Otoacoustic Emissions and Auditory Foveae in Mole-Rats

Otoacoustic Emissions and Auditory Foveae in Mole-Rats Pim van Dijk1, Bert Maat1, Sabine Begall2, E. Pascal Malkemper3, Kai R. Caspar2,4, Geoffrey A. Manley5,a) 1 University of Groningen, University Medical Center Groningen, Department of Otorhinolaryngology/Head and Neck Surgery, Groningen, The Netherlands. 2 Department of General Zoology, University of Duisburg-Essen, Essen, Germany. 3 Max Planck Research Group Neurobiology of Magnetoreception, Max Planck Institute for Neurobiology of Behavior – caesar, Bonn, Germany. 4 Institute for Cell Biology, Heinrich Heine University, Düsseldorf, Germany. 5 Cochlear and Auditory Brainstem Physiology, Department of Neuroscience, School of Medicine and Health Sciences, Cluster of Excellence "Hearing4all", Carl von Ossietzky University Oldenburg, 26129 Oldenburg, Germany. a) Corresponding author: [email protected] Abstract. Auditory foveae are a rare phenomenon. To date, two kinds of auditory foveae have been described. The first kind, in constant-frequency bats, is accompanied by extremely sharp tuning in the narrow frequency range of the foveal region. The second kind, found in birds in the barn owl and in mammals in mole-rats, shows no increased tuning sharpness, and appears to be merely an expansion of the neural representation of an important range of frequencies. Studies of distortion-product otoacoustic emissions (DPOAE) in mole-rats have led to contradictory conclusions. Whereas Kössl et al. [1] found nothing unusual in DPOAE of Ansell’s mole-rat, Pyott et al. [2] reported an absence of DPOAE in both the related Damaraland mole-rat and the naked mole-rat. To clarify this discrepancy, and examine the effect of the auditory fovea as reported in neural mapping data for Ansell’s mole-rat by Müller et al. [3], we measured both DPOAE and stimulus-frequency emissions (SFOAE) in Ansell’s, the Damaraland, and naked mole-rats. Our data revealed robust emissions, especially in the region of the fovea. Despite relatively high auditory thresholds over the whole hearing range, when SFOAE or DPOAE fell within the foveal frequencies, their amplitudes were similar to those of other mammals that have much more sensitive hearing. This suggests that within the fovea, a larger number of hair cells is activated by the stimuli used, and indicates that other mole-rat species may also have foveae in a similar frequency range. DPOAE amplitude and phase characteristics were different between the foveal and non-foveal frequency ranges but similar across the three species, suggesting similar cochlear mechanics. INTRODUCTION Mole-Rats Hearing in the strictly subterranean mole-rats of the genera Fukomys and Heterocephalus (family Bathyergidae) has been extensively studied [1 – 10]. In the underground tunnels where they live, frequencies above 1 kHz are strongly attenuated over distances of a few meters, while lower frequencies may experience a sound-pressure amplification; this is known as the ‘stethoscope effect’ [8, 11, 12]. The hearing range of some subterranean rodents is conspicuously narrow and (at 60 dB SPL) restricted to low frequencies between 20 Hz and 6 kHz [4, 5, 13]. Reflecting the acoustic properties of their tunnels, auditory sensitivity is greatest at frequencies around 1 kHz, although hearing thresholds in this frequency range are still much higher than those of most surface-dwelling rodents [9, 14]. Furthermore, behavioral studies point to a loss of directional hearing in bathyergids, as shown by the inability of naked mole-rats to localize short sound bursts [5]. While mole-rat hearing capabilities appear poor in comparison to other rodents, the rich vocal repertoire of Mashona and naked mole-rats suggest that the auditory sense is nonetheless highly relevant [15, 16]. Furthermore, Müller et al. [3] found an acoustic fovea at low frequencies; frequencies between 0.6 and 1 kHz occupied more than 5 mm of the total length of the basilar membrane (12.9 mm), whereas all other frequency ranges occupied less than 1 mm per octave. Whether similar acoustic foveae are present in other African mole-rats remains to be demonstrated, but great morphological similarities of the organ of Corti across the sister genera Fukomys and Cryptomys suggest that this might indeed be the case [17]. Recently, Pyott et al. [2] suggested that another peculiarity should be added to the unusual auditory characteristics of some African mole-rats: the absence of distortion-product otoacoustic emissions (DPOAE). They interpreted this as indicating the lack of active amplification in mole-rat cochleae. That finding directly contradicted the earlier paper of Kössl et al. [1], which reported robust DPOAE in Ansell’s mole-rat Fukomys anselli. To clarify this discrepancy, we recorded both DPOAE and SFOAE in three mole-rat species: Ansell’s, the Damaraland (Fukomys darlingi), and naked mole-rats (Heterocephalus glaber). METHODS Animal husbandry and acoustic measurements We studied OAE in three species of African mole-rats, (Fukomys darlingi, n = 3, five ears, F. anselli, n = 2, four ears, and Heterocephalus glaber, n = 3, three ears, one ear of each animal). The subjects had a mean age of 2 ± 0.8 years and were housed at the University of Duisburg-Essen. Details on animal husbandry have been described elsewhere [9]. The Fukomys mole-rats were anaesthetized by injecting 6 mg / kg body mass ketamine and 2.5 mg / kg body mass xylazine i.m. [18] and a slightly higher dosage (9 mg /kg body mass ketamine and 3.4 mg / kg body mass xylazine i.m.) for H. glaber. The body temperature was maintained by a non-electric delta-phase isothermal heating pad (Braintree Scientific, Braintree, MA, USA) and repeatedly checked with a rectal electrode. During the measurements, the animal was placed within a custom-made semi-anechoic chamber (size: 115 cm x 80 cm x 120 cm, see [19] for details). OAE measurements took place in March and November 2022 on three bathyergid species, the Mashona mole-rat (F. darlingi, n = 3, four ears), the Ansell’s mole-rat (F. anselli, n = 1, two ears), and the naked mole-rat (H. glaber, n = 1, one ear). For a detailed description of the experimental setup and the anesthesia regime, we refer to [20]. In brief, DPOAE were generated using two primary tones with frequencies f1 and f2 and a f2/f1 ratio ranging from 1.1 to 1.4, with level L1=60 dB SPL and L2=50 dB SPL. The stimuli were 1200 ms long, including a 10 ms rise and fall time. The primaries were delivered by separate speakers. The amplitude of 2f1-f2 and 2f2-f1 were computed by fitting a sinusoidal to the distortion products. SFOAE were generated with a probe tone of 40 dB SPL and a masker tone of 70 dB SPL. The frequency of the probe tone was in the range of 300-4000 Hz. The frequency of the masker tone was 43 Hz below the probe tone. Probe and masker had a stimulus length of 800 ms, including a 10 ms rise and fall time. Probe and masker tones were delivered by separate speakers. The SFOAE was extracted by subtraction of the response of the simultaneous probemasker stimulation from the sum of the responses from probe-only and masker-only. Amplitude and phase of the SFOAE were computed by fitting a sinusoidal function to the SFOAE. The noise floors of both measurements were estimated by fitting a sinusoidal signal with a frequency + and – 0.1 x (f2-f1). Average noise floors were in the range of -4 to -24 dB SPL across the frequency range of the OAEs. The experiments were conducted in accordance with the German Regulations for Laboratory Animal Science (GVSOLAS) and approved by the North Rhine-Westphalia State Environment Agency (LANUV, permit number: 8120.04.2019.A354). RESULTS SFOAE measurements SFOAE in all three species were typically detectable over a range of about 2 octaves (Fig. 1). The amplitude profiles showed a fine structure with lobes of amplitude maxima separated by often deep amplitude notches; these regions showed phase shifts (Fig. 1B). At frequencies between the amplitude notches, the phase showed a linear relation with frequency, corresponding to a median group delay of 1.65 ms (interquartile range 1.36-2.11 ms). FIGURE 1: Characteristics of SFOAE. (A) SFOAE amplitude in three animals from three species. (B) corresponding phase characteristics. Phase traces were shifted by multiples of 2π to improve visibility. The solid black lines are least-squares fits to linear portions of the phase traces, where the horizontal extent of the lines indicates the data range included in the fit. To facilitate estimation of the phase shifts, the dashed lines were extrapolated from the linear fits. Phase shifts (vertical orange bars) were estimated as the vertical separation of extrapolated curves at the frequency of amplitude dips in the corresponding curves in panel A. (C) Group delays estimated from phase slopes in 8 ears of 6 animals. The frequency axis indicates the center frequency of the corresponding linear fit. The dashed curve corresponds to a group delay inversely proportion to frequency. The solid curve is a fit to the data. (D) Phase shifts in the same animals and ears. DPOAE measurements DPOAE in the three species were measured at ratios of f2 to f1 between 1.1 and 1.4 in 5 ears of 5 animals. The highest-amplitude DPOAE up to 12 dB SPL were detected for 2f1-f2 (Fig. 2). The high-side DPOAE at 2f2-f1 were always weaker; their levels did not exceed 5 dB SPL. Except in H. glaber, DPOAE were detected between ~0.3 and 3 kHz. In H. glaber, 2f1-f2 was detected from ~0.5 up to at least 4 kHz (Fig. 2E). Figure 3 shows maps of DPOAE phase and amplitude in F. darlingi in a format used before by Knight and Kemp [21] for data from humans, and Meenderink et al. [22] for frog data. The amplitude panels replot the data in Figure 2C. In the phase plot (panel 3B), the phase contours around 1 kHz, where the amplitudes peaked, are approximately vertical and closely spaced. This indicates that with DP frequency, the phase increases with a corresponding delay of 0.99 (s.d. 0.35) ms. At higher frequencies, the phase contours are nearly horizontal, indicating a nearly constant phase and shortened delay. FIGURE 2: Amplitudes of DPOAE of frequency 2f1-f2 (left column) and 2f2-f1 (right column) for various ratios f2/f1 of the stimulus frequencies ranging from 1.1 to 1.4. (A) and (B): F. anselli, (C) and (D): F. darlingi, (E) and (F): H. glaber. The dashed curves show the noise levels. The colors correspond across the panels and reflect the ratio f2/f1 of the stimulus frequencies as shown in panel A. FIGURE 3: DPOAE amplitude (panel A) and phase map (panel B) in F. darlingi. In each panel, data for 2f1-f2 (top half of panels) and 2f2-f1 (bottom half of panels) are shown for ratios f2/f1 from 1.1 to 1.4. The data in the amplitude map is also shown in Fig. 2C. Contours correspond to constant amplitude and phase, respectively. DISCUSSION No evidence for a Lack of Hair-Cell Active Processes Together with the report from Kössl et al. [1], our data indicate that the inability to measure DPOAE reported by Pyott et al. [2] were very likely the result of not gaining an adequately free access to the tympanic membrane. In molerats, the external ear canal is narrow, long, and convoluted, there are aggregations of cerumen, and its entrance is guarded by hairs, making it very difficult to produce an open connection to an acoustic system. The suggestion that mole-rats lack an active process is therefore no longer an issue. Nonetheless the relatively high thresholds found in mole-rats cannot so easily be explained. Brückmann and Burda [6] found the average best threshold in behavioral tests of Fukomys anselli (therein: Cryptomys sp.) to be 24 dB SPL at 0.8 kHz. Müller and Burda [23] recorded auditorysystem potentials from the brainstem and midbrain, also in F. anselli, and reported very similar thresholds to those reported from behavior. Our DPOAE and SFOAE data, however, show evoked emission amplitudes that are very similar to data reported from surface-dwelling rodents such as guinea pigs and chinchillas [24]. This strongly suggests that the middle ear cannot be inefficient, at least in the frequency range that dominates in our data (0.3 to 3.0 kHz). Evoked Emissions and the Auditory Fovea The clearest result of our measurements is the high amplitudes found in both DPOAE and SFOAE in the frequency range described as the foveal region in Ansell’s mole-rat Fukomys anselli. In that region, the space constant, i.e. the space along the cochlea occupied by one octave, is five times larger than in the other cochlear regions. Kössl et al. [1] showed that the suppression tuning curves of DPOAE are not more sharply tuned in the foveal region than outside it; the fovea is thus not specialized for very high frequency selectivity. In this respect it resembles the foveal region in the barn owl, within which neural tuning hardly changes [25], DPOAE amplitudes were highest, but the tuning selectivity of DPOAE suppression curves was broad [26]. Such foveae in mole-rats and owls can thus be interpreted as frequency regions of very high numbers of hair cells and nerve fibers underlying the neural processing of other features of acoustic signals, such as their frequency as indicated through phase locking, their relative timing, etc. This is supported by the extremely high phase-locking capabilities of barn owl auditory-nerve fibers [25]. In the individual ears that showed the highest DPOAE amplitude (and presumably an optimal connection between the acoustic system and the eardrum), the frequency selectivity for different primary-tone ratios was almost identical between ratios 1.1 and 1.3 (Fig. 3). All amplitude curves for these ratios peak in the center of the foveal region. In barn owl DPOAE, the optimal frequency ratio for the primary tones peaked when f1 was just above the region where the fovea begins (~5 kHz), i.e. when f2 was clearly within the fovea ([26], their Fig. 2). Measurements of evoked emissions will also be influenced by the number of hair cells that respond to the stimuli. For equivalent stimuli, the amplitudes of DPOAE and SFOAE in mole-rats from the foveal region were roughly the same as those reported for humans ([28, 29]) and for guinea pigs and chinchillas [24]. Thus, despite auditory thresholds that are 15 to 20 dB higher than those of these other rodents, SFOAE amplitudes did not differ. This is presumably related to the poor tuning and concomitant increase in hair-cell numbers in the foveal region that respond to each acoustic stimulus. However, mole-rat SFOAE were detectable only over less than 3 octaves. That region was flanked by narrow regions of low amplitude that show phase shifts (Fig. 1B). SFOAE amplitude and phase behavior was similar to that in other mammals. The amplitude patterns indicate that SFOAE arise due to two interfering components, as suggested for other mammals [30]. DPOAE phase patterns can be interpreted in the framework of two mechanisms contributing to DPOAE generation (21, 30]: a wave-fixed or nonlinear distortion component, and a place-fixed, or reflection, component. The wave-fixed component originates from the overlap region of the two primary tones, which is primarily near the f2 place. Its phase is relatively independent of the frequency of the stimuli, resulting in horizontal contours in the mole-rat phase maps for frequencies above the fovea (Fig. 3b). The place-fixed component is a wave generated by the nonlinear distortion that travels apically, and is reflected. Its hallmark is the changing phase of the DPOAE, shown in the mole-rat phase maps by approximately vertical contour lines (Fig. 3B). In mole-rats, the DPOAE in the frequency range of the (presumed) fovea is dominated by a place-fixed component where the amplitude is mainly determined by enhancement and reflection at the DP place. For frequencies above the fovea, DPOAE is dominated by a wave-fixed generator. Are there foveae in all mole-rats? Our data indicate that the patterns described for the amplitudes and phases of DPOAE and SFOAE in F. anselli are also found in the two other species measured. Similarity of the overall amplitude profiles suggests that F. darlingi and H. glaber also have a fovea, where the center frequency is slightly (F. darlingi), or much higher (H. glaber) compared to F. anselli. Unfortunately, the amplitudes recorded in F. darlingi and H. glaber were in general smaller and thus less clear, and this may be related to the difficulty in gaining optimal access to the tympanic membrane. Thus, there is a need for further measurements in these species. ACKNOWLEDGMENTS PvD and BM received funding from the Heinsius Houbolt Foundation. EPM thanks the Max Planck Society for funding. GAM received travel support from the Cluster of Excellence “Hearing4All” in Oldenburg, funded by the DFG (German Research Foundation) under Germany's Excellence Strategy – EXC 2177/1 - Project ID 390895286. REFERENCES 1. M. Kössl, G. Frank, H. Burda and M. Müller, J. Comp. Physiol. A. 178, 427–434 (1996). 2. S. J. Pyott, M. van Tuinen, L. A. Screven, K. M. Schrode, J.-P. Bai, C. M. Barone, S. D. Price, A. Lysakowski, M. Sanderford, S. Kumar, J. Santos-Sacchi, A. M. Lauer and T. J. Park, Curr. Biol. 30 (22), 4329–4341 (2020). 3. M. Müller, B. Laube, H. Burda, and V. Bruns, J. Comp. Physiol. A. 171, 469–476 (1992). 4. R. S. Heffner and H. E. Heffner, Hear. Res. 62 (2), 206–216 (1992). 5. R. S. Heffner,and H. E. Heffner, J. Comp. Neurol. 331 (3), 418–433 (1993). 6. G. Brückmann and H. Burda, J. Comp. Physiol. A. 181: 83-88 (1997). 7. M. J. Mason, H. L. Cornwall, and E. S. J. Smith, PLoS ONE 11 (12), e0167079 (2016). 8. K. Okanoya, S. Yosida, C. M. Barone, D. T. Applegate, E. F. Brittan-Powell, R. J. Dooling, and T. J. Park, J. Comp. Physiol. A. 204, 905–914 (2018). 9. K. R. Caspar, A. Heinrich, L. Mellinghaus, P. Gerhardt, and S. Begall, J. Exper. Biol. 224 (22), jeb243371 (2021). 10. G. A. Manley, A. Maat, S. Begall, P. Malkemper, K. R. Caspar, L. Moritz and P. van Dijk, Hear. Res. 445, 108994 (2024). 11. G. Heth, E. Frankenberg and E. Nevo, Experientia 42, 1287–1289 (1986). 12. S. Lange, H. Burda, R. E. Wegner, P. Dammann, S. Begall and M. Kawalika, Naturwiss 94, 134–138 (2007). 13. R. S. Heffner and H. E. Heffner, Hear. Res. 46 (3), 239–252 (1990). 14. P. Gerhardt, Y. Henning, S. Begall and E. P. Malkemper, J. Exper. Biol. 220 (23), 4377–4382 (2017). 15. V. Dvořáková, E. Hrouzková and R. Šumbera, Bioacoustics 25 (3), 253–266 (2016). 16. A. J. Barker, U. Koch, G. R. Lewin and S. J. Pyott. ”Hearing and vocalizations in the naked mole-rat”. In: Hearing and vocalizations in the naked mole-rat. The Extraordinary Biology of the Naked Mole-Rat, edited by R. Buffenstein, T.J. Park and M.M. Holmes, Advances in Experimental Medicine and Biology 1319, 157–195 (2021). 17. S. Lange, “Sinnesökologie Afrikanischer Sandgräber (Bathyergidae) am Beispiel von Hörund Geruchssinn“. Doctoral dissertation. Universität Duisburg-Essen (2006). 18. A. Garcia Montero, H. Burda, and S. Begall, Vet. Anaesth. Anal. 42, 187–191 (2015). 19. E. P. Malkemper, V. Topinka, H. Burda, Hear. Res. 320, 30-37 (2015). 20. G. A. Manley, A. Maat, S. Begall, P. Malkemper, K.R. Caspar, P. van Dijk, Hear. Res. in press. 21. R.D. Knight and D.T. Kemp, J. Acoust. Soc. Am. 107 (1) (2000). 22. S. W. F. Meenderink, P. M. Narins, P. van Dijk, JARO 6, 37-47 (2005). 23. M. Müller and H. Burda, Naturwiss. 76, 134-135 (1989). 24. M. A. Berezina-Greene and J. J. Guinan Jr., JARO 16, 679–694 (2015). 25. C. Köppl, J Neurophysiol 77:364-377 (1997) 26. G. Taschenberger and G. A. Manley, Hear. Res. 123, 183-200 (1998). 27. C. Köppl, O. Gleich and G. A. Manley, J. Comp. Physiol. A. 171, 695–704 (1993). 28. J. B. Dewey and S. Dhar, JARO 18, 89–110 (2017). 29. Y. Wang, Z. Qi, Y. Mengmeng, J. Wang and R. Chen, Brain Sci. 9, 255 (2019). 30. C. L. Talmadge, Long, G. R. A. Tubis, S. Dhar. J. Acoust. Soc. Am. 105 (1) (1999). COMMENTS AND DISCUSSION Vaclav Vencovsky: In the discussion, you suggest that the DPOAE phase in Fig. 3 indicates that in the fovea region, the DPOAE signal is dominated by the place-fixed reflection component whereas at higher frequencies wave-fixed nonlinear distortion was dominant. In that case, for the fovea region, the phase slope of DPOAEs could be roughly equivalent to the phase slope of SFOAEs. Is this the case in your data? Authors: The figure below shows DPOAE group delays. The average DPOAE group delay was 0.99 (s.d. 0.35) ms. For SFOAE, the average delay was 1.73 (s.d. 0.57) ms. Thus, the DPOAE delay was systematically shorter than that of SFOAE. SUPPLEMENTAL FIGURE: DPOAE group delays. The symbol colors correspond to mole-rat species; see Fig. 1. Circles: 2f1-f2; diamonds: 2f2-f1. The black line is a regression curve through the data points. Bas Meenderink: An analysis of DPOAE phase at various ratios was also performed by Meenderink and Van Dijk (2005) for the frog basilar papilla. Like the foveal frequency range in your Fig. 3, the phase contours were vertical. In other words, DPOAE phase was essentially independent of the frequency ratio of the stimulus tones and was dominated by a reflection component. However, in the frog the DPOAE data are mainly above 1 kHz. In the molerat the DPOAEs are also lower in frequency. I think there is a key difference between lowand high-frequency DPOAEs in the mammalian cochlea, and the strict ‘rules’ for the distortion and reflection sources presumably only apply for higher-frequency OAEs. Reference: S.W.F. Meenderink, P.M. Narins, P.M. and P. Van Dijk, J. Assoc. Res. Otolaryngol. 6, 37-47 (2005). Authors: In support of your remark, the Knight and Kemp analysis shows phase contour lines that tend to slope down at lower frequency (near 1 kHz, their Figure 6 and 7). This is similar to our results in Fig. 3 and may indeed indicate a deviation from the strict distortion-reflection dichotomy. The vertical contour lines we observed in the (supposed) foveal frequency range may thus merely be a characteristic of low-frequency mammalian emissions, rather than being indicative of a fovea. Nevertheless, Müller et al. [3] did find an acoustic fovea in Fukomys. The phase contours may thus reflect a fovea. Chris Bergevin: The DPOAEs in your study were measured at stimulus level L1=60 dB SPL and L2=50 dB SPL. For the SFOAE measurement, the probe tones were 40 dB SPL. The difference in delays between the two measures may thus reflect differences in the stimulus levels, where weaker stimuli in SFOAE resulted in longer delays. Authors: We agree that this may be an explanation for the difference between DPAOE and SFOAE delays.