Introduction
The mammalian inner ear is a highly complex sensory organ. Almost all the cell types of the mouse inner ear are derived from the otic placode and are formed on embryonic day (E) 8.5 [
1]. The primordia of the mouse auditory and vestibular organs are formed in the ventral and dorsal parts of the otic vesicle, respectively, at E10.5 [
1,
2]. At E12.5, the progenitor cells in the cochlear prosensory epithelium begin to exit the cell cycle sequentially from the apical to the basal turn along with the expression of P27kip1 [
3]. Hair cell (HC) differentiation begins in the midbasal part of the cochlear turn at E13.5 and spreads to the basal and the apical regions along with HC marker expression [
2]. During inner ear development, the proliferation and differentiation of the progenitor cells are precisely regulated by several signaling pathways and transcription factors, including the Notch and Wnt signaling pathways [
4].
Wnt signaling participates in maintaining the pluripotency of otic progenitors in inner ear development [
5–
7].
Lgr5, the most-studied downstream gene of the Wnt signaling pathway, is expressed in embryonic (E15.5–E18.5) and neonatal cochlear progenitors [
8,
9]. Lgr5
+ cells can act as progenitors of HCs and supporting cells (SCs) in the embryonic and postnatal stages because of their ability to self-renew, proliferate, and differentiate [
10–
13]. Chai
et al. showed that the proliferation and differentiation of Lgr5
+ cells are regulated by Wnt inhibitors and enhancers, suggesting that Lgr5
+ cells serve as Wnt-responsive progenitor cells [
13]. Wnt agonist treatment in the neonatal cochlea enhances Lgr5
+ progenitor cell proliferation and increases HC formation [
5,
6].
Notch signaling is known as a fundamental pathway that regulates the cell-fate determination in the inner ear. The Notch signaling receptors (Notch1–4), ligands (Jag1 and Delta1), and effectors (Hes1, Hes5) were first identified in the otic placode and are consistently active throughout the entire process of cochlear development [
14–
16]. During the early stage, Jag1-mediated Notch signaling is required for establishing the prosensory regions of inner ear [
17], whereas during the later stage, Notch signaling serves as a negative regulator for HC formation in the auditory sensory epithelium by “lateral inhibition,” and the loss of Notch signaling generates supernumerary ectopic HCs at the expense of SCs [
18–
22]. During HC regeneration, Notch signaling was upregulated after HC damage in mature avian auditory sensory epithelium, which prevents SCs from regenerating excessive HCs [
23].
In previous studies [
24,
25], we explored the roles of Notch signaling in postnatal mouse cochlea by specifically deleting Notch1 in Sox2
+ SCs, which resulted in the upregulation of Wnt signaling, followed by SC proliferation and HC generation in cochlea. To further dissect the interaction between Notch and Wnt signaling during early mouse cochlea development, we specifically down and upregulated Notch signaling in Wnt-responsive Lgr5
+ cells during the developmental stage at which the prosensory region is determined (E10.5–E14.5) and during the developmental stage at which HCs and SCs undergo differentiation in the cochlea (E14.5–E18.5). The prosensory region of the cochlea was evaluated by staining P27Kip1 and Jag1, and the HCs and SCs were labeled with Myo7a and Sox2, respectively, to investigate cell differentiation in the organ of Corti. We found that Notch1 deletion in Lgr5
+ cells starting from E10.5 showed no evident effect on prosensory region formation or on progenitor cell proliferation, whereas Notch1 deletion in Lgr5
+ cells from E14.5 onwards drove the quiescent progenitor cells to re-enter the cell cycle and differentiate into extra HCs. By contrast, the overexpression of Notch1 in embryonic Lgr5
+ cells showed no evident effects on the cochlear prosensory epithelia formation or HC differentiation. These results suggest that Notch signaling plays an important role in maintaining the homeostasis of the cochlear sensory epithelium in terms of cell numbers and structures, and the manipulation of the Notch signaling pathway may provide a new route for HC regeneration in the mammalian cochlea.
Materials and methods
Lgr5-EGFP-IRES-CreERT2 (Lgr5CreERT2/+, Stock No. 008875), Notch1flox/flox (Notch1f/f, Stock No. 007181), and RosaNotch1 (Notch1OE/+, Stock No. 008159) mice were purchased from the Jackson Laboratory. To generate Lgr5CreERT2/+;Notch1flox/flox embryonic mice, male Lgr5CreERT2;Notch1flox/flox mice were mated with female Notch1flox/flox mice overnight, and the mice were separated the next day, and this was considered E0.5. Lgr5CreERT2/+;Notch1OE/+ mice were generated by mating male Lgr5CreERT2/+ mice with female Notch1OE/+ mice. Mice of both sexes were used in the study. Approximately 80 mL of 40 mg/mL tamoxifen (Sigma-Aldrich) was injected intraperitoneally into the pregnant mice at E10.5 or E14.5. Exactly 5 mg/mL of 5′-ethynyl-2′-deoxyuridine (EdU, Invitrogen) was injected into the mice once daily during E10.5−E14.5 or E14.5−E18.5.
Embryos were harvested from timed pregnant females. Whole heads were fixed in 4% PFA (Sigma-Aldrich) for 1 h at room temperature and immersed in a graded series of sucrose solutions. After washing with 10 mmol/L PBS, 14 µm-thick cross-sections were prepared with a cryostat microtome (Leica). Whole-mount specimens were prepared by delicate dissection, incubated for 1 h in a solution of 10% donkey serum, 1 mmol/L PBS, and 1% Triton X-100, and incubated with primary antibody overnight at 4 °C. The primary antibodies were goat anti-Sox2 (1:500 dilution; Santa Cruz Biotechnology), chicken anti-EGFP (1:800 dilution; Abcam), rabbit anti-Jag1 (1:500 dilution; Santa Cruz Biotechnology), mouse anti-P27kip1(1:200 dilution; BD), and rabbit anti-Myo7a (1:800, Proteus Bioscience). Corresponding donkey anti-rabbit, anti-goat, anti-chicken, or anti-mouse secondary antibodies conjugated with Alexa Fluor® 488, Alexa Fluor® 568, or Alexa Fluor® 647 (ThermoFisher Scientific, 1:500 dilution) were used. The appropriate secondary Alexa Fluor-conjugated antibodies were incubated for 1 h at room temperature. EdU was detected with Alexa Flour azide using the Click-iT EdU Imaging Kit (Invitrogen) in accordance with the manufacturer’s protocol. Fluorescent images were acquired using a Leica SP8 confocal microscope.
Results
Lgr5 was expressed in the embryonic cochlear epithelium
Lgr5 expression has been observed in the embryonic (E15.5–E18.5) cochlear duct [
9,
14]. To confirm that Wnt-responsive cells exist in the prosensory epithelium, we initially investigated Lgr5 protein expression in the developing cochlea using Lgr5-EGFP-IRES-CreERT2 mice. Sox2 and Jag1 were used as markers of the prosensory epithelium that produces the organ of Corti [
7,
26]. At E13.5, Lgr5 protein was expressed in the ventral part of the mouse cochlear duct (Fig. 1), including the Sox2
+ and Jag1
+ sensory epithelium, suggesting that Wnt signaling is active in the ventral cochlear duct and may play roles in cochlear sensory epithelium formation.
Notch1 deletion in Lgr5+ cells in the otic vesicle showed no remarkable effect on prosensory region formation
To explore the role of Notch inhibition in Wnt-responsive otic progenitor cells, we generated Lgr5
CreERT2/+; Notch1
flox/flox mice, in which constitutive deletion of Notch1 in Lgr5
+ cells was achieved by tamoxifen administration at E10.5. Lgr5
CreERT2/+ mice served as control. Embryos were harvested at E14.5. During auditory sensory epithelium development, the progenitor cells exit the cell cycle sequentially from the apical to the basal regions along with the expression of P27kip1 [
3]. We found that the expression pattern of P27kip1 was similar to the expression patterns of P27kip1 in the control and Lgr5
CreERT2/+;Notch1
flox/flox mice (Fig. 2). In addition, the location of Jag1
+ cells was comparable between the two groups (Fig. 2), indicating that Notch1 deletion in Lgr5
+ cells of the otic vesicle exhibited no considerable effect on prosensory epithelium formation.
Notch1 deletion in Lgr5+ cells of the prosensory epithelium drove the quiescent progenitors to re-enter the cell cycle and give birth to new HCs
To explore the role of Notch inhibition in the cochlear prosensory epithelium, we injected tamoxifen into pregnant Lgr5CreERT2/+;Notch1flox/flox mice at E14.5. EdU was injected once daily for the labeling of the proliferating cells, and embryos were harvested at E18.5. Lgr5CreERT2/+ mice served as control. Myo7a and Sox2 were used as the markers of HCs and SCs, respectively. In the control group, one row of inner HCs and three rows of outer HCs were organized neatly in the organ of Corti at E18.5, and no EdU+/Myo7a+ cells were observed (Fig. 3). In the Lgr5CreERT2/+;Notch1flox/flox mice, several extra EdU+/Myo7a+ and EdU+/Myo7a+/Sox2+ cells were observed in the pillar cell region, which is the region between the inner and outer HCs, of the cochlear apical and middle turn. Many EdU+/Sox2+ cells were observed in the SC layer in Notch1-deficient cochleae (Fig. 3). In the control group, no EdU+/Myo7a+ cells were found in the whole cochlear epithelium (n = 4). In the Lgr5CreERT2/+;Notch1flox/flox group, the number of EdU+/Myo7a+ cells was remarkably increased to 12.42±3.11/cochlea (n = 5). These results indicated that Notch1 inhibition drove the Lgr5+ cells in the pillar cell region to re-enter the cell cycle, proliferate, and differentiate into new HCs.
The overexpression of Notch1 in Lgr5+ cells showed no evident effects on cochlear sensory epithelium development
To explore the role of Notch1 overexpression in Wnt-responsive Lgr5+ cells, we generated Lgr5CreERT2/+;Notch1OE/+ mice in which the constitutive overexpression of Notch1 in Lgr5+ progenitor cells was achieved by tamoxifen administration at E10.5 or E14.5, and embryos were harvested at E14.5 or E18.5. Lgr5CreERT2/+ mice served as control. We found no differences between the two groups regarding the size of the cochlear sensory epithelium, the expression pattern of Sox2 and Jag1 at E14.5 (Fig. 4A and 4B), or the number of HCs and SCs at E18.5 (Fig. 4C and 4D). This result suggested that Notch1 overexpression exhibited no evident effect on cochlear prosensory epithelia formation or HC differentiation.
Discussion
Cochlear sensory epithelium formation during embryonic development in mammals requires the spatial and temporal orchestration of multiple signaling pathways. Notch and Wnt are two fundamental and conserved pathways that are active throughout the entire developmental process of the inner ear [
4]. Notch signaling acts as a promoter for the determination of the prosensory domain through the “lateral induction” effect during the early development of the inner ear and is involved in the pattern formation of the specialized mosaic structure of the organ of Corti through the “lateral inhibition” effect during HC and SC differentiation [
14,
15,
26,
27]. Wnt/β-catenin signaling has been implicated in maintaining the pluripotency of progenitors during inner ear development, and Wnt-responsive Lgr5
+ cells are progenitor cells in the inner ear that can be activated to proliferate and differentiate into HCs under certain conditions [
6,
11,
12]. In our previous study, we showed that Notch signaling acts as a negative regulator of the proliferation of Wnt-responsive Lgr5
+ progenitors in the mouse cochlea during the neonatal period [
24]. However, the interactions between these two complex pathways during inner ear development must be further dissected.
Notch signaling activation in all otic epithelial cells in the mouse inner ear induces ectopic prosensory region formation, and this event is followed by HC and SC differentiation [
26–
28]. Notch1 overexpression in the otic vesicle in Col2a1-Cre mice with a Tet-On/Cre induction system causes ectopic sensory HCs and SCs in the non-sensory regions of the cochlea [
27], and Notch1 overexpression in the ectoderm in Pax2-cre mice causes otic placode expansion [
14]. The lateral induction of the prosensory region and the ectopic HCs induced by the overexpression of Notch1 occur in an age-dependent manner, and no ectopic HCs are generated in the cochlea after E13.5 [
26,
29]. In the current study, we specifically regulated Notch signaling in Lgr5
+ cells at different developmental stages. According to the GFP staining of Lgr5
+ transgenic mice (Fig. 1), the Lgr5
+ cells were mainly located in the ventral part of the cochlear duct that generates the prosensory region of the cochlea at E10.5–E14.5. The overexpression of Notch1 in Lgr5
+ cells during this stage resulted in no considerable difference in the range of the prosensory region labeled either by Jag1 or by P27Kip1 when compared with controls, and no ectopic HCs were found in the prosensory region (Fig. 4). Furthermore, Notch signaling overexpression in Lgr5
+ cells at E14.5–E18.5 showed no effect on the differentiation of HCs and SCs. The lack of effect after the upregulation of Notch1 in Lgr5
+ cells may be due to endogenous activation and functional saturation of Notch signaling in the prosensory and sensory regions during inner ear development [
16,
30].
When we conditionally deleted Notch1 in Lgr5
+ cells starting from E10.5, the markers for the prosensory domain still appeared at the proper time and location at E14.5 (Fig. 2), and progenitor cell proliferation was not affected. This finding may be explained by gene redundancy in the Notch signaling pathway, and the orchestration of the lateral induction and inhibition of Notch signaling in the developing inner ear occurs in a ligand-dependent manner [
31], and the visualization of Notch1 proteolysis showed that Notch1 is nearly undetectable during the lateral induction of the prosensory domain [
32]. Our results further demonstrated that Notch1 does not play a core role in Lgr5
+ cells during the early mammalian inner ear development stages.
In a previous study, we showed that the deletion of Notch1 in Sox2
+ cells promotes extensive re-entry into the cell-cycle by Lgr5
+ progenitor cells in the neonatal mouse cochlea [
24]. However, given that the transgenic mice used in our previous study cannot completely eliminate the increased cochlear regeneration through Sox2 haploinsufficiency effects, which was observed in a later study [
33], we further investigated the effects of Notch1 deletion specifically in the Lgr5
+ cells during E14.5–E18.5 in the current study. Notch1 activity considerably increases during the lateral inhibition process during inner ear development [
32]. As expected, we observed the proliferation of SCs, followed by the differentiation of HCs, in the cochlear sensory epithelium (Fig. 3), especially in the inner pillar cell region. However, Angelika
et al. reported that Notch signaling is not necessary for the differentiation and maintenance of pillar cell fate, which are distinguished by Hey2 expression [
19]; whereas the quantitative high-resolution cellular map of the organ of Corti showed that other Notch effector genes,
Hes5,
Hey1, and
Lfng, are expressed in the inner and outer pillar cells [
34], which suggests that the Notch signaling plays important roles in pillar cells. In current study, we found that the pillar cells displayed great regenerative potential and generated EdU
+/Myo7a
+ HCs after Notch1 deletion in Lgr5
+ cells (Fig. 3), suggesting that Notch signaling served as a negative regulator for the proliferation of Lgr5
+ progenitors in mouse cochlea development, which is consistent with the results in our previous paper [
24].
We concluded that the roles of Notch1 in Wnt-responsive Lgr5+ cells are unidirectional and stage dependent during mouse inner ear development. Specifically upregulating Notch signaling in Lgr5+ cells exhibited no effect on the prosensory region determination or cell differentiation in the organ of Corti. Specifically downregulating Notch1 in Lgr5+ cells during E14.5–E18.5 caused cell cycle re-entry and mitotic HC generation in the sensory region of the cochlea, which suggests that Notch signaling serves as a negative regulator in Lgr5+ progenitor activation. No effect was observed from upregulation or downregulation of Notch1 in Lgr5+ cells during E10.5–E14.5, which might be caused by the endogenous Notch signaling activation or gene redundancy in Notch signaling during cochlear prosensory domain determination. Thus, our findings improved the understanding of the interactions between Notch and Wnt signaling.
Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature