Large-scale annotated dataset for cochlear hair cell detection and classification

Our sense of hearing is mediated by cochlear hair cells, of which there are two types organized in one row of inner hair cells and three rows of outer hair cells. Each cochlea contains 5-15 thousand terminally differentiated hair cells, and their survival is essential for hearing as they do not regenerate after insult. It is often desirable in hearing research to quantify the number of hair cells within cochlear samples, in both pathological conditions, and in response to treatment. Machine learning can be used to automate the quantification process but requires a vast and diverse dataset for effective training. In this study, we present a large collection of annotated cochlear hair-cell datasets, labeled with commonly used hair-cell markers and imaged using various fluorescence microscopy techniques. The collection includes samples from mouse, rat, guinea pig, pig, primate, and human cochlear tissue, from normal conditions and following in-vivo and in-vitro ototoxic drug application. The dataset includes over 107,000 hair cells which have been identified and annotated as either inner or outer hair cells. This dataset is the result of a collaborative effort from multiple laboratories and has been carefully curated to represent a variety of imaging techniques. With suggested usage parameters and a well-described annotation procedure, this collection can facilitate the development of generalizable cochlear hair-cell detection models or serve as a starting point for fine-tuning models for other analysis tasks. By providing this dataset, we aim to give other hearing research groups the opportunity to develop their own tools with which to analyze cochlear imaging data more fully, accurately, and with greater ease.

Intermediate Cells of Dual Embryonic Origin Follow a Basal to Apical Gradient of Ingression Into the Lateral Wall of the Cochlea

Intermediate cells of the stria vascularis are neural crest derived melanocytes. They are essential for the establishment of the endocochlear potential in the inner ear, which allows mechanosensory hair cells to transduce sound into nerve impulses. Despite their importance for normal hearing, how these cells develop and migrate to their position in the lateral wall of the cochlea has not been studied. We find that as early as E10.5 some Schwann cell precursors in the VIIIth ganglion begin to express melanocyte specific markers while neural crest derived melanoblasts migrate into the otic vesicle. Intermediate cells of both melanoblast and Schwann cell precursor origin ingress into the lateral wall of the cochlea starting at around E15.5 following a basal to apical gradient during embryonic development, and continue to proliferate postnatally.

Neural crest inducing signals

The formation of the neural crest has been traditionally considered a classic example of secondary induction, where signals form one tissue elicit a response in a competent responding tissue. Interactions of the neural plate with paraxial mesoderm or nonneural ectoderm can generate neural crest. Several signaling pathways converge at the border between neural and nonneural ectoderm where the neural crest will form. Among the molecules identified in this process are members of the BMP, Wnt, FGF and Notch signaling pathways. The concerted action of these signals and their downstream targets will define the identity of the neural crest.

Specification of the neural crest occurs during gastrulation and requires Pax7

The neural crest is a stem population critical for development of the vertebrate craniofacial skeleton and peripheral ganglia. Neural crest cells originate along the border between the neural plate and epidermis, migrate extensively and generate numerous derivatives, including neurons and glia of the peripheral nervous system, melanocytes, bone and cartilage of the head skeleton. Impaired neural crest development is associated with human defects, including cleft palate. Classically, the neural crest has been thought to form by interactions at the border between neural and non-neural ectoderm or mesoderm, and defined factors such as bone morphogenetic proteins (BMPs) and Wnt proteins have been postulated as neural crest-inducers. Although competence to induce crest cells declines after stage 10 (ref. 14), little is known about when neural crest induction begins in vivo. Here we report that neural crest induction is underway during gastrulation and well before proper neural plate appearance. We show that a restricted region of chick epiblast (stage 3-4) is specified to generate neural crest cells when explanted under non-inducing conditions. This region expresses the transcription factor Pax7 by stage 4 + and later contributes to neural folds and migrating neural crest. In chicken embryos, Pax7 is required for neural crest formation in vivo, because blocking its translation inhibits expression of the neural crest markers Slug, Sox9, Sox10 and HNK-1. Our results indicate that neural crest specification initiates earlier than previously assumed, independently of mesodermal and neural tissues, and that Pax7 has a crucial function during neural crest development.

Molecular mechanisms of neural crest induction

The neural crest is an embryonic cell population that originates at the border between the neural plate and the prospective epidermis. Around the time of neural tube closure, neural crest cells emigrate from the neural tube, migrate along defined paths in the embryo and differentiate into a wealth of derivatives. Most of the craniofacial skeleton, the peripheral nervous system, and the pigment cells of the body originate from neural crest cells. This cell type has important clinical relevance, since many of the most common craniofacial birth defects are a consequence of abnormal neural crest development. Whereas the migration and differentiation of the neural crest have been extensively studied, we are just beginning to understand how this tissue originates. The formation of the neural crest has been described as a classic example of embryonic induction, in which specific tissue interactions and the concerted action of signaling pathways converge to induce a multipotent population of neural crest precursor cells. In this review, we summarize the current status of knowledge on neural crest induction. We place particular emphasis on the signaling molecules and tissue interactions involved, and the relationship between neural crest induction, the formation of the neural plate and neural plate border, and the genes that are upregulated as a consequence of the inductive events.

Timing and competence of neural crest formation

Neural crest cells can be induced by an interaction between neural plate and ectoderm. To clarify the timing and nature of these inductive interactions, we have examined the time of competence of the neural plate to become neural crest as well as the time of neural fold specification. The neural plate is competent to respond to inductive interactions with the nonneural ectoderm for a limited period, rapidly losing its responsive ability after stage 10. In contrast, nonneural ectoderm from numerous stages retains the ability to induce neural crest cells from competent neural plate. When neural folds are explanted to test their ability to produce neural crest without further tissue interactions, we find that folds derived from all rostrocaudal levels of the open neural plate are already specified to express the neural crest marker Slug. However, additional signals may be required for maintenance of Slug expression, since the transcript is later down-regulated in vitro in the absence of tissue interactions. Taken together, these results suggest that there are multiple stages of neural crest induction. The earliest induction must have occurred by the end of gastrulation, since the newly formed neural fold population is already specified to form neural crest. However, isolated neural folds eventually down-regulate Slug, suggesting a second phase that maintains neural crest formation. Thus, induction of the neural crest may involve multiple and sustained tissue interactions.