|Hair cells, found within the cochlea of the inner ear,
transduce soundwaves into nerve impulses and
send them to the auditory centers of the brain .
An alternative approach to surgery, regeneration therapy, is quickly becoming the focus of research attention. The field has been active since the discovery in the 1980’s that hair cells in neonatal chicks were able to regenerate after acoustic/ototoxic insult, a scenario which was once thought impossible. Since then, research has made use of the naturally regenerative qualities of hair cells in bird models, the easy access of the auditory system in zebrafish for in vivo studies, and mammalian cochlear explants as the foci for potential models of treatment. Regeneration of hair cells is thought to rely on the supporting cells which underlie the auditory epithelia. These cells, either through mitosis and subsequent differentiation, or through direct transdifferentiation, have been shown to act as a reservoir for the growth of new hair cells. Genetic assays have been used to identify genes being up- and down-regulated following hair cell injury. Of particular interest are those involved in the Notch signalling pathway – a pathway which has been shown to inhibit hair cell regeneration in rodent models. Inhibition of this pathway via pharmacological administration has been shown to restore hearing in mice following acoustic insult. These findings suggest that hair cell regeneration therapy may be a realistic alternative to surgery in the not so distant future.
Animal models are an invaluable tool for scientific research as a whole. The processes of model organisms may be investigated and manipulated in ways which are either technically or ethically impossible in humans. Comparative studies between species help to elucidate mechanisms previously unknown, with the goal of someday having research apply to the human specimen. Of particular use to hair cell regeneration research have been various species of birds, the zebrafish, as well as rodents. Many other models have been used (e.g. Drosophila), but only the three here mentioned will be highlighted.
The concept of hair cell regeneration was born in the late 1980s, with independent studies by Contache and Cruz, showing that hair cell regeneration was possible in chicks after damage to their basilar papilla (homologous to the organ of Corti in mammals). This was shown through two different methods: Contache exposed the chicks to high amplitude pure-tone noise, thus damaging hair cells specific to the applied frequency, while Cruz administered the ototoxic agent Gentamicin, leading first to the damage of high-frequency hair cell receptors (at the proximal end of the basilar papilla), and spreading to the distal, low-frequency hair cells. Both studies noted complete morphological and functional recovery within weeks of the insult. This was an interesting find, as hair cells were thought be unable to repair after damage. Further research found that it was not the damaged hair cells themselves which were giving rise to the new generation, but the support cells which surround hair cells within the auditory epithelia. With these findings in mind, research was undertaken to characterize the cellular mechanisms which constitute the process of hair cell regeneration. Regeneration via transdifferentiation and mitosis have both been found to occur, each with unique signalling cascades. Perhaps the most valuable finding was identifying the role of the Atoh1 transcription factor. This protein, first identified as a homolog of the Drosophila protein 'atonal', has been found to play a key role in the process of hair cell differentiation following insult. It is this pathway which is currently being tested as a therapeutic target for hair cell regeneration in mammals.
|Zebrafish Lateral Line|
|The lateral line organ of the zebrafish has been a very useful target for hair cell research.
Shown here are neuromast cells along the lateral line.
Figure adapted from Villablanca et al. 2006 
The zebrafish is a popular model in scientific research for several reasons. This is largely due to the characteristics of the developing embryo: it is large, transparent, has a relatively short gestation period, and is able to develop outside of the mother, thus allowing for easy experimental manipulation and observation. Additionally, the zebrafish has had its genome sequenced, thus making it a valuable species in genetic studies. Of particular interest to hair cell research is the naturally regenerative qualities of cells within the zebrafish's lateral line organ. The lateral line organ acts as a sensory organ in fish, traversing the length of the animal on either side. In zebrafish, the lateral line is easily accessible, making it a useful target for experiments, and contains hair cells homologous to those found within the organ of Corti. All such qualities present in the zebrafish have helped researchers to identify genes active (or repressed) following hair cell damage and during the process of regeneration. Various studies have focused on the actions of Growth Hormone,
Rb-Raf-1 interactions, and Notch-Atoh1 signalling pathways, to name just a few.
Hair cell regeneration research, as with scientific research in general, has benefited immensely from the use of rodents models. In regards to hair cell regeneration, rodents share many characteristics with humans. The organ for hearing in mice is fundamentally the same as that of humans and is thus subject to similar restraints. Notably, damaged hair cells of mice do not regenerate naturally. Thus, discoveries made with rodents are directly applicable to humans. Of great use have been explant studies, in which cells of the rodent cochlea are cultured in vitro and tested separate from the animal. Research from other models (birds, zebrafish etc.) often point towards useful pathways to target, homologous pathways are found in rodents, tested in vitro (i.e. on cochlear explants) and, if successful, tested in the live rodent. This was the case for the Notch-Atoh1 pathway which has recently gathered much attention (see Mizutari et al. discussed below). Previous to this study, explant models have helped to test the effects of several effectors of the Notch pathway, eventually culminating in what may be a viable in vivo treatment for hearing loss.
Models for Hair Cell Regeneration
|Hair Cell Regeneration|
|Mitosis of supporting cells and subsequent differentiation into hair cells
is a potential route for regeneration of the damaged cochlea
The first hair cells seen to appear in the regeneration process are sourced from the supporting cells of the basilar papilla. Supporting cells, relatively undifferentiated cells to begin with, undergo a process known as transdifferentiation, in which they transform from supporting cell to hair cell. The cues for transdifferentiation are not certain. Two hypotheses currently stand. The first, supposes that supporting cells which undergo transdifferentiation are intrinsically set to do so. These cells, it is proposed, are 'at the ready' for transdifferentiation, awaiting the cue of damaged/lost hair cells. Evidence supporting this view links similarity in structure and function: approximately 4% of supporting cells contain uniquely structured nuclei, while roughly 4% of supporting hair cells exhibit stem cell-like characteristics during embryonic development, having the capacity to divide more than once. Alternatively, and more popular, is the microenvironment hypothesis. This approach suggests that damaged hair cells release signals locally, which are received by the surrounding supporting cells, acting as a cue for transdifferentiation. One protein which is of particular interest is the T-cell restricted intracellular antigen-related (TIAR) protein. Translocation of TIAR from the nucleus to the cytoplasm is an early indicator of apoptosis in many tissues - an event which is observed in hair cells after Gentamicin administration. Interestingly, the translocation of TIAR to the cytoplasm occurs close to the time that supporting cells begin to transdifferentiate, leading to the tempting hypothesis that TIAR is somehow linked to the transdifferentiation signalling cascade.
Mitosis, the process of cell division, is the second, and perhaps most prominent method of hair cell regeneration. Unlike transdifferentiation, which is the direct conversion of a cell from one type to another, mitosis involves cell proliferation. In the instance of hair cell damage, where the number of lost cells may be in the thousands , the generation of new cells is necessary if the initial undamaged state of the cochlea is to be recovered. Research has shown that the timeline of regeneration via mitosis vs. transdifferentiation is different, with transdifferentiation occurring earliest, and mitosis occurring at a later stage. This method is believed to provide optimal regeneration in the auditory epithelia, with early recovery being provided by the cell population already present, and later recovery requiring the growth and proliferation of the surrounding supporting cells.
The Notch Pathway
|Blocking Notch signalling with γ-secretase inhibitor 'LY411575' allows for
Atoh1 expression and transdifferentiation of supporting cells into hair cells.
Notch is a transmembrane receptor known for its role in communication between adjacent cells during embryogenesis. When a cell is differentiating, it will express a Notch ligand which interacts with the Notch receptors of adjacent cells. This interaction triggers a cascade of intracellular events within the adjacent cell which prevent it from entering the same differentiation pathway. This local communication is necessary for the formation of functionally distinct cells which arise from the same progenitor pool . Notch-mediated lateral inhibition is seen in the developing embryo as well as between cells of the cochlea. Notch activation activates γ-secretase, which acts downstream to express the Hes5 gene. Hes5, when present in the cell, prevents the expression of the Atoh1 transcription factor - a factor known to induce hair cell differentiation.
With the Nothch-Atoh1 pathway in mind, Mizutari et al. set out to find a γ-secretase inhibitor that could be administered to the supporting cells of the cochlea. It was hypothesized that γ-secretase inhibition would disinhibit Atoh1 from Hes5, thus leading to supporting cell differentiation into hair cells (as previous literature would suggest ). The inhibitor LY411575 was selected to be given to the mice following auditory insult to the hair cells. It was found that injection of the inhibitor through the semipermeable membrane
of the round window (the structure separating the middle and inner ear) could be done without toxic effects to the animal.
|An increased density of hair cells is seen in explants
treated with LY411575 compared to control.
Figure adapted from Mizutari et al. 2013 
The presence of myosin VIIa (a marker specific to hair cells) following LY411575 administration confirmed that the newly generated cells were indeed hair cells. Furthermore, the new cells exhibited bundle-like structures - a common feature of hair cells. This increase in hair cells was accompanied by a decrease in the presence of supporting cells (which can be detected by the presence of Sox2). To test whether the newly developed hair cells were coming from the supporting cell population, supporting cells were marked with green fluorescent protein (GFP) previous to injury. The newly generated cells (found one month after trauma) were seen to be positive for both myosin5IIa and GFP, indicating that the source of the new hair cells was indeed from the supporting cell population, a finding consistent with the transdifferentiation hypothesis.
Real-time PCR analysis showed that Hes5 mRNA increases to twice its pre-trauma amount following acoustic insult, decreasing to pre-trauma levels after 3 days. This effect was blocked completely with the administration of LY411575. Likewise, the administration of LY411575 led to sustained upregulation of Atoh1 expression following injury. The effect however, was not seen uniformly across the basilar membrane. Recovery was mostly limited to the middle of the cochlea, with the basal end showing no recovery at all. As the cochlea exhibits tonotopic organization, this limited recovery would be reflected in functional recovery (i.e. varied frequency sensation). Auditory brainstem responses (ABRs) - a method of detecting frequency perception in animal models - reflected this varied recovery, with best results between 8-16 kHz (the frequency correlating to the middle of the cochlea).
The work of Mizutari et al. is a milestone in hair cell regeneration research, but there is still plenty of work to be done. For example the functional recovery that was shown was only partial - within a small range of frequencies. Although the tonotopic layout of the human cochlea is different than that of the mouse, recovery in humans is likely to be similarly limited. Additional research directed at recovery of hair cells within the base and apex of the cochlea would be warranted. Furthermore, the recovery is only in instances of acoustic damage, and the treatments are administered rapidly to a population of young mice. These may be limitations to the experiment itself, but there is a broad range of research (some of which is mentioned above) tackling the problem from many angles. It is unlikely that any one method of treatment will lead to full recovery of the auditory epithelia. It is through the combined efforts across research modalities that auditory restoration via hair cell regeneration will become a viable option within the near future.