The long-term changes in supporting cell morphology and viability in the hair cell damaged cochlea are well documented 
. However, much less is known about the acute effects of gentamicin exposure and hair cell loss on supporting cell specific gene expression, morphology, and viability. To characterize the molecular and cellular changes of supporting cells after acute hair cell loss we employed a well-characterized early postnatal cochlear culture system in combination with an optimized gentamicin based hair cell-ablation assay. The use of supporting cell specific Cre/loxP based labeling technique and GFP reporter line (Lfng/GFP) allowed us to monitor changes in supporting cell phenotype and behavior during and shortly after gentamicin insult. Our findings suggest that acute gentamicin based hair cell-depletion in the murine cochlea only modestly affects supporting cell viability, but induces a series of morphological changes in supporting cells. In avians, supporting cells engulf and phagocytose dying hair cells 
. Similarly, in the acutely gentamicin damaged cochlear epithelia, hair cell membrane debris were found deep in the supporting cell layer, seemingly engulfed by supporting cells. After 3 DIV, hair cell debris were cleared from the epithelium and spreading supporting cells filled the voids left by hair cells. However, in contrast to avian auditory supporting cells, which re-enter the cell cycle in response to hair cell damage 
, auditory supporting cells in the murine hair cell-depleted cultures failed to re-enter the cell cycle and remained postmitotic.
Recent studies demonstrated that Notch signaling continues to be active in the early postnatal supporting cells and genetic disruption or inhibition of Notch signaling using GSI results in robust production of ectopic hair cell-like cells in the early postnatal cochlea 
. However, the role of Notch signaling in the hair cell-damaged cochlea is less clear. Using the experimental paradigm described here, we were able to examine how hair cell loss alters Notch effector and target expression in supporting cells. Our data suggest that in the early postnatal cochlea Notch signaling is only modestly weakened by the loss of hair cells, and identify Hes1, Hey1, Hey2, HeyL, and Sox2 as targets and likely Notch effectors of this hair cell-independent mechanism of Notch signaling. Moreover, we identify two distinct differences between Notch signaling in the intact and hair cell-ablated cochlea. First, Hes5 expression is almost completely lost after hair cell loss, suggesting that in the early postnatal cochlea, Hes5 expression is heavily dependent on hair cell-mediated Notch signaling. Second, our data suggest a different mechanism of regulation for Hes1 and Hey2 in the hair cell-depleted cochlea. In the intact auditory sensory epithelium, inhibition of Notch signaling does not significantly reduce Hes1 or Hey2 transcript levels 
. However, in the hair cell damaged cochlea, loss of Notch signaling results in significant reduction of Hey2 and Hes1 expression. The altered Notch responsiveness is likely due to the loss of hair cell-derived signals in the hair cell damaged cochlea. In the intact cochlea, Hey2 is co-regulated by Fgf and Notch signaling 
. The loss of inner hair cell derived Fgf8 signaling in the hair cell damaged cochlea, likely disrupts Fgf signaling, rendering Hey2 Notch responsive.
The Notch ligand involved in this hair cell-independent mechanism of Notch signaling is unknown and experiments investigating the nature of the Notch ligand and its expressing cell type are in progress. One good candidate is the supporting cell-specific Notch ligand jagged1 (Jag1), which continues to be expressed by supporting cells even after extensive hair cell loss 
. Alternatively, it is conceivable that DNER, or a yet uncharacterized Notch ligand expressed by nerve fibers of the innervating spiral ganglion could activate Notch receptors on contacting supporting cells 
It has been proposed that the newly formed hair cell-like cells seen after prolonged Notch inhibition are the result of supporting cells trans-differentiating into hair cells 
. Employing Cre-loxP based labeling system, we are able to demonstrate that indeed, supporting cells are the source of the newly generated hair cell-like cells in our experimental paradigm. Moreover, we provide evidence that supporting cell specific features are rapidly down regulated and that these newly formed hair cell-like cells undergo a differentiation and maturation program, which resembles hair cell development in vivo
. After 3 DIV, newly formed hair cell-like cells display gene expression profile characteristics of newly differentiated hair cells: they express “early” hair cell specific genes (e.g Atoh1, Pou4f3, Nhlh1, Fgf8, Otof, Myo6, Parv) but lack expression of “late” hair cell specific markers (Tmc1, Tmc2, Strc, Ocm, Slc26a5). After an additional 3 DIV (6 DIV), the in vitro
generated hair cells up-regulate “late” hair cell specific genes and bear actin rich stereocilia-like hair cell bundles. After 6–7 DIV, hair cell-like cells showed electrophysiological properties similar to immature hair cells, rather than supporting cells. Supporting cells in untreated and gentamicin treated cultures had low membrane resistances (around 20 MΩ), due to their connections to neighboring supporting cells via gap junctions, and exhibited large membrane potential oscillations that occur throughout the network of supporting cells 
. In contrast, hair cell-like cells as well as control hair cells had high membrane resistances of hundreds of MΩ and did not show membrane potential oscillations. These observations suggest that the newly differentiated hair cell-like cells have lost their functional coupling with their neighboring cells and possibly have down-regulated supporting cell specific programs and transformed into hair cell-like cells.
The molecular mechanisms that determine an inner versus an outer hair cell fate are largely unknown. Forced expression of Atoh1 in early postnatal Prox1 positive supporting cells converted these supporting cells into hair cell-like cells but these cells surprisingly failed to induce prestin expression or other outer hair cell specific characteristics 
. In contrast, 90% of hair cell-like cells in Notch inhibited hair cell-ablated cultures expressed the outer hair cell specific protein prestin, indicating that regenerated hair cell-like cells differentiate predominantly into outer hair cell-like cells. The bias of newly generated hair cell-like cells towards an outer hair cell fate seen in the experimental paradigm described here should facilitate efforts to uncover the underlying molecular pathways involved in establishing outer and inner hair cell specific fates.
Recent studies in mice suggest that hair cell regeneration occurs in the adult vestibular utricle 
. The relative low rate of non-mitotic hair cell regeneration in the adult utricle can be significantly enhanced by treatment with GSI 
. To date there is no evidence for spontaneous hair cell regeneration in the adult mammalian auditory sensory epithelium. However, there is growing evidence that mammalian auditory supporting cells have the intrinsic ability to function as hair cell progenitors. Once dissociated and cultured in a permissive environment, immature early postnatal auditory supporting cells generate hair cells through both mitotic and non-mitotic mechanisms 
. Similarly, in the intact early postnatal cochlea, re-expression of the transcriptional activator Atoh1 
and/ or Notch inhibition results in the generation of new hair cells 
. Interestingly, these studies uncovered a difference in “hair cell generation competency” between the cochlear apex and cochlear base, with the highest number of new hair cells generated in the apex and the lowest number or no hair cells generated in the base. Similarly, our data demonstrate that supporting cells localized at the cochlear mid-apex readily trans-differentiate into hair cell like cells, whereas supporting cells in the cochlear base only infrequently do so. One possible explanation for the difference in apical and basal supporting cell competency might be the difference in supporting cell “age”. Cochlear differentiation occurs in a basal to apical gradient, with basal supporting cells differentiating first. It is possible that the apical to basal competence gradient mirrors an age dependent decline of supporting cell plasticity in the mammalian cochlea. Alternatively, it is possible that apical and basal supporting cells and/ or their environments are inherently different. Recent findings by Edge and colleagues demonstrate that in adult mice GSI-stimulated supporting cell to hair cell conversion only occurs in the cochlear apex but not in more basal regions of the cochlea 
. Future studies are required to determine the molecular underpinnings for the “supporting cell competency gradient” in the mammalian cochlea.
In conclusion, our studies revealed that Notch signaling is still active in the hair cell damaged cochlea and we identify Hes1, Hey2, Hey1, HeyL, and Sox2 as Notch targets and likely Notch effectors of this hair cell-independent Notch signaling mechanism. Using a Cre/loxP based labeling system we demonstrate that loss of Notch signaling in the hair cell-damaged cochlea allows for supporting cells to trans-differentiate into hair cell-like cells. In addition, we demonstrate that these newly formed hair cell-like cells undergo a hair specific differentiation/maturation program, subtype specific specialization, and exhibit basic electrophysiological properties similar to early postnatal hair cells rather than supporting cells. We anticipate that the experimental paradigm described here will allow future discoveries of additional regulators and modulators of auditory supporting cell plasticity.