The inner ear is part of the anatomical structure of the ear and is located in the cavity of the sacral rock (25). Through a series of complicated canals, it forms a structure called the labyrinth, which consists of the bony and membranous labyrinths (25). The bony labyrinth contains the vestibular system, which is associated with balance, and the cochlea, which is associated with hearing (26). The cochlea converts sound into nerve impulses that transmit external signals to the central nervous system through the auditory nerve, thus resulting in hearing (26). Inner ear development is a complex process involving a series of morphological changes caused by the interaction between ectodermal epithelial cells and mesenchymal cells (27).

Brn4 is abundantly expressed in the mesenchymal cells of the embryonic ear (Fig. 2), which not only benefits the survival of spiral ganglion neurons (SGNs), but also stimulates SGNs to extend axon bundles and eventually form synapses with surrounding hair cells (28). The gap junctions between cochlear cells facilitate K⁺ transmission, thus resulting in a high resting potential for sensory hair cell excitation (29). Brn4 was reported to promote the assembly and localization of connexins at the cochlear support cell borders, thereby maintaining proper intra-cochlear lymphatic potential for the efficient transmission of sound (22). Mice with targeted mutations in Brn4 demonstrated abnormal development in the inner ear interstitial compartment and showed signs of numerous behavioral changes associated with auditory and vestibular system function defects, including vertical nodding movement, gait abnormality and hearing loss (30). As both the temporal bone and cochlea originate from mesenchymal cells of the embryonic ear, anatomical and histological analyses have indicated that Brn4 mutant mice have temporal bone dysplasia, enlargement of the inner ear canal, constriction in the bony labyrinth, thinning of the chamber wall, cochlear dysplasia, reduction in cochlear coiling and changes in the size and shape of the stapes footplate (30).

DFN3 is the most common type of X-chromosome-linked hereditary hearing loss (31). More than two decades ago, Brn4 was found to be a specific pathogenic factor in this disease (32–34). The pathogenic molecular mechanism of DFN3 simultaneously involves the sound conduction and sensory system and is based on structural changes in the Brn4 protein. Imaging examinations in patients have also revealed that Brn4 mutations can result in cochlear morphological abnormalities, internal auditory canal expansion and thickening of the stapes footplate (35). Cerebrospinal fluid was observed to flow into the vestibule through the dilated inner auditory canal, and consequently promoted perilymphatic gusher following the removal of the stapedial footplate during inner ear surgery (36). The disease is characterized by severe mixed hearing loss in males, whereas female carriers usually do not experience deafness. However, female carriers may have varying degrees of hearing threshold changes, mild semicircular canal dilatation and postural abnormalities (37). In previous years, genetic sequencing of families with DFN3 has revealed that almost all affected individuals have Brn4 gene segments with nonsense, missense or frameshift mutations, and even de novo genome deletion, thus resulting in abnormal protein structure and function (37–44). Chromosome translocations and deletions have also been detected upstream of Brn4 (up to 900 kb upstream), suggesting that some cis-regulatory elements may be present in these regions to regulate the gene expression (45). Imaging analyses of patients with DFN3 with Brn4 mutations revealed that the patients also had a high incidence of hypothalamic malformation, in addition to inner ear structural abnormalities such as stapes fixation and multiple vestibular diverticulum (46,47).

Although Brn4 gene mutations have been widely confirmed to cause severe inner ear dysplasia, its specific molecular mechanism remains unclear. Moreover, how to effectively prevent the occurrence and development of this genetic disease and how to correct the abnormal cochlear structure caused by the mutation of Brn4 require further research.

The pancreas originates from the dorsal and ventral buds of the endoderm, which are later referred to as the dorsal and ventral lobes; these develop into the ‘tail’ and ‘head’ of the adult pancreas, respectively (48). The pancreas functions as both an exocrine and endocrine gland. The endocrine function involves the pancreatic islets, which lack a fixed form and are found scattered throughout the pancreas (49). The most representative cells in the islets are glucagon-secreting α cells (15–20% of islet cells) and insulin-secreting β cells (65–80% of islet cells) (50). The balance of insulin and glucagon is crucial for the maintenance of blood glucose homeostasis (50).

Islet α cells are the first endocrine cells formed in the developing pancreas on embryonic day 9.5 in mice, and Brn4 was revealed to promote their differentiation (Fig. 2) (51–53). Islet α cells first express the proglucagon gene, and the produced proglucagon is further processed into glucagon, which increases blood glucose levels (53). Previous studies identified that the proximal G1 promoter element of the proglucagon gene, an AT-rich DNA sequence, was the binding site for Brn4 (24). Upregulated expression levels of Brn4 have been detected in islet cell lines producing glucagon and rat pancreatic islet cells, thus indicating its importance in pancreatic development. Heller et al (23) reported that Brn4, expressed in the E10 pancreas (Fig. 2), was the only confirmed islet α cell-specific transcription factor. However, Brn4 deletion did not generate abnormalities in pancreatic bud formation, pancreatic islet α cell numbers or physiological function in mice. Therefore, the synthesis and secretion of glucagon appears to depend on the participation of numerous factors, which may overlap in function.

Clinically, diabetes is typically treated with oral medication or insulin injection (54). The complications of diabetes have long been a prominent cause of human disabilities such as musculoskeletal abnormalities and microangiopathy (54,55). One previous study discovered that the differentiated islet cells from human marrow mesenchymal cells highly expressed pancreatic transcription factors, such as Brn4, in induced islet-like cell aggregates, suggesting that Brn4 may be targeted to induce the regeneration of islet cells to treat clinical diseases including diabetes, pancreatic cancer and abnormal glucose metabolism (56).

Self-proliferating neural stem cells (NSCs) are present throughout life in mammalian brains. These cells are mainly distributed in the subventricular zone and the hippocampal subgranular zone, and they can differentiate into neurons and glial cells (57). Since Brn4 was found to mediate striatal neuron precursor differentiation in 1999 (58), its role in neural regeneration has attracted significant attention (Fig. 2). Nuclear receptor-related 1 (Nurr1), a midbrain dopaminergic (DA) neuron-specific transcription factor, was reported to promote the differentiation of NSCs into DA neurons, increase the levels of dopamine in the striatum and ameliorate behavior defects in rats with Parkinson's disease. The expression of Brn4 also promoted the Nurr1-induced significant increase in the viability and maturity of DA neurons (59). Brn4 was also discovered to induce the expression of glial cell line-derived neurotrophic factor, which contributed to the survival and maturation of DA neurons, together with its receptors, GDNF family receptor α-1 and Ret (60). In addition, following denervation surgery, the expression levels of Brn4 in the hippocampus of rats were found to be upregulated, and reached a peak at 14 days post-injury (61). Functional experiments have indicated that Brn4 not only promotes the differentiation of NSCs into neurons but also enhances the maturation of neurons (61,62). This change in Brn4 may be due in part to the activation of the PI3K/AKT signaling pathway caused by upregulated expression levels of insulin-like growth factor (63). In addition, the paired box protein Pax 6/barrier-to-autointegration factor complex was demonstrated to drive neurogenesis by directly activating the expression of transcription factors, such as Brn4, during neurodevelopment (64). Investigations to determine target genes of Brn4 using RNA sequencing have revealed that, alongside the expression of Brn4, the expression levels of genes involved in neuronal development and maturation were upregulated, whereas those of genes associated with maintaining the pluripotency of NSCs were downregulated. Furthermore, C-terminal-binding protein 2 and Notch2 were suggested to be direct targets of Brn4 (65,66).

Fetal brain tissue can be used for cell replacement therapy in neurodegenerative diseases, such as Parkinson's disease, but it is limited by donor deficiency and cannot be standardized (67). Fortunately, the induced differentiation into specific cell groups after the expansion of NSCs in vitro can solve this problem. For example, in both the midbrain and the hippocampus, Brn4 promoted the differentiation of NSCs into neurons (Fig. 2) and the maturation of new neurons (58–62,68). Therefore, Brn4 may provide a target for novel drugs in the treatment of neurodegenerative diseases.

In 1998, Thomson et al (69) proposed that blastocyst-derived embryonic stem cells may be used as a clinical treatment for diseases such as type 1 diabetes and Parkinson's disease. Since this discovery, the cell transplantation theory has gradually become a research focus. However, there are ethical problems in obtaining stem cells from embryos, and transplant rejection is common during allogeneic transplantation (69). Therefore, how to obtain cell sources conveniently and quickly and eliminate rejection reactions has become an urgent problem that must be solved. In 2006, Takahashi and Yamanaka (70) identified four factors, Oct3/4 (also known as POU5F1), Krüppel-like factor 4 (KLF4), Sox2 and c-Myc together referred to as OKSM, that induced embryonic or adult mouse fibroblasts to differentiate into induced pluripotent stem cells (iPSCs) in vitro; the results were subsequently confirmed in human cells (71). However, the activation of c-Myc, a proto-oncogene, was revealed to result in the formation of tumors in mice transplanted with iPSCs (70). Subsequently, some researchers found that induced NSC (iNSC) colonies were produced after culturing mouse iPSCs induced by OKSM in NSC-specific medium for ≥8 days (72). As with Oct4, Brn2 and 4 are also members of POU protein family and have important effects on nervous system development (5). Therefore, some researchers are focusing on the roles of Brn2 and 4 in cell reprogramming.

The co-overexpression of Brn2, Sox2 and forkhead box G2 was identified to be able to reprogram human or mouse fibroblasts and astrocytes into neural precursor cells, which are also called induced DA precursors, owing to their DA differentiation ability (73–75). Under spontaneous neuronal differentiation condition, these induced neural precursor cells were observed to differentiate into glutamatergic and GABAergic neuronal subtypes (73). In previous studies, Brn2 combined with guanine nucleotide-binding protein subunit β-like protein and myelin transcription factor 1-like protein was used to reprogram somatic cells into functional neurons (76–79). A previous study suggested that Brn2 itself may be sufficient to achieve the aforementioned somatic cell reprogramming effect, and that changes in cell culture conditions determine the direction of induction (80).

Brn4, in combination with Sox2, KLF4 and c-Myc is called BSKM; BSKM was found to successfully induce fibroblast reprogramming into iNSCs (Fig. 2). These cells, after being transplanted into the stem cell pool in the adult mouse brain, exhibited true pluripotency in terms of proliferation and differentiation into neurons and glial cells (81,82). The reprogramming efficiency was further improved by the addition of E47 or transcription factor 3 (83). Oct4 and Brn4 were discovered to bind to Sox2 and form a dimer in a similar manner by cooperativity by sequencing, a method to determine the cooperativity parameters of transcription factors (84). Although both OKSM and BSKM can reprogram somatic cells into iNSCs (85), they exhibit several differences; for example, owing to the subtle differences between Oct4 and Brn4 in the POU_S domain (86), BSKM was observed to directly transform somatic cells into iNSCs without requiring the state of transient iPSCs induced by OKSM (Fig. 2) (81,86). Thus, BSKM may have lower tumorgenicity and an enhanced clinical practical value.

Moreover, reprogramming human or mouse fibroblasts or astrocytes into functional neurons or neuronal precursor cells by Brn4 is not uncommon (87–89), and these studies suggested that neurons induced by Brn4 in vitro could be used as cell resources for substitutive therapy to treat nerve injury or neurodegenerative diseases. However, numerous issues require further investigations and research, such as whether induced neurons can completely replace the damaged cells; whether these cells have the long-term characteristics of nerve cells in producing and conducting excitation; and how to accurately intervene to modulate the direction, cycle and speed of cell differentiation.