Date Published: June 14, 2018
Publisher: Public Library of Science
Author(s): Hanae Lahlou, Alejandra Lopez-Juarez, Arnaud Fontbonne, Emmanuel Nivet, Azel Zine, Bruce B Riley.
The inner ear represents a promising system to develop cell-based therapies from human induced pluripotent stem cells (hiPSCs). In the developing ear, Notch signaling plays multiple roles in otic region specification and for cell fate determination. Optimizing hiPSC induction for the generation of appropriate numbers of otic progenitors and derivatives, such as hair cells, may provide an unlimited supply of cells for research and cell-based therapy. In this study, we used monolayer cultures, otic-inducing agents, Notch modulation, and marker expression to track early and otic sensory lineages during hiPSC differentiation. Otic/placodal progenitors were derived from hiPSC cultures in medium supplemented with FGF3/FGF10 for 13 days. These progenitor cells were then treated for 7 days with retinoic acid (RA) and epidermal growth factor (EGF) or a Notch inhibitor. The differentiated cultures were analyzed in parallel by qPCR and immunocytochemistry. After the 13 day induction, hiPSC-derived cells displayed an upregulated expression of a panel of otic/placodal markers. Strikingly, a subset of these induced progenitor cells displayed key-otic sensory markers, the percentage of which was increased in cultures under Notch inhibition as compared to RA/EGF-treated cultures. Our results show that modulating Notch pathway during in vitro differentiation of hiPSC-derived otic/placodal progenitors is a valuable strategy to promote the expression of human otic sensory lineage genes.
Hearing loss and vestibular dysfunction are the most common sensory deficits in humans . The inner ear is a highly specialized sensory organ containing auditory and vestibular hair cells (HCs) that transduce mechanical energy into electrical energy for transmission to the central nervous system . During otic development, HCs in the inner ear are derived from the differentiation of early otic progenitor cells through a precise temporally and spatially-coordinated pattern of gene expression orchestrated by complex signaling cascades [3_4]. A normal human cochlea contains approximately 16,000 sensory HCs forming one row of inner HCs and three rows of outer HCs. They are limited in number and are susceptible to damage from a variety of insults, ranging from ototoxic drugs to loud noise exposure, genetic mutations, or the effects of aging. In contrast to the avian cochlea able to regenerate lost HCs [5–6], the mature mammalian cochlea is unable to spontaneously regenerate HCs leading to permanent hearing loss.
Regenerative medicine offers reasonable expectations for the potential treatment of inner ear disorders through the replacement of lost or damaged sensory cells. Initial advances in the differentiation of murine ESCs/iPSCs into HC-like cells [9–15] have paved the way for similar progress with pluripotent stem cells of human origin. Compelling evidence accumulated over the last decade supports hiPSC technology as offering a promising future for stem cell research, disease modeling and cell-based therapies in different types of tissues. In the case of the inner ear, one of the challenges is to better define the developmental pathways and their sequential activation/inactivation to allow the efficient in vitro production of human inner ear otic/placodal progenitors and their further differentiation into otic sensory cell-like cells. In the present study, our aim was to promote human otic/placodal induction processes and the generation of otic sensory lineage cells by exploring the effects of timely modulations of major pathways during hiPSC differentiation. Our results show that the expression of otic/placodal markers can be induced through the activation of the FGF signaling pathway by FGF3 and FGF10 ligands, suggesting that human placodal development and otic induction from hiPSCs is also an FGF-dependent process, as previously demonstrated with lineage guidance of both mESCs [9–14] and hESCs [17–20]. FGF activation has emerged as a prominent player in promoting otic-epibranchial progenitor identity during early development [51–52]. Indeed, ectopic expression of FGF3 or FGF10 during mouse embryogenesis induces the formation of ectopic otic vesicles expressing some otic markers i.e., PAX2 . Complementary loss-of-function approaches in zebrafish revealed that high levels of PAX2A and PAX8 favor otic differentiation, whereas low levels increase cell numbers in epibranchial ganglia . In addition, previous data reported that FGF signaling is important at early stages to induce expression of PAX2 and PAX8 required for otic induction through differential regulation of competence factors FOXI1, SOX3 and FGF24 . Interestingly, in our gene expression analysis of differentiated cells at day 13, we found a significant upregulation of PAX2 in parallel to a downregulation of SOX3, known as a pro-neural ectoderm lineage marker. In addition to PAX2/PAX8 expression, our results showed a population of hiPSC-derived otic/placodal progenitors that co-expressed other markers, such as DLX5 and GATA3 which are generally found in the native otic placode. While not specific individually, the combined expression of multiple otic/placodal gene markers is thought to be a good indicator of otic lineage identity [10–13]. Our results suggest that FGF treatment of hiPSCs in monolayer cultures promotes human otic/placodal progenitors while reducing or suppressing mesendoderm and pro-neural ectoderm lineages. The generation of different otic/placodal lineages implies the homogeneous nature of differentiation allowed by adherent monolayer system [19, 56] compared to EB-aggregates or 3D-based culture strategies [12, 57]. In the second step of the procedure, we tested which in vitro conditions would enhance the ability of hiPSC-derived otic progenitors to differentiate into human otic sensory cells. To this end, we explored the differentiation potential of otic progenitor cells under Notch inhibition (i.e. DBZ-treated) as compared to RA/EGF treatment previously used to induce HC-like cells in vitro from hESCs . We used multiple initial HC gene markers (ATOH1, POU4F3 and MYO7A) to examine a possible sensory cell identity after challenging human otic progenitors with either DBZ compound or RA/EGF supplements for an additional 7 days in culture. The otic sensory lineage specification depends on the proneural gene ATOH1 and its interactions with other transcription factors . ATOH1 is necessary and in some contexts sufficient for early inner ear HC development . The POU4F3 and MYO7A have been found expressed in differentiating HCs during early inner ear development [57–58] shortly after ATOH1 and are considered to be among the crucial initial HC markers in otic sensory lineage studies [29, 59]. Of interest, our qPCR analysis revealed a significant upregulation of ATOH1 expression when early otic progenitor cells were exposed to RA/EGF. This observation fits with the roles of EGF and retinoid pathways in inner ear development and with a previous study on hESCs . RA has also been shown to regulate otic vesicle formation in vivo [44–45], whereas EGF ligands i.e., EGF/TGF-alpha promote proliferation and/or maintenance of inner ear progenitor cells in vitro [46–47]. In contrast, for otic progenitors maintained 7 days in vitro under Notch inhibition, we observed a significant increase in the relative gene expression of both ATOTH1 and MYO7A as compared to their levels in age-matched RA/EGF-treated cultures. In addition, challenging otic progenitors with DBZ resulted in the generation of half (~50% of total) being MYO7-immunopositive compared to around 5% within the RA/EGF-treated cultures. Our immunostaining results provide additional insight and support the efficient promotion of otic sensory lineage achieved under Notch modulation, which led to the differentiation of sensory cell populations that were double immuno+ for MYO7A and POU4F3. Interestingly, using an embryoid body model, Costa et al.  generated HC-like cells (i.e., MYO7A immuno+) using mESCs by genetic programming through combined overexpression of ATOH1, GFI1 and POU4F3 transcription factors. Interestingly, compared to RA/EGF cultures, DBZ cultures showed a concomitant decrease in the expression of the bHLH gene HES5 and the Fringe gene LNFG both of which are components of the Notch pathway. This result is in accordance with our previous observation of a significant downregulation of HES5 following pharmacological inhibition of Notch in mouse inner ear tissue specific-stem cells differentiated in a sphere model . Furthermore, a recent study demonstrated that initial pro-sensory cell fate is regulated by Fringe activity, requires low levels of Notch signaling and is sensitive to changes in Notch signaling in the developing organ of Corti . Another subsequent cell fate decision operates via a lateral inhibition mechanism to sort out HCs and supporting cells during inner ear development [23_25]. This second cell fate process was reported as independent of Fringe activity and much less sensitive to small changes in Notch activity . Taking into account the increase in the expression of otic sensory markers and downregulation of HES5 and LNFG known to modify Notch signal transduction properties , it is reasonable to presume that in our in vitro differentiation system, a pharmacological reduction of Notch activity would affect the Fringe-dependent otic sensory cell fate process. Our results revealed that hiPSC-derived otic progenitors were capable of differentiation into cells expressing markers for otic sensory lineage. This differentiation ability towards otic sensory lineage was enhanced by a pharmacological modulation of Notch pathway in vitro. Although, human otic progenitors differentiated either under EGF/RA or DBZ expressed initial HC markers MYO7A and POU4F3, their expression was not sufficient to promote the formation of stereocilia, as reported with inner ear organoids . However, the 3D culture protocol used to generate functional HC-like cells from hESCs in this previous study involved multiple steps and was time-consuming (up to 75 days) as compared to the monolayer culture system used the present study (20 days). In addition, mature HCs differentiated in organoid aggregates may not be appropriate for further cell transplantation experiments that require hydrogel-free culture of differentiated cell progenitors. The lack of hair bundle-like structures in our hiPSC culture system suggested that human otic sensory cells were at a nascent state of commitment to HC phenotype and had failed to continue final maturation.