≈400 Hz, and go off in pursuit. The male's auditory system is selectively tuned to frequencies of ≈300–400 Hz, as shown long ago (2). This process seems simple enough, but there's more to it. The very small wings of a mosquito (2–3 mm) are very inefficient radiators of their 300- to 400-Hz flight tones because they are very small compared with the wavelength of a 400-Hz tone (800 mm), as shown by the acoustical analysis of Bennet-Clark (4). For the male mosquito, this means that unless a female is flying very nearby, a distance of millimeters to a few tens of centimeters away, he should not hear her (4). Now enter Jackson and Robert.

Their experiments (1) show that the mechanical response of the male antennae of the Tanzanian mosquito Toxorhynchites brevipalpis exhibits a markedly nonlinear response, decomposable into three distinctly different levels of mechanical excitability or dynamics. This complex response in his hearing is triggered when the male detects the “song” of a female flying by. The authors used synthetic models of fly-by songs that they constructed from recordings of both freely flying and tethered flying females. The first stage, G-1, is a brief amplification that is triggered by an intensity level corresponding to a male–female distance of ≈19 mm. This stage is immediately followed by a second, compressive phase if the female remains “locked in” range, but when her song wanes (she is “flying by” the male in these experiments), a second stage of gain, G-2, kicks in that keeps sensory acuity temporally elevated. Importantly, this elevation persists beyond the stimulus itself for more than a second, maintaining a heightened sensitivity, which the authors assert is hysteresis in the gain function (1). Thus, once a female is detected by the male, G-1 can be reinstated, and if the female is within 2 cm of the male, he can home in on the song/female; if she flies past, however, the antennal gain continues to be (re)enforced by G-2 and would, by the authors' calculations, be at heightened sensitivity by a separation distance for at least another 5 mm, beyond the gain from the initial amplification at G-1. It's rather like the lingering scent of perfume, which briefly persists even after its owner has passed by.

These distances may seem trivial at human scale, but considering that mosquitoes are very small and that the intensity of the acoustic signal in the particle velocity field falls off as the inverse cube of distance between sender and receiver (not as the inverse square that predicts sound attenuation in the pressure field, as is the case in human hearing), the male's ability to hear a female would be expected to drop off steeply with separation distance. Moreover, the particle-velocity field contains directional information, so courtship in mosquitoes on the wing is clearly a game of millimeters.

Is this nonlinear auditory tuning in mosquitoes due to active processes as it is in vertebrates? The authors (1) point out that if the mosquitoes are dead, none of the three nonlinear phases can be evoked by acoustic stimulation. The metabolic dependence of nonlinear processes in mosquito hearing, which they surmise may have origins with the sensory receptor cells themselves, was addressed by Goepfert and Robert (5, 6). Importantly, they proved that sounds could be generated from the antenna itself, recording its vibrations by Doppler laser vibrometry, occurring either spontaneously (no acoustic stimulation) or evoked (with brief bouts of stimulation). This situation is analogous to the spontaneous and evoked otoacoustic emissions David Kemp reported in 1979 (7) that catalyzed and inspired the active-hearing movement among mammalian auditory scientists. The work on the active mechanics of mosquito hearing is formally similar to vertebrate and mammalian hearing, at the microscopic level of receptor cells and possibly even micromechanical mechanisms. The pivotal role of metabolically dependent nonlinear processes in hearing in such taxonomically disparate taxa as insects and mammals is intriguing. It may not be altogether surprising when viewed from an evolutionary perspective; vertebrate hair cells and the scolopidial receptors of mosquitoes are both of ciliary origin.

If we zoom back out to the macroscopic organism level, it is worth examining the function of active hearing. The authors interpret their findings in terms of how active hearing maximizes the search strategies a male might use to find a female (figure 5 in ref. 1). However, a flying male also must deal with his own flight tone, which could differ from a female's. Any acoustic mechanism that would make a female's tone “stand out” from his would be beneficial, and indeed, most animals have mechanisms for selective attention (SA) that serve to alert (momentarily heighten perceptual acuity) an animal to the presence of a salient stimulus, such as social signals, predatory warning sounds, environmental noises, etc. In human hearing, SA is engaged in the “mixture party effect,” where, even in a noisy, crowded room of colleagues, overhearing one's own name can shift all of one's auditory attention to zoom in on that conversation. In mammals, SA is brain-centered, not ear-centered. If the active process in mosquito hearing is thought of as a form of SA, then it may have an ear-centered SA. It is an intriguing thought that perhaps this ear-centered SA might also initiate processes in the male mosquito's brain that could engage central processes further reinforcing its acousticomotor systems to purse a just-overheard female.

Is this ascribing too much behavioral plasticity to a creature that lives for a few weeks in adulthood and whose behavioral repertoire seems to be limited to finding a female to mate with (males) and, after mating, for the female to bite a warm mammal and suck its blood to help make eggs? A recent paper by Gibson and Russell (8) suggests that the mating ritual of mosquitoes is not simply cruise, catch, and mate. Gibson and Russell also studied acoustic behavior in T. brevipalpis, but their focus was on what happens when a male actually catches up with a female. Science has focused on the hearing of male mosquitoes, but Gibson and Russell (8) listened to both sexes, especially during courtship. The flight tones of males and females do not differ significantly (≈420 Hz), although wing-beat frequency varies between 300 and 500 Hz. Their remarkable finding is that when a male and female mosquito, tethered and “flying,” are brought within earshot of each other, both sexes bring their flight tones into near-perfect convergence, with an accuracy that extends out to the first four harmonics and within a reaction time of ≈6 seconds. Same-sex interactions do not result in convergent flight tones for either sex; instead, their tones pointedly diverge. Presumably, this behavior could serve as a way for each to assess the fitness of the other as a reproductive partner or could simply be a way to avoid same-sex courtship. In any case, there is ample room for speculation about active hearing in light of these findings.

Finally, this discussion is about mosquitoes, and as interesting as might be the desiderata of their aural sex, what comes mostly to mind when we see or hear one is to swat it dead. These studies were done on a nonbiting mosquito, T. brevipalpis, but the deservedly infamous ones, including the malarial mosquito Anopheles gambiae and the yellow fever mosquito Aedes aegypti, warrant similar investigation. Although it remains to be seen whether these studies will result in devices that make our backyards safe from pesky and potentially health-endangering mosquitoes, their findings make us draw back and marvel at the biomechanical ingenuity of nature's sensory systems in all creatures, great or small.