Transcriptomic Analysis Reveals an Altered Hcy Metabolism in the Stria Vascularis of the Pendred Syndrome Mouse Model

Purpose Slc26a4−/− mice exhibit severer defects in the development of the cochlea and develop deafness, while the underlying mechanisms responsible for these effects remain unclear. Our study was to investigate the potential mechanism linking SLC26A4 deficiency to hearing loss. Materials and Methods RNA sequencing was applied to analyze the differential gene expression of the stria vascularis (SV) from wildtype and Slc26a4−/− mice. GO and KEGG pathway analysis were performed. Quantitative RT-PCR was applied to validate the expression of candidate genes affected by Slc26a4. ELISA and immunofluorescence technique were used to detect the homocysteine (Hcy) level in serum, brain, and SV, respectively. Results 183 upregulated genes and 63 downregulated genes were identified in the SV associated with Slc26a4 depletion. Transcriptomic profiling revealed that Slc26a4 deficiency significantly affected the expression of genes associated with cell adhesion, transmembrane transport, and the biogenesis of multicellular organisms. The SV from Slc26a4−/− mice exhibited a higher expression of Bhmt mRNAs, as well as altered homocysteine (Hcy) metabolism. Conclusions The altered expression of Bhmt results in a dramatic change in multiple biochemical reactions and a disruption of nutrient homeostasis in the endolymph which may contribute to hearing loss of Slc26a4 knockout mouse.


Introduction
Pendred syndrome, characterized by deafness with enlargement of aqueduct and goiter, is caused by mutations of SLC26A4, one of the most prevalent causes of hereditary hearing loss globally [1]. A large-scale study showed mutations of SLC26A4 that occur in approximately 5-10% sensorineural hearing loss children among a variety of different ethnic populations [2]. The onset of hearing loss in patients with Pendred syndrome is variable. Some patients may have hearing loss at birth, while others suffer fluctuating, or progressive hearing loss during their childhood, which may provide a window of opportunity for treatment [3].
To investigate the underlying mechanism of the Pendred syndrome, researchers have created a mouse model for study.
The Slc26a4 knockout mice suffer severe hearing loss and malformation of cochlea at birth. Previous studies looking at the Slc26a4 knockout mice have found that pendrin, encoded by the Slc26a4 gene, locates in root cells of the outer sulcus, cells overlying the spiral prominence, and spindleshaped cells of the stria vascularis in cochlea. Meanwhile, three main changes occur in the inner ears of Slc26a4 mutant mouse. (1) A lack of pendrin causes the acidification of the endolymph [4]. Pendrin is an anion exchange protein, similar to Cland HCO 3-exchangers that participate in fluid regulation of the endolymph [5]. In the absence of pendrin, HCO 3-secretion into the endolymph is affected, which causes the endolymph to become acidic and leads to inhibition of Ca 2+ reabsorption [4,6]. (2) The scala media is enlarged due to the dysfunction of the endolymphatic sac.
The cochlear lumen is formed by the balance of the fluid secretion in the vestibular labyrinth and fluid absorption in the endolymphatic sac, from embryonic day (E) 13.5 and 14.5 [7]. Thus, dysfunction of fluid absorption in the endolymphatic sac, due to Slc26a4 mutation, results in an enlarged volume of the scala media [8,9]. (3) The expression of Kcnj10 is reduced, and this inwardly rectifying K + channel subunit, located in the intermediate cells of the stria vascularis (SV), is responsible for the establishment of the endolymphatic potential (EP) [10,11]. Wangemann et al. have reported an absence of Kcnj10 in the intermediate cells of Slc26a4 mutant mice at 1-4 months of age, and this lack of channel protein causes a decrease in the EP, which may directly lead to deafness [12]. Among these three changes, the pathology in the SV is the initial and predominant cause of the hearing loss associated with Slc26a4 -/mice. However, how Slc26a4 deficiency affects morphology and function of SV remains unknown.
In this study, we used RNA sequencing methodology to compare differences in the transcriptomes in the SV from wildtype and Slc26a4 mutant mice. We found that Slc26a4 deficiency significantly affected SV genes required for multicellular organism biogenesis. Furthermore, the SV from Slc26a4 -/mice exhibited higher expression of Bhmt mRNAs, as well as a dramatic reduction of Hcy, suggesting abnormality of Hcy metabolism as a novel mechanism by which SLC26A4 affects hearing.

Materials and Methods
2.1. Animals. Slc26a4 knockout mice were obtained from the Jackson Laboratory, and Calca knockout mice were from Cyagen Co. LTD. All animals were maintained in heterozygotes in SPF circumstances and adapted to the experimental circumstances for a week before the experiments. All experimental procedures were carried out in accordance with the Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee of the Shanghai Jiao Tong University Affiliated Sixth People's Hospital [Permit number: SYXK 2016-0020].

RNA-seq.
Wildtype and Slc26a4 -/mice were sacrificed at P14. The stria vascularis were isolated from inner ear and then lysed in Trizol (Invitrogen). Total RNA was isolated according to the manufacturer's instructions. All samples were sequenced on an Illumina HiSeq2500 platform at 15 million 100-bp single reads per sample. After quality control of the sequencing libraries, reads were trimmed and mapped against the Ensembl genome annotation and the human genome assembly (hg19/GRCh38) using Tophat2. Reads mapping to ribosomal RNAs or the mitochondrial genome were removed.

Quantitative
Real-Time RT-PCR. Total RNA from wildtype and Slc26a4 -/stria vascularis was extracted with TRIzol reagent according to the manufacturer's instructions (Invitrogen). One microgram of total RNA was reverse transcribed using the ReverTra Ace® qPCR RT Kit (Toyobo, FSQ-101) according to the manufacturer's instructions. A SYBR RT-PCR kit (Toyobo, QPK-212) was used for quantitative real-time PCR analysis. The relative mRNA expression of different genes was calculated by comparison with the control gene Gapdh (encoding GAPDH) using the 2 -△△Ct method (Table 1).

Immunostaining.
Wildtype and Slc26a4 -/mice were sacrificed at P14. The cochleae were isolated from temporal bone, fixed by 4% PFA, dehydrated by 30% sucrose, and cut into frozen sections. The section was then permeabilized with 0.1% Triton X-100 in PBS and blocked with 1% BSA in PBS. The samples were incubated with anti-Hcy (PAD984Ge01, Clonecloud) at 4°C overnight and then secondary antibody or DAPI at room temperature for 30 min. The stria vascularis were dissected from lateral wall and its actin structure were shown by phalloidin staining. Samples were examined and the figures were acquired with an LSM 710 confocal laserscanning microscope (Carl Zeiss, Oberkochen, Germany) at 20x or 63x magnification.
2.5. Enzyme-Linked Immunosorbent Assay. The blood plasma and brain homogenate from wildtype and Slc26a4 -/mice were prepared and persevered at -20°C until analysis.
Hcy level was measured using ELISA Kit for Homocysteine (CED984Ge, Clone-Cloud) according to the manufacture's instruction. Absorbance was assessed by an ELISA microplate reader at 450 nm and the results were expressed in pg/ml.

ABR Threshold
Testing. The hearing condition of mice was detected by auditory brain stem response (ABR) thresholds. The mice used for ABR testing were fully anesthetized by a mixture of ketamine and xylazine. The initial dose of ketamine and xylazine was 85 and 15 mg/kg, respectively. The stimulus generation and biosignal acquisition parameters were similar to those used in a previous study [13]. Auditory stimuli were presented from a speaker 10 cm ahead of the mouse's head, needle electrodes were positioned in the vertex of the head and the reference, and grounding electrodes were positioned posterior to the external auditory canals. The tone bursts were generated by TDT RP2.1 Realtime signal processor, (10-ms duration with 0.5-ms rise/fall time presented at 21.1/s intervals) at 4, 8, 16, 24, and 32 kHz with SigGen software. The biological signals picked up by the electrodes were led to a RA4PA preamplifier from Tucker-Davis Technologies (TDT System III; Alachua, FL, USA). The evoked ABR responses were amplified 20x by a PA4 preamplifier (TDT) and averaged 1000 times. The intensity of tone bursts started at 90 dB SPL and declined by a step of 10 dB, except around threshold where it was 5 dB. The threshold was determined based on visibility and reproducibility of negative-positive waves. Threshold was defined as the lowest sound level that produces a reproducible response.

Statistical Analysis.
The results are represented as the mean ± s:e:m:, and statistical significance between groups was determined using an unpaired t-test or the Mann-Whitney U test. The GraphPad Prism software 8.0 was used for all analyses, and a * p < 0:05 was considered statistically significant.  Compared to those from Slc26a4 +/+ mice, the cochlear lumen of the inner ears, isolated from 6-week-old Slc26a4 -/mice, is significantly enlarged. Furthermore, the SV can be visualized      7 Neural Plasticity through the pigmentation of the intermediate cells [14], and the lateral wall appears darker and more obvious in Slc26a4 +/+ mice when compared to Slc26a4 -/- (Figure 1(a)).

Basilar membrane Lateral wall
The inner layer of the SV is formed by marginal cells facing the endolymph [10]. Therefore, we used the actin stain phalloidin, to profile the structure of these marginal cells in SV. Indeed, enlargement of marginal cells was seen in Slc26a4 -/mice, while the marginal cells from Slc26a4 +/+ mice were normal in both size and organization [15] (Figure 1(b)).

Transcript Regulation in the Stria Vascularis Caused by Slc26a4 Deletion.
To determine the role of SLC26A4 in the function of the cochlea, wildtype and Slc26a4 -/mice were sacrificed and their SV were carefully isolated. Total RNA from pairs of SVs was extracted and then sent for nextgeneration sequencing and data from each sample generated on average 24-Mb of clean reads, after low-quality filtering, and these clean reads were mapped for reference. Each of the data sets contained 24-Mb reads and a mapping rate of 92-93%. Moreover, we counted the number of identified expressed genes and calculated their proportion and distribution to the total gene number in the database for each sample.
The correlation for gene expression levels among the samples is a key criterion to determine whether the experiments are reliable and whether the samples chosen are reasonable, and principal component analysis was performed to assess gene expression levels (Figure 2(a)).
Volcano and Scatter plots showed differentially expressed transcripts for a fold change of >2 in the Slc26a4 deficient group when compared to the wildtype group (183 upregulated genes and 63 downregulated genes) (Figures 2(b) and 2(c)).

Gene Ontology Analysis of the Differential Genes.
To better understand the associated functions of these differentially expressed genes in Slc26a4-mediated cochlear function, gene ontology (GO) analysis was used to perform enrichment analysis and classifications (Figures 3(a) and 3(b)). GO analysis identified enriched biological processes associated with "biological adhesion," "cellular component organization or biogenesis," "multicellular organismal processes," and "developmental processes," indicating that a strong multicellular organism biogenesis process occurred during the Slc26a4mediated maintenance of cochlear homeostasis. Furthermore,  Neural Plasticity enriched cellular component terms associated with "membrane," "organelle," "extracellular region," and "membraneenclosed lumen" were identified implying that diverse cellular components were involved in Slc26a4-mediated cochlear function. Enriched molecular functions were defined as associated with "structural molecule activity," "catalytic activity," and "molecular transducer activity," indicating that formation of organelle structure and intracellular signal transduction were the major molecular functions for Slc26a4.
3.4. Analysis of Important KEGG Pathways. We next used the differential genes for KEGG pathway enrichment, and the results showed that genes involving in the "ECM-receptor interaction pathway," "focal adhesion pathway," "aldosteroneregulated sodium reabsorption pathways," and "taste transduction pathways" were significantly enriched, indicating that loss of Slc26a4 genes may lead to disrupted transmembrane cell communication and transport in the SV (Figures 3(c) and 3(d)).
Collectively, GO and KEGG pathway analysis suggested an essential role for SLC26A4 in the regulation of extracellu-lar structure formation, cell-to-cell or cell-to-ECM adhesion, and transmembrane transport.
We noted that one of the important downregulated genes in the SV of Slc26a4 deficient mice was Gpr176, which encodes a G-protein-coupled receptor [16] and has previously been reported to be associated with adult-onset autosomal dominant cerebellar ataxia, with deafness and narcolepsy [17]. Quantitative RT-PCR analysis confirmed the downregulation of Gpr176 in the SV of Slc26a4 -/mice (Figure 4(c)), suggesting a reduction of GPR176-mediated signaling upon Slc26a4 deficiency.
Another important gene found was Calca and therefore, we generated a Calca knockout mouse using CRISPR/Cas9 Cndp2  Neural Plasticity technology and carefully examine its ABR threshold. The results showed that Calca +/mice had equivalent hearing to the wildtype controls (Figure 4(d)), suggesting that at least heterozygous deletion of Calca had no significant effect on hearing.
3.6. Altered Hcy Metabolism in the Stria Vascularis of Slc26a4 -/-Mice. We next determined the expression of Bhmt, as well as several other enzymes required for Hcy metabolism including Bhmt2, Ahcy, and Cbs in the SV and brain from Slc26a4 -/mice using quantitative RT-PCR. The results showed that Slc26a4 deficiency significantly increased Bhmt expression in the SV but not the brain, while the expressions of other enzymes involved in Hcy metabolism were similar in both groups (Figures 5(a) and 5(b)). Interestingly, we also observed a significant increase in the Bhmt pseudogene (Bhmt-ps), in the SV of Slc26a4 -/mice but not in brain, which was similar to the levels of Bhmt itself, suggesting that there may be a common regulation of Bhmt and Bhmt-ps in the SV. Thus, we applied immunostaining assays using an anti-Hcy polyclonal antibody to measure levels in the SV from wildtype and Slc26a4 -/mice and found that Hcy levels were dramatically decreased in the SV of Slc26a4 -/mice ( Figure 6(a)). However, we did not observe a significant change in serum Hcy levels in these mice when compared to the wild types, based on an anti-Hcy ELISA (Figures 6(b) and 6(c)).

Discussion
Hearing loss could be caused by genetic factors, aging, chronic cochlear infections, infectious diseases, ototoxic drugs, and noise exposure [18][19][20][21][22][23][24]. Most of the hearing loss is due to the damage of hair cells and spiral ganglion neurons and has been extensively investigated in many previous studies [25][26][27][28][29][30], while the detailed mechanism of stria vascularisrelated hearing loss, such as Pendred syndrome, has not been well known. SLC26A4 affects cochlear function both in humans and mice; however, the mechanism by which SLC26A4 achieves this is unclear [15]. Transcriptome sequencing has been extensively used in many previous studies to investigate the detailed mechanism in the inner ear [31][32][33][34][35][36][37]. In our study, we applied a transcriptome sequencing approach to examine the levels of differentially expressed genes in the SV from Slc26a4 -/mice. Using bioinformatic analysis, we identified 183 upregulated genes and 63 downregulated genes in the SV associated with Slc26a4 depletion. Furthermore, the cellular functions of these genes are related to cell communication, extracellular matrix organization, and transmembrane transportation. Therefore, our findings suggest a novel mechanism by which SLC26A4 can affect cochlear function.
Previous studies have shown that macrophage invasion contributes to degeneration of the SV in the Slc26a4 -/mouse model [14,38]. Consistent with this, we found an additional 10 Neural Plasticity two well-characterized inflammation regulating genes, Alox15, that encoded arachidonate 15-lipoxygenase [39] and Arrb1, encoding β-arrestin 1 [40,41], in top downregulated genes in the SV of Slc26a4 -/mice. Quantitative RT-PCR results also showed that the expression levels of Arrb1 and Alox15 were decreased in the SV of these knockout mice. These findings suggest a possible role for tissue inflammation in the degeneration of the SV in the Pendred syndrome mouse model [42,43]. The Calca gene, encoding calcitonin and alpha-calcitonin gene-related peptide [44], is one of the top downregulated genes found in our study. As reported in a previous study, Calca is expressed in mouse cochlea at an early stage during development [45]. We therefore obtained a Calca mutant mouse to determine whether the low expression of Calca affected hearing. Since Calca plays a critical role in multiorgan development [46,47], Calca -/mice, created by CRISPR/-Cas9 methodology, were unhealthy and died soon after birth. Therefore, we were forced to generate a Calca +/heterozygous mouse model to determine hearing. The ABR result showed that there was no difference in the hearing threshold between the wildtype and the Calca +/animals. No significant hearing loss may be due to the heterozygotic nature of the mice used. Therefore, further research is needed to determine whether more subtle, or hidden hearing loss is present in these mice. Also, a new Calca minus model is needed to further investigate its effect upon hearing.
Our transcriptome sequencing results found that the most distinctive upregulated gene in the SV from Slc26a4 -/mice was Bhmt, which encodes betaine Hcy S methyltransferase, whose activity is required for the transfer of a methyl group from betaine to Hcy, a nonproteinaceous sulfur amino acid [48,49]. Malfunction of Bhmt leads to hyperhomocysteinemia and several previous epidemiological and experimental studies have revealed a correlation between hyperhomocysteinemia and hearing loss [50]. Previous study showed that Hcy level plays an important part in keeping the integrity of SV, which is the foundation of SV to establish the EP [51]. We consistently observed a dramatic decrease in Hcy from the SV of Slc26a4 -/mice when compared to wildtypes, and recently, nutritional imbalance is emerging as a causative factor associated with hearing loss. Furthermore, Teresa Partearroyo et al. have reported that Bhmt plays a central role in the homeostasis 11 Neural Plasticity of methionine metabolism in the cochlea and its deficiency in mice causes increased susceptibility to noise-induced hearing loss [52,53]. Moreover, a decrease in Bhmt expression and an increase of Hcy in the SV is thought to explain the hearing loss phenotype seen in Cx30 -/mice [54]. However, one difference between our findings and previous studies is that in the SV of Slc26a4 -/mice, Bhmt is upregulated, which leads to decreased amounts of Hcy, which is part of the folate and methionine cycles. The elevated expression of Bhmt leads to an increased consumption of Hcy and folate and production of methionine and S-Adenosylmethionine, which serve as the major donors of methyl groups in the synthesis of hormones, nucleotides, and membrane lipids. It is speculated that the altered expression of Bhmt results in a dramatic change in multiple biochemical reactions and a disruption of nutrient homeostasis in the endolymph. However, whether the decrease in Hcy levels accounts for cochlear dysfunction and hearing loss in Pendred syndrome mouse models requires further investigation.

Conclusion
Transcriptomic profiling revealed that Slc26a4 deficiency significantly affected the expression of genes associated with cell adhesion, transmembrane transport, and the biogenesis of multicellular organisms. The altered expression of Bhmt results in a dramatic change in multiple biochemical reactions and a disruption of nutrient homeostasis in the endolymph which may contribute to hearing loss of Slc26a4 knockout mouse.

Data Availability
Data are available from the author Wenyue Xue (xuecin-dy0107@163.com) upon reasonable request and with permission of the Department of Otolaryngology-Head and Neck Surgery, Shanghai Jiao Tong University Affiliated Sixth People's Hospital.