Prestin is an anion transporter dispensable for mechanical feedback amplification in Drosophila hearing
- 1.5k Downloads
In mammals, the membrane-based protein Prestin confers unique electromotile properties to cochlear outer hair cells, which contribute to the cochlear amplifier. Like mammals, the ears of insects, such as those of Drosophila melanogaster, mechanically amplify sound stimuli and have also been reported to express Prestin homologs. To determine whether the D. melanogaster Prestin homolog (dpres) is required for auditory amplification, we generated and analyzed dpres mutant flies. We found that dpres is robustly expressed in the fly’s antennal ear. However, dpres mutant flies show normal auditory nerve responses, and intact non-linear amplification. Thus we conclude that, in D. melanogaster, auditory amplification is independent of Prestin. This finding resonates with prior phylogenetic analyses, which suggest that the derived motor function of mammalian Prestin replaced, or amended, an ancestral transport function. Indeed, we show that dpres encodes a functional anion transporter. Interestingly, the acquired new motor function in the phylogenetic lineage leading to birds and mammals coincides with loss of the mechanotransducer channel NompC (=TRPN1), which has been shown to be required for auditory amplification in flies. The advent of Prestin (or loss of NompC, respectively) may thus mark an evolutionary transition from a transducer-based to a Prestin-based mechanism of auditory amplification.
KeywordsPrestin Mechanical amplification of sound Auditory amplification Drosophila hearing NompC
Fast Fourier transform
Inner hair cell
Outer hair cell
Polymerase chain reaction
Hearing is a physiologically delicate process. The sensitivity of vertebrate ears results from the intricate and unique molecular transduction machinery that resides in the hair bundles of their sensory hair cells. Within the bundles, the interplay between force-gated transducer channels and adaptation motors generates forces that amplify sound-induced bundle movements and thereby boost vertebrate auditory performance while in mammals, an additional amplification mechanism is provided by a Prestin-mediated electromotile activity (Hudspeth 2008). Understanding the molecular, and mechanistic, details of auditory transduction is essential for understanding the bases of hearing and the causes of deafness.
In mammals, the exquisite sensitivity and frequency selectivity is achieved through an active amplification of sound-evoked vibrations of the basilar membrane (BM) within the cochlea. BM vibrations deflect specialized microvilli, or stereocilia, of inner ear hair cells and thereby directly open mechano-electric transducer channels (Fettiplace and Hackney 2006; Peng et al. 2011). Two types of auditory hair cells exist, both of which harbor mechanosensory transducer channels: (1) inner hair cells (IHCs) represent the ‘sensory arm’ of cochlear function and send the auditory information, created by the BM vibrations, to the brain; (2) outer hair cells (OHCs) represent the ‘motor arm’ of cochlear function and respond to transduction-induced changes in membrane potential by contractions and expansions of their cell bodies, thereby amplifying the BM vibrations. The molecular motor responsible for this electromotility of OHCs is Prestin, a member of the SLC26 anion transporter family (Dallos and Fakler 2002). Loss of Prestin in mice raises the threshold at which they respond to sound, and a mutation in Prestin may cause hereditary hearing loss in humans (Liberman et al. 2002; Liu et al. 2003).
Like vertebrate hair cells, the auditory neurons of the fruit fly D. melanogaster are endowed with an active, force-generating process that involves mechanically gated transducer channels and adaptation motors acting in concert to boost hearing (Göpfert et al. 2005; Nadrowski et al. 2008). The macroscopic performance of the entire D. melanogaster ear is governed by the properties of these active transducer modules, and the fly’s external antennal sound receiver can be used to directly probe auditory transducer function in vivo in both wild-type and mutant flies (Albert et al. 2007; Kamikouchi et al. 2010).
D. melanogaster responds to sound through the Johnston’s organ (JO), which is made up of an array of ciliated neurons in function analogous to the IHCs of the mammalian cochlea (Eberl 1999). The motor mechanism by which sound-induced vibrations are actively amplified by the JO is unknown but the origin of this amplification is attributable to JO neurons (Göpfert et al. 2005). JO does not have OHCs to amplify vibrations. Instead, the dendritic cilia of JO neurons may contract to boost the vibratory response to faint sounds (Göpfert and Robert 2002; Kernan 2007; Lee et al. 2008). A transducer-based, quantitative model, which assumes that (as yet molecularly unknown) motor proteins act in series with the actual transducer channels, was found to explain the mechanical and electrical responses of the fly’s ear to small disturbances (Nadrowski et al. 2008). An alternative hypothesis as to how JO neurons might amplify sound is through a Prestin-based electromotility. In this hypothesis, the D. melanogaster homolog Prestin (dpres) may contribute to ciliary motility by causing contraction in the membranes of the JO neurons.
The possibility that dpres contributes to transducer-based active amplification of the sound-induced vibrations has escaped rigorous molecular evolutionary analysis of Prestin orthologs in representative species which suggested that Prestin-based electromotility exists only in mammals (Okoruwa et al. 2008; Tan et al. 2011). In fact, specific residues have been identified as being essential for Prestin’s evolution from an anion transporter to a protein with electromotility responsible for OHC electromotility necessary for sound amplification (Schaechinger et al. 2011; Tan et al. 2011). New research, however, questions the time at which this electromotility evolved in metazoans. While an early study in chickens revealed no Prestin-based electromotility (He et al. 2003), a more recent study showed Prestin-like signatures in chicken short hair cell stiffness (orthologous function to mammalian outer hair cells), active amplification and frequency tuning of the auditory organ (Beurg et al. 2013). This suggests that the amplifying function of Prestin could be more highly conserved than previously supposed.
The goal of this study was to determine whether dpres is necessary for non-linear mechanical amplification in the hearing organ of D. melanogaster. This was accomplished by examining the expression pattern of dpres, and by analyzing the phenotypes of newly generated dpres loss-of-function mutants. The results demonstrate that dpres is expressed in JO neurons yet is not required for auditory amplification in D. melanogaster. Instead, dpres function may be restricted to the role as anion antiporter, consistent with previous reports (Hirata et al. 2012a, b).
The D. melanogaster strains carrying w 1118 , dpres-Gal4, and dpres mutants (engineered from homologous recombination and P-element mobilization described below) were used in the analysis. All flies were raised at 25 °C and 60 % humidity.
The 5′ regulatory sequences of dpres were cloned into the pPTGal vector (Sharma et al. 2002) and injected into w 1118 embryos to produce a transgenic line. The resulting line was crossed to a UAS-EGFP reporter to determine the dpres expression pattern.
The construct was injected into fly embryos to make a transgenic line that was then crossed to FLP, SceI for targeted mutagenesis. Transgenic larvae were heat-shocked at 37 °C for 30 min to produce a circularized DNA element that recombined with dpres. Stable lines were produced from the progeny and screened using PCR to identify lines with a disrupted dpres gene. Line 339 was confirmed as a dpres mutant by long-template PCR and therefore renamed dpres 339 .
A P-element transposase line [Dr, P(Δ2–3)] was crossed to the P-element insertion lines KG05103 and EY04256 to generate EY04256 Dr, P(Δ2–3/KG05103 dysgenic flies, which were then crossed to CxD/TM3. Offspring were collected to create stable lines that were screened by PCR to identify imprecise excisions that removed dpres. Of the 101 balanced lines isolated, 92 were homozygous viable. Using PCR on 72 viable or lethal lines, we determined that one viable line, 51C, was a dpres deficiency spanning from the KG05103 insertion site past exon two (Fig. 1).
RT-PCR primers were designed against the dpres gene. PCR was performed on mRNA from late-stage embryos, larvae and adults to determine expression pattern of dpres mRNA from controls and homologous recombinants.
Electrophysiology was performed on the antenna with methods described by Eberl et al. (2000). Briefly, the fly was immobilized in a pipette tip trimmed to allow the head to protrude. The pulse song, a component of the fly’s courtship song, was played on a loudspeaker and delivered to the fly through Tygon tubing to preserve near-field sound properties. The sound stimulus intensity was measured at 5.3 mm/s at the position of the antennae, using a calibrated Emkay NR3158 particle velocity microphone (Knowles). One electrolytically sharpened tungsten electrode was inserted in the junction between the first and second antennal segments while a second electrode penetrated the dorsal head cuticle as a reference. The signals were subtracted with a DAM50 differential amplifier (World Precision Instruments) and then digitized and normalized using a virtual instrument designed with Superscope 3.0 (G. W. Instruments) software. Responses to ten presentations of the stimulus were averaged, and the peak amplitudes of sound-evoked potentials (SEPs) were measured.
Confocal microscopy was performed using a Zeiss LSM510 microscope. Whole mount flies were prepared by immersing in 50 % glycerol and mounting on depression slides. Antennae were visualized by dissection and staining with rhodamine phalloidin.
Laser Doppler vibrometry
Adult D. melanogaster aged 1–5 days after eclosion were mounted on a Teflon rod and immobilized using dental glue. A laser Doppler vibrometer (LDV; PSV-400 with an OFV-70 close-up unit and a DD-5000 displacement decoder, Polytec GmbH, Waldbronn, Germany) was used to record the oscillations of the arista. Signals were generated and recorded in the PSV 8.6 Scanning Vibrometer Software (GmbH, Waldbronn, Germany) and later processed in Sigmaplot (Systat Software Inc) to extract best frequencies and energy gain (more details in Göpfert et al. 2005). Mutant and control data were tested for normality (Shapiro–Wilk) and equal variance (rejection thresholds set to p < 0.05 in both cases) prior to further statistical analysis. All data presented here passed both the normality and equal variance test and were analyzed using a two-tailed t test.
Anion transport measurements
A dpres-GFP fusion construct (in pEGFP-N2 vector) was transfected into CHO cells using JetPEI transfection reagent (Polyplus). Patch-clamp recordings of electrogenic transport currents were performed as described (Schaechinger and Oliver 2007; Schaechinger et al. 2011). Briefly, cells with unequivocal membrane fluorescence were selected (24–48 h after transfection). Whole-cell patch-clamp recordings were carried out at room temperature (20–22 °C) with an EPC10 amplifier (HEKA electronics, Lambrecht, Germany) controlled by Patchmaster software (HEKA). Electrogenic anion transport was measured as the ionic transport current in response to command voltage ramps (−100 to +100 mV; 300 ms). Background currents recorded before application of extracellular transport substrates were subtracted from current traces. For comparison between cells, current densities were calculated by normalizing to whole-cell capacitance as a measure of membrane area.
Patch electrodes were filled with intracellular solution containing 160 mM CsCl, 1 mM HEPES, 1 mM K2EGTA, adjusted to pH 7.3 with KOH. Extracellular transport substrates were locally applied to the cells via a gravity-fed application capillary. Extracellular solution contained 10 mM NaCl, 150 mM Na-gluconate, 10 mM HEPES, and 2 mM Mg-gluconate, adjusted to pH 7.4 with NaOH. For extracellular application of divalent anions, 10 mM SO4 or oxalate substituted for chloride. For application of bicarbonate, 25 mM HCO3 − was exchanged for equimolar Cl− and the solution was equilibrated with 5 % CO2.
dpres is expressed in auditory neurons
The D. melanogaster gene CG5845 encodes dpres (Fig. 1a) (Weber et al. 2003). dpres is located at polytene chromosome position 75A9 on chromosome arm 3L (Fig. 1a). The predicted gene product is a 742 amino acid protein with sulfate transporter domains that overall shares 18 % identity and 31 % similarity with mammalian Prestin (Fig. S1).
If dpres were functionally orthologous to human Prestin, acting as a piezo crystal-like motor that augments the ears’ mechanical responses to sound, we would expect dpres to be expressed in the neurons of JO, and possibly other chordotonal organs. Therefore, we first characterized the dpres expression pattern.
dpres mutants have no effect on SEPs
To reconcile the lack of electrophysiological phenotype in the homologous recombinant allele dpres 339 , we wondered whether the truncated proteins, up to 250 residues, from the aberrant transcripts could be partially functional. Therefore, we wanted to generate a deletion that completely removes the dpres coding sequences. To this end, we took two approaches. We first made use of the w hs element inserted into the homologous recombinant allele, providing w+ eye color, as a substrate for gamma-ray mutagenesis to select for w− derivatives. Two w− derivatives recovered were found to have cytologically visible deletions in the 75AB region of polytene chromosomes, and to form a lethal complementation group with Df(3L)BSC8. Because these deletions in all combinations removed other vital genes in addition to dpres, they were not useful for testing dpres function. In the second approach, we made use of two P-element insertions near dpres to generate imprecise excisions. To allow the possibility of recovering a synthetic deletion between these two elements by P-induced recombination, we generated flies containing both P-elements and the P-element transposase. From this we recovered dpres 51C , an imprecise excision of the KG05103 element with a deletion extending from the KG05103 insertion site into the dpres gene, thus removing the entire upstream regulatory region and the first two exons. As with dpres 339 , we found that SEPs were unaffected in the dpres 51C allele (Fig. 3).
Mechanical amplification of sound is independent of dpres function
The role of dpres in mechanical amplification of sound
Of the nine mammalian members of the SLC26 family, only Prestin is known to confer electromotility when transfected into mammalian cells (Zheng et al. 2000; Oliver et al. 2001; Lohi et al. 2002). Recent structural and functional analysis of Prestin suggests that the molecular mechanism underlying electromotility was derived evolutionarily from a primordial anion transport mechanism. Thus the protein domains that mediate anion transport, including a central anion binding site, are also critically involved in the electromotile function (Schaechinger et al. 2011; Gorbunov et al. 2014). Yet, by generating chimeric molecules between mammalian mechanically active Prestin and a non-mammalian transporter ortholog, it was shown that in principle, Prestin can mediate both anion transport and electromechanical activity at the same time (Schaechinger et al. 2011). It is thought that the movement of a voltage sensor, most likely related to the interaction of anions with the protein, changes the area occupied by the Prestin molecule within the plane of the membrane (Oliver et al. 2001; Dallos and Fakler 2002; Song and Santos-Sacchi 2013). When many Prestin molecules are clustered in the baso-lateral membrane and change shape at the same time, the outer hair cell length changes (Dallos et al. 1993; Iwasa 2001). An increase in cilia membrane tension in response to decreases in sound intensity would be an analogous function in D. melanogaster. We found evidence for specific dpres expression in the auditory sensory neurons. However, our study with flies lacking dpres found that dpres is dispensable for fly hearing.
The anion antiporter activity of dpres
The primordial function of mammalian Prestin as an anion transporter, rather than a motor, is further supported by Albert et al. (2007) who show that the zebrafish Prestin ortholog does not possess the same electromotility as mammalian Prestin, along with altered kinetics and voltage dependence. In fact, non-mammalian Prestin orthologs are electrogenic anion transporters that exchange monovalent versus divalent anions (Schaechinger and Oliver 2007; Deng et al. 2013). The explanation for lack of a dpres role in D.melanogaster hearing is supported by the studies of Okoruwa et al. (2008), who surveyed the electromotility of Prestin homologs among vertebrates, by looking at protein sequence conservation. They found that the electromotility appears to be a derived character that is distinctly mammalian, especially in eutherian mammals. They argue that two insertion/deletion events and subsequent evolution conferred fast electromotility in the transmembrane anion transporter domain of Prestin. While this divergence to electromotility in mammalian Prestin was thought at the time to coincide with anatomical innovations in the outer hair cells and surrounding cochlear tissue of eutherian lineage, namely the reduction in number of rows of IHCs and in OHCs and changes in the tectorial membrane structure, the inference of chick Prestin’s role in short hair cell amplification (Beurg et al. 2013) suggests that electromotile properties may have predated the mammalian anatomical arrangement.
Because the dpres gene has been functionally retained in the D. melanogaster genome, is expressed in sense organs at all major stages of its life cycle, and its product mediates robust anion transport (Fig. 5) (Hirata et al. 2012a, b), we expect that dpres acts as an anion transporter in the normal function of the neuron. If that is the case, one might still expect to find that dpres mutants have reduced hearing due to defects in functional neurons. A possible explanation for the lack of a quantifiable dpres mutant phenotype is the presence of other proteins that may serve a redundant function to dpres in the chordotonal organ. A search of genes differentially expressed in the JO (Senthilan et al. 2012) identified ten solute carriers, though none closely related to dpres. The possibility remains, though, that one of these transporter genes serves a redundant function throughout the mechanosensory tissues in which it is expressed and can thus compensate for the loss of dpres function, or that functionally equivalent solute carriers replacing dpres are not selectively expressed in the mechanosensory organ. Alternatively, it is possible that functional consequences of dpres mutation in the JO can only be revealed under certain conditions that differ from the standard laboratory environment, such as at the fringes of tolerable temperature or under conditions of cell stress.
Are acquisition of Prestin electromotility and loss of nompC coeval?
Prestin’s derived cochlear amplifier function meanwhile has been reported to have evolved in both avian and mammalian lineages—lineages that do not contain nompC orthologs, while its sole (ancestral) function as an anion transporter is present in those species that have retained NompC as a functional mechanotransduction channel (Fig. 6). The association of Prestin’s derived function with loss of the nompC gene suggests an interesting hypothesis in which the non-linear sound amplification function of NompC has been replaced by Prestin. This replacement may be facilitated by the fact that nompC mutations in D. melanogaster reduce, but do not eliminate, auditory function (Eberl et al. 2000; Effertz et al. 2011). We are excited to see further functional and evolutionary analysis on both Prestin and nompC to test if this is an accurate reflection of the evolution of auditory sound amplification. Key progress in confirming an association between Prestin electromotility and loss of NompC will be made by analysis of reptiles, among which NompC homologs can be found in turtles and alligators, but so far not in lizards (Fig. 6). Reports of active amplification mechanisms in reptiles (Stewart and Hudspeth 2000) have yet to be linked to, or distinguished from, Prestin-based mechanisms (Fig. 6).
We would like to thank the University of Iowa Central Microscopy Research Facility for use of the Zeiss LSM510 confocal microscope (Grant Number 1 S10 RR12916-01). Dominic Evans performed the laser vibrometry on dpres mutants. Katie Bunce conducted the P-element mobilization experiments that led to the 51C mutation. Matthew Topping provided constructive criticism during manuscript preparation. Fly stocks were obtained from the Bloomington Drosophila Stock Center at Indiana University and P-element constructs were originally obtained from the Drosophila Genome Research Center. This work was supported by NIH RO1 DC004848 to Daniel F. Eberl, a Ruth L. Kirschstein NRSA Post-Doctoral Research Fellowship DC006550 to Janice Fritz and grants from the HFSP (RGY0070/2011), BBSRC (BB/L02084X/1) to Joerg Albert, and grants from the DFG (GO 1092/2, SPP1608 and OL 240/4-1, SPP1608) to Martin C. Göpfert and Dominik Oliver. This work was also facilitated by the Iowa Center for Molecular Auditory Neuroscience supported by NIH P30 core grant DC010362 to Steven Green. All applicable international, national, and/or institutional guidelines for the care and use of animals were followed.
- Weber T, Göpfert MC, Winter H, Zimmermann U, Kohler H, Meier A, Hendrich O, Rohbock K, Robert D, Knipper M (2003) Expression of prestin-homologous solute carrier (SLC26) in auditory organs of nonmammalian vertebrates and insects. Proc Natl Acad Sci USA 100:7690–7695PubMedCentralPubMedCrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.