Comparative functional characterization of novel non-syndromic GJB2 gene variant p.Gly45Arg and lethal syndromic variant p.Gly45Glu

Study samples

Individuals with hearing loss received GJB2 (GenBank ID: NG_008358.1) genotyping, performed by the clinical diagnostic Stanford Molecular Pathology laboratory, as part of routine medical care. Genomic DNA was isolated from peripheral blood according to standard procedures. The second exon of the GJB2 gene, containing the entire coding sequence, was PCR amplified and products were purified using either the Qiaquick PCR Purification Kit (Qiagen, Valencia, CA, USA) or Exo-SAP-IT (GE Healthcare Amersham, Pittsburgh, PA, USA) according to the respective protocols. Purified products were then sequenced with fluorescent di-deoxy terminators (Life Technologies, Grand Island, NY, USA) and electrophoresed on a 3130xl genetic analyzer (Life Technologies, Grand Island, NY, USA). Mutation Surveyor DNA Variant Analysis software (SoftGenetics, State College, PA, USA) facilitated the evaluation of sequence quality and the detection of sequence changes.

Two samples were selected for this study under an approved Stanford IRB protocol. Written consent was not required for this left-over DNA as per Stanford IRB approved protocol (protocol 8353, assurance number FWA00000935 (SU)). One sample from a Eurasian individual carried the novel c.133G>A, p.Gly45Arg variant. Because of incomplete family history and limited clinical testing of affected individuals, the pattern of inheritance, while most likely dominant, could not be established with certainty. The proband had post-lingual development of, thus far, mild hearing loss. A second sample from a patient with Japanese ancestry carried two variants, c.134G>A (p.Gly45Glu) and c.408C>A (p.Tyr136Stop).

Bioinformatic pathogenicity predictions of GJB2 variant p.Gly45Arg

The functional effects of the novel c.133G>A, p.Gly45Arg variant were predicted using the Sorting Intolerant from Tolerant (SIFT) program (http://sift.jcvi.org), Protein Analysis Through Evolutionary Relationships (PANTHER) (http://www.pantherdb.org; Thomas et al., 2003), cSNP scoring tool, Polymorphism Phenotyping v2 (PolyPhen-2) (http://genetics.bwh.harvard.edu/pph2; Adzhubei et al., 2010) and Functional Analysis Through Hidden Markov Models (FATHMM) (http://fathmm.biocompute.org.uk/index.html; Shihab et al., 2013). Additionally, we used Mutation Prediction (MuPred) (http://mutpred.mutdb.org; Li et al., 2009) and Predictor of Human Deleterious Single Nucleotide Polymorphisms (PhD-SNP) (http://snps.biofold.org/phd-snp/phd-snp.html; Capriotti, Calabrese & Casadio, 2006) to assess predicted degree of pathogenicity. The basis on which these algorithms could be applied and the use of these algorithms to analyze GJB2 gene missense mutations including p.Gly45Glu were previously described (Yilmaz, 2015).

Cloning and determination of allele configuration

Residual DNA samples from the two study subjects were used to clone the whole GJB2 coding region from wild-type genomic sequence and the proband sequences. The sequence containing c.133G>A, p.Gly45Arg was cloned into the pDrive vector (Qiagen, Valencia, CA, USA) using the following primers: Cx26 forward: 5′-CCCCTCGAGAGATGG ATTGGGGCACGCTGCAGACGATC-3′and Cx26 reverse: CCCGGATCCCGGTAGGTCC ACCACAGGGAGCCTTCGATGCG. XboI and BamHI restriction sites were introduced into the primers for further subcloning and are shown underscored in italics. The coding sequence of the individual carrying the c.134G>A, p.Gly45Glu and c.408C>A, p.Tyr136Stop variants was also cloned using these primers in order to determine the allele configuration of the two substitutions. When it was determined that the two variants were in cis configuration, the sequence was cloned in the same fashion using a different reverse primer, 5′-CCCGGATCCGAAACTGGCTTTTTTGACTTCCCAGAACA-3′, which substituted a Tyr amino acid for the Stop codon at position 136. Clones were selected using 50 µg/mL Ampicillin (Thermo Fisher Scientific, Carlsbad, CA, USA) and miniplasmid preparations and restriction enzyme analyses were performed to identify positive clones. These were then sequenced using Sanger sequencing with fluorescent di-deoxy terminators and electrophoresed on a 3730xl genetic analyzer (Life Technologies, Grand Island, NY, USA).

Plasmid construction

The coding regions of the wild-type GJB2 sequence, the p.Gly45Arg (G45R) variant sequence and the truncated sequence representing the p.Gly45Glu (G45E) and p.Tyr136Stop (Y136X, thus named because the protein is truncated at codon 136, and X is a one-letter abbreviation for a Stop codon) double variants in cis (G45R/Y136X and G45E/Y136X) were sub-cloned from the pDrive clones into the XhoI and BamHI sites of the pAcGFP1-Hyg-N1 and the pDsRed-Monomer-N1 vectors (Clontech, Mountain View, CA, USA). The fluorescent proteins in these and in the subsequently described constructs were located at the C-terminus of the Cx26 protein. p.Gly45Glu and p.Tyr136Stop were created separately using the QuikChange II Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA). The following primers were used for PCR amplification: p.Gly45Glu forward: 5′-AAAGGAGGTGTGGGAAGATGAGCAGGCCG-3′; p.Gly45Glu reverse: 3′-TTTCCTCCACACCCTTCTACTCGTCCGGC-5′; p.Glu45Gly forward: 5′-CTCCCTGTGGTGGACCTAAACAAGCAGCATCTT-3′ and p.Glu45Gly reverse: 3′-GAGGGACACCACCTGGATTTGTTCGTCGTAGAA-5′. The mutated bases are underscored in the primer sequences and correspond to the following: G>A at GJB2 coding nucleotide 134 resulting in Gly>Glu at codon 45 in the GJB2-WT-pAcGFP1-Hyg-N1 and GJB2-WT-pDsRed-Monomer-N1 constructs and A>G at the same nucleotide in the GJB2-G45E/Y136X-pAcGFP1-Hyg-N1 and GJB2-G45E/Y136X-pDsRed-Monomer-N1 constructs, resulting in a reversion of the Glu variant to wild-type Gly at codon 45. Positive pAcGFP1-Hyg-N1 clones were selected using 50 µg/mL ampicillin (Thermo Fisher Scientific, Carlsbad, CA, USA) and positive pDsRed-Monomer-N1 clones using 50 µg/mL kanamycin (Thermo Fisher Scientific, Carlsbad, CA, USA). Miniplasmid preparations and restriction enzyme analyses were performed to identify positive clones. All positive clones were then sequenced to verify the presence of the introduced changes and to ensure that the PCR amplification had not introduced erroneous sequence changes.

To create the fusion proteins of Cx26 and tags V5 and FLAG, adaptor-duplexes with the specific tag sequences flanked with 5′BamHI and 3′BsrGI restriction sites were synthesized, and subsequently subcloned into the 5′BamHI and 3′BsrGI restriction sites of the GJB2-pAcGFP1-Hyg-N1 constructs, replacing the GFP1 sequence with that of the epitope tag (Zhang, Scherer & Yum, 2011).

Constructs GJB2-pcDNA3.1(-)-EGFP and GJB2-pcDNA3.1(-)-mRFP, which were generously provided by Drs. Ogawa and Akiyama (Ogawa et al., 2014), were used to generate the GJB2-G45R-pcDNA3.1(-)-EGFP, GJB2-G45R-pcDNA3.1(-)-mRFP, GJB2-Y136X-pcDNA3.1(-)-EGFP and GJB2-Y136X-pcDNA3.1(-)-mRFP constructs by site directed mutagenesis as described above. The following primers were used for PCR amplification: Gly45Arg forward: 5′-CAAAGGAGGTGTGGAGAGATGAGCAGGCC-3′; Gly45Arg reverse: 3′-GGCCTGCTCATCTCTCCACACCTCCTTTG-5′; Glu45Gly forward: 5′-AAAGG AGGTGTGGGGAGATGAGCAGGCCG-3′ and Glu45Gly reverse: 3′-CGGCCTGCTCAT CTCCCCACACCTCCTTT-5′, respectively. The mutated bases are underscored in the primer sequences and correspond to the following: G>A at GJB2 coding nucleotide 133 resulting in Gly>Arg at codon 45 of constructs GJB2-pcDNA3.1(-)-EGFP and GJB2-pcDNA3.1(-)-mRFP and A>G at coding nucleotide 134 resulting in Glu>Gly at codon 45 of constructs GJB2-G45E/Y136X-pcDNA3.1(-)-EGFP and GJB2-G45E/Y136X-pcDNA3.1(-)-mRFP.

Transient transfection and fluorescence microscopy

HeLa cells, which are devoid of Cx26 expression, were purchased from American Type Culture Collection (Manassas, VA, USA) and were cultured according to standard procedures in Eagle’s Minimum Medium (EMEM) supplemented with 10% fetal bovine serum and 100U/100 µg/mL penicillin/streptomycin (Thermo Fisher Scientific, Carlsbad, CA, USA). Twenty-four hours before transfection, the HeLa cells were trypsinized and plated into a six-well plate with 18 × 18 mm² glass cover-slips in each well and containing medium without antibiotics. The transfection was performed using 5 µg of the expression vectors described above and 10 µL of Lipofectamine 2000 reagent (Thermo Fisher Scientific, Carlsbad, CA, USA) per well, according to the manufacturer’s instructions. The transfections with different constructs (wild-type, p.Gly45Arg, p.Gly45Glu/p.Tyr136Stop, p.Gly45Glu and p.Tyr136Stop) were performed in the following combinations: cis, trans, heterozygous and homozygous. The p.Gly45Glu and p.Tyr136Stop variants were co-transfected in both cis and trans as a control in order to confirm that the truncating variant reverses the deleterious effects of p.Gly45Glu in cis. The expression of the Cx26 constructs was analyzed 48 h after transfection. HeLa cells were fixed for 10 min at room temperature with 4% paraformaldehyde in phosphate buffered saline (PBS) and permeabilized with 0.2% Triton X-100 (Sigma, St. Louis, MO, USA) after washing with PBS. Coverslips were mounted with Prolong Gold antifade reagent with DAPI stain (Thermo Fisher Scientific, Carlsbad, CA, USA) on glass slides and the fluorescence was visualized and photographed using the confocal laser-scanning microscope Zeiss LSM 700 (Zeiss United States, Pleasanton, CA, USA).

Inositol trisphosphate (IP₃)-uncaging with calcium imaging and fluorescence recovery after photo bleaching (FRAP) assays

Live cells were transfected, grown on coverslips and mounted using a confocal imaging chamber (Warner Instruments, Hamden, CT, USA). After initial equilibration with Hanks’ balanced salt solution (HBSS) (Life Technologies, Grand Island, NY, USA), cells were loaded for 60 min at room temperature with 1 µM caged-IP₃ (Enzo Life Sciences, Inc., Farmingdale, NY, USA), 5 µM Rhod2-AM (Life Technologies, Grand Island, NY, USA), 1 µM Calcein Violet-AM (Life Technologies, Grand Island, NY, USA), 250 µM sulphinpyrazone, and 0.01% w/v pluronic F-127 to prevent dye sequestration and secretion. We imaged the cells after washing for 30 min in 250 µM sulphinpyrazone/Optimem (Life Technologies, Grand Island, NY, USA) using a 20× air objective mounted on a confocal microscope (LSM700; Zeiss United States, Pleasanton, CA, USA. The parallel loading of caged IP3, Calcium sensor and the FRAP dye in combination with their spectral profiles allows for simultaneous imaging of IP3-coupling and FRAP, because both protocols use the 405 nm laser line for photo activation and photo bleaching, respectively. In this experiment we temporally separated both imaging protocols in two sections (IP3 and FRAP), both starting with identical laser irradiation steps, because repeated scanning with the 405 nm laser line, which is necessary to record Calcein Violet recovery, may result in random IP3-uncaging events in cells next to the intentionally irradiated cell.

Single Connexin26-EGFP positive cells were subjected to two consecutive focal irradiation intervals at 405 nm for a duration of 7.5 ± 3.1 s (mean ± SD, n = 12) depending on cell size. After the first irradiation interval we tested for IP₃-coupling by monitoring relative calcium concentrations reflected by Rhod2-AM fluorescence intensity levels (excitation: 555 nm, emission filter: LP 560 nm) for 1 min in one second intervals. IP₃ coupling was assessed by measuring Rhod2-AM fluorescence intensity levels in the uncaged cell and the cell directly coupled with the irradiated cell.

The second laser irradiation was initiated 3 min after the first interval and the same Connexin26-EGFP expressing cell was focally irradiated a second time using the same parameters as described earlier. Microscope detector settings were changed to imaging for 10 min in one second intervals using an excitation of 405 nm and an emission filter of SP 490 nm to record Calcein Violet fluorescence intensity levels after photo bleaching. Using these parameters the target cell was bleached to at least 50% fluorescence intensity levels compared to the time point before laser irradiation. For dye coupling analysis, we determined fluorescence intensity levels within the bleached (b) area and a control unbleached (u) area. We calculated ratios of fluorescence intensities (fb/fu) for the complete time course. Dye recovery of more than 100% relative fluorescence intensity relates to the consecutive IP3/FRAP assay with two irradiation steps. FRAP was recorded after the second irradiation step, 3 min after the first laser pulse, which is why Calcein violet may recover within the next 10 min beyond intensity levels measured before the second pulse eventually. Gap junctions were blocked by washing cells with 100 µM carbenoxolone/Optimem (Life Technologies, Grand Island, NY, USA).