Brownell discovered that mammalian OHCs were able to alter their length when electrically stimulated [16]. Following the discovery of “electromotility”, the study of its mechanism and role in the vertebrate auditory system became one of the most exciting areas in hearing research. Experiments which detected cellular motility even after the degradation of the cell’s content via internal tryptic digestion suggested that a molecular motor in the plasma membrane drives the mechanism of force generation [17]. The surface area of the plasma membrane was covered to nearly 70% by prestin. OHCs demonstrate piezoelectric properties with an efficiency of conversion from mechanical force to electrical charge is approximately four times greater than that of the best man-made piezoelectric material [18].
Prestin typically shares the protein structure of the SLC26A family: a conserved central region of hydrophobic amino acids with N- and C-terminal residues on the cytoplasmic side of the plasma membrane. The sulfate transporter (SulTP) sequence is located in the hydrophobic core, while a STAS domain with clusters of charged residues is present in the C-terminal region. Amino acids in the SulTP domain are almost identical among mammalian species, such as humans, mice, rats, and gerbils [19]. Although prestin was identified 20 years ago, its experimental 3D structure is still unavailable. Prestin may contain specific domains that serve as the ‘voltage sensor’ (to detect voltage change) and the ‘actuator’ (to generate length change and force). Their fundamental characteristics and mechanisms, however, remain unexplored. Approximately 200 amino acid residues have since been mutated, to determine the mechanism of action in the voltage sensor and identify sequences critical for prestin function [20]. Another approach to probe the region responsible for motor capability involves locating residues that are conserved in mammalian prestin, but variable in non-mammalian prestin orthologs. To further understand the molecular and cellular mechanisms underlying this mysterious motor protein, we attempted the mapping, sequencing, and cloning of a non-mammalian prestin ortholog using RNA-Seq.
The bullfrog SLC26A5 cDNA is 2,292 bp long, and encodes a predicted polypeptide of 763 amino acid residues. After isolating the prestin gene from the inner ear cDNA of the American bullfrog, we generated a stable cell line transfected with this new coding gene. Confocal images localized the heterologously-expressed frog prestin in the plasma membrane (Fig. 1B). NLC was measured both in HCs from the frog’s inner ear (Fig. 3A–C), and in HEK293T cells expressing frog prestin, to analyze its functional property (Fig. 3D). For each cell type, four parameters (Qmax, Clin, V1/2, and z) of NLC were calculated. The charge density in frog HCs was higher than that recorded in HEK293T cells (Fig. 4A, B). Based on the fundamental assumption that a direct relationship exists between the molecular density of the protein in the cell membrane and the amount of charge recorded by the electrode [21], it is reasonable to conclude from our results that the density of endogenously-expressed prestin is higher than when expressed in the cell line. The z values obtained here should be noted (Fig. 4D). The absence of a significant difference in the z value between the two cell types suggests that the same charge is moving through the transmembrane electrical field within the protein. Alterations in intracellular ion concentration can shift the V1/2 direction within a range of − 180 and greater than 100 mV [22]. We observed a shift in V1/2 to a more positive direction from the frog AP cells, due to variations in intracellular conditions between these and HEK293T cells (Fig. 4C). Our experiment demonstrated that the new coding gene could encode a functionally active protein conferring NLC to both frog HCs and the mammalian cell line. As previous studies have shown that non-mammalian prestin does not demonstrate motor capability, and that the motor function of prestin is a newly derived molecular property exclusive to mammals [9, 13, 23], we did not attempt to examine the motor capability of frog prestin in the present study. The lack of motor function in non-mammalian prestin indicates that the ‘voltage sensor’ and ‘actuator’ in the molecule may evolve independently and have different structural bases.
Analysis of the gerbil and bullfrog prestin amino acid sequences revealed approximately 40% identity among the two species (Fig. 1). Pendrin, the closest mammalian prestin paralog, carries 40% sequence identity and exhibits no voltage-dependent NLC [24]. Zebrafish prestin carries more than 50% sequence identity compared to that of mammals, and possesses no electrophysiological characteristics. As the voltage sensing range of zebrafish prestin is not within the range of − 150 mV to 100 mV, uncertainty remains as to whether a two-state Boltzmann function is appropriate for its description [23]. Studies on avian species revealed that two types of HCs occur in the chick inner ear, neither of which possess voltage-dependent non-linear capacitance [9]. However, contrasting results from immunolabeling studies by Maryline Beurg confirmed the presence of chicken prestin in the hair cell lateral membrane, and demonstrated that HCs of the chicken auditory papilla possessed NLC [25]. NLC from cPres-expressing cell lines was not measured in the present study (Fig. 5C). Previous studies focusing on amphibian prestin are quite rare, thereby limiting knowledge of its function. NLC measurements from xPres-expressing cell lines revealed that xPres possessed similar electrophysiological features as rPres (Fig. 6). If the functional evolution of prestin is characterized by a gradual gain of NLC as demonstrated in a previous study, frog prestin would therefore be evolutionarily more advanced than avian and teleost prestin [11, 13]. The presence of NLC in frog prestin might suggest a common mechanism within the protein structure for their functional significance both in mammalian and amphibian prestin. The predicted 3D structure of gerbil and frog prestin showed that they shared a similar framework (Fig. 2C). It is reasonable to assume that the voltage sensor of prestin consisted of residues present in the frog prestin sequence, but absent in pendrin and chick prestin. Further comparative studies may reveal the molecular peculiarities underlying the mechanisms of prestin.