Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 May 1.
Published in final edited form as: J Invest Dermatol. 2019 Nov 6;140(5):1035–1044.e7. doi: 10.1016/j.jid.2019.09.022

Allele-Specific Small Interfering RNA Corrects Aberrant Cellular Phenotype in Keratitis-Ichthyosis-Deafness Syndrome Keratinocytes

Ming Yang Lee 1, Hong-Zhan Wang 2, Thomas W White 2, Tony Brooks 3, Alan Pittman 4,5, Heerni Halai 1, Anastasia Petrova 1, Diane Xu 1, Stephen L Hart 6, Veronica A Kinsler 6,7, Wei-Li Di 1
PMCID: PMC7395648  NIHMSID: NIHMS1613077  PMID: 31705875

Abstract

Keratitis-ichthyosis-deafness (KID) syndrome is a severe, untreatable condition characterized by ocular, auditory, and cutaneous abnormalities, with major complications of infection and skin cancer. Most cases of KID syndrome (86%) are caused by a heterozygous missense mutation (c.148G>A, p.D50N) in the GJB2 gene, encoding gap junction protein C×26, which alters gating properties of C×26 channels in a dominant manner. We hypothesized that a mutant allele-specific small interfering RNA could rescue the cellular phenotype in patient keratinocytes (KCs). A KID syndrome cell line (KID-KC) was established from primary patient KCs with a heterozygous p.D50N mutation. This cell line displayed impaired gap junction communication and hyperactive hemichannels, confirmed by dye transfer, patch clamp, and neurobiotin uptake assays. A human-murine chimeric skin graft model constructed with KID-KCs mimicked patient skin in vivo, further confirming the validity of these cells as a model. In vitro treatment with allele-specific small interfering RNA led to robust inhibition of the mutant GJB2 allele without altering expression of the wild-type allele. This corrected both gap junction and hemichannel activity. Notably, allele-specific small interfering RNA treatment caused only low-level off-target effects in KID-KCs, as detected by genome-wide RNA sequencing. Our data provide an important proof-of-concept and model system for the potential use of allele-specific small interfering RNA in treating KID syndrome and other dominant genetic conditions.

INTRODUCTION

Keratitis-ichthyosis-deafness syndrome (KID syndrome [MIM 148210]) is a rare, autosomal-dominant condition characterized by ocular and auditory impairment and hyperkeratotic skin lesions (Burns, 1915). Ocular involvement includes photophobia and neovascularization, progressively reducing visual acuity (Caceres-Rios et al., 1996), and auditory impairment features sensorineural hearing loss (Patel et al., 2015). Skin involvement consists of erythrokeratodermic or verrucous plaques and palmoplantar keratoderma with alopecia and/or onychodystrophy (Caceres-Rios et al., 1996; Coggshall et al., 2013). Moreover, KID syndrome is frequently complicated with chronic, opportunistic cutaneous infection, resulting in failure to thrive and, in severe cases, septicemia (Coggshall et al., 2013). Patients are at increased risk of developing skin malignancies, particularly squamous cell carcinoma (Coggshall et al., 2013). These complications can have a significant impact on life expectancy. Current cutaneous treatment is limited to symptomatic management, including retinoids to attempt to improve the skin barrier and antifungal and antibacterial agents for infection control (Coggshall et al., 2013). There is an unmet need for targeted treatment of this condition.

The cause of KID syndrome was identified as heterozygous missense mutations in the GJB2 gene (Richard et al., 2002; van Steensel et al., 2002), which encodes a transmembrane gap junction (GJ) channel-forming protein, C×26. GJs are clustered intercellular structures found in virtually all contacting cell types, enabling direct cell-cell communication via the exchange of ions, nutrients, and signaling molecules with a molecular weight <1 kDa (Elfgang et al., 1995; Levit et al., 2015). Connexin proteins, the constituents of GJs, can oligomerize to form hexameric structure known as connexons. On the plasma membrane, connexons can either function alone as hemichannels or dock with a compatible connexon from the adjacent cell membrane to form a GJ (Laird, 2006). Undocked hemichannels serve as a conduit between the cytoplasm and the extracellular space of the cell, whereas GJs couple the cytoplasm of adjacent cells electrically and biochemically (García et al., 2016).

To date, 12 missense GJB2 mutations have been reported in KID syndrome, of which the mutation c.148G>A, resulting in the substitution of aspartic acid for asparagine at codon 50 (p.D50N), is by far the most common mutation, accounting for 86% of cases in the largest European cohort (Mazereeuw-Hautier et al., 2007). At least 10 out of the 12 identified mutations, including p.D50N, have been associated with aberrant hemichannel behavior (Donnelly et al., 2012; García et al., 2016; Lee et al., 2009; Mese et al., 2011), presented as elevated membrane currents (Lee et al., 2009), enhanced permeability to small-molecule tracers (Mese et al., 2011), and/or enhanced ATP release in response to specific stimuli (Donnelly et al., 2012; García et al., 2016). Therefore, hemichannels have been considered a potential therapeutic target when developing new KID syndrome treatment (Levit et al., 2015; Xu et al., 2017). Recent work has shown that mefloquine, a US Food and Drug Administrationeapproved antimalarial drug, potently suppresses aberrant hemichannels in primary keratinocytes (KCs) from a transgenic mouse model with a heterozygous p.G45E mutation in GJB2 (Levit et al., 2015). Very recently, the monoclonal antibody abEC1.1 was developed, which specifically suppressed hemichannels formed by C×26 wild-type (WT), p.G45E, or p.D50N mutants (Xu et al., 2017). However, it is unclear whether those strategies can discriminate mutant GJB2 alleles from WT. This concern is particularly important given the context that most KID syndrome mutants exert dominant effects on co-expressed WT connexins (Di et al., 2005; García et al., 2015). In the last decade, allele-specific small interfering RNA (AS-siRNA) technology has shown strong therapeutic potential in treatment of dominant genetic disorders and brought clinical benefits to a patient with pachyonychia congenita (Trochet et al., 2015).

We present a specific and effective AS-siRNA against the GJB2 c.148G>A (p.D50N) mutation, which successfully rescues the abnormal cellular phenotype in patient-derived KCs. Our approach could potentially be a novel future therapy for this debilitating and life-limiting condition.

RESULTS

Patient-derived KCs with a heterozygous c.148G>A (p.D50N) mutation had aberrant GJ and hemichannel behavior and caused hyperkeratotic skin morphology

Previous studies on mutant GJB2 expression, distribution, and function have relied largely on ectopic expression of homozygous GJB2 mutations in Xenopus oocytes (Lopez et al., 2013) and HeLa cells (Press et al., 2017). These models, however, do not accurately represent the genetic state in patients with KID syndrome, who are heterozygotes for GJB2 mutations, and therefore have limitations when used in preclinical evaluation for new therapeutic strategies. To overcome this, primary KCs isolated from a fresh skin biopsy of a patient with KID syndrome heterozygous for c.148G>A (p.D50N) mutation in GJB2 were immortalized using the lentiviral vector encoding HPV type 16 E6/E7 cDNA.

The GJB2 mutation c.148G>A was confirmed in the immortalized cells (KID-KCs) (Figure 1a). KCs obtained from a healthy donor and immortalized with the same protocol were used as a control (control-KCs). Both immortalized cell lines showed polygonal morphology with various sizes in early passages but were more uniformly shaped in later passages (Figure 1b). The cell morphology and genotype were monitored over a propagation period up to 45 passages. There was no apparent change in morphology, and the mutation in KID-KCs was expressed stably. The expression of endogenous GJB2 mRNA in KID-KCs and control-KCs was examined by quantitative reverse transcriptaseePCR (qRTPCR) using specific primers for total GJB2. A reduction of 44% in total GJB2 mRNA in KID-KCs was seen compared with that in control-KCs (n = 4, P = 0.24) (Figure 1c). The PCR amplicon of KID-KCs was sequenced, and the chromatogram exhibited a 1:1 ratio for WT:mutant allele peaks at the c.148 locus (Figure 1a), suggesting similar mRNA expression levels for both WT and mutant alleles in KID-KCs.

Figure 1. Genotype, morphology, GJB2 expression, and subcellular localization in keratinocytes.

Figure 1.

Open in a new tab

(a) cDNA sequences of GJB2 from the patient with KID syndrome and the healthy donor. (b) Morphology of keratinocytes from healthy donor and patient with KID syndrome at early passages (P1 or P5) and late passages (P35 or P45). Bar = 100 μm. (c) mRNA expression of total GJB2, the WTallele, and the mutant allele in keratinocytes from healthy donor and patient with KID syndrome determined by qRT-PCR. (d) Total C×26 protein (asterisk, at26 kDa) expression was examined by immunoprecipitation and immunoblotting, which shows decreased C×26 expression in patient cells. (e) Expression of C×26 detected by immunofluorescence staining, where gap junction plaques can be found at cell-cell junctions in normal keratinocytes (arrows), whereas C×26 in keratinocytes from the patient with KID syndrome was localized discretely in the cytoplasm (arrowheads). Bar = 40 μm. (f) E-Cad was stained in green color. C×26 expression in the skin is shown, where punctate staining of C×26 was observed (arrows). The dotted lines show dermal-epidermal junction. Bar = 40 μm. ***P < 0.001. Cont, keratinocytes derived from the healthy donor; E-Cad, E-cadherin; KID, keratitis-ichthyosis-deafness; N.S., not significant; qRT-PCR, quantitative reverse transcriptase-PCR; WT, wild-type.

The expression of the C×26 protein was also examined by immunoblotting. Several anti-C×26 antibodies have been used in immunoblotting previously (Press et al., 2017; Yum et al., 2007), but most were used to detect C×26 in rodent cells or tissues or in HeLa cells ectopically expressing C×26, and only a few were able to detect endogenous C×26 in cultured human KCs. This is possibly because of a low endogenous C×26 level or a lack of anti-C×26 antibodies with sufficient affinity and/or specificity. We tested six antibodies and found that the pair of a rabbit polyclonal antibody (Thermo Fisher Scientific, Waltham, MA, 13–8100) and a mouse monoclonal antibody (Merck Millipore, Burlington, MA, MABT198) gave a clear band at 26 kDa when used for immunoprecipitation experiments (Figure 1d). Immunoprecipitation using this pair of antibodies showed a reduction in C×26 protein expression in KID-KCs compared with control-KCs, which was consistent with the qRT-PCR results.

The distribution of C×26 in the cells was determined using immunostaining. In control-KCs, punctate or plaque-like C×26 staining was observed at cell-cell contact sites (Figure 1e), which were indicated by membranous staining of E-cadherin, suggesting that WT C×26 was able to traffic to the plasma membrane and form GJ plaques. By contrast, C×26 in KID-KCs failed to accumulate at membrane regions but showed a primarily discrete punctate staining pattern in the cytoplasm (Figure 1e). Although a small portion of GJ plaques overlaid with E-cadherin at the plasma membrane, they were smaller in size than those observed in control-KCs. Interestingly, the immunostaining pattern of the patient skin did not show a striking reduction or mislocalization of C×26 expression (Figure 1f). This could be explained, in part, by the in vitro culture condition, which rendered KCs more proliferative and less differentiative, given that C×26 is predominantly expressed in differentiated KCs (Churko and Laird, 2013; Martin et al., 2014).

The function of GJ intercellular communication in KID-KCs was assessed by scrape-loading dye transfer using a neurobiotin tracer and compared with control-KCs. Neurobiotin diffused extensively from initially scrape-loaded cells to neighboring cells in control-KCs (Figure 2a). In contrast, the diffusion of neurobiotin in KID-KCs reduced markedly and was almost confined to the first line of the scrape-wounded cells. Quantification of the images revealed a reduction of 58% in diffusion area in KID-KCs compared with that in control-KCs (n = 3 each, P < 0.01), suggesting that the GJ channels formed in KID-KCs were defective (Figure 2b).

Figure 2. Abnormal gap junction and hemichannel behavior in KID-KCs.

Figure 2.

Open in a new tab

The gap junction intercellular communication in KID-KCs (KID) or control-KCs (Cont) was examined by the SLDT assay, and the hemichannel activity was examined by whole-cell patch clamp and NB uptake. (a) Representative images of SLDT in the keratinocytes using NB tracer (red). (b) Data analysis shows that KID-KCs had impaired ability to transfer NB to adjacent cells. (c) Representative patch clamp records from single keratinocytes in response to the voltage step protocol from −110 mV to +110 mV in 20 mV increments. (d) The plot of current density against membrane voltage reveals aberrantly enhanced hemichannel activity in KID-KCs. (e) Representative images of NB uptake (red), with the nuclei stained with DAPI (blue). (f) Data analysis shows increased uptake of NB in KID-KCs. All data are presented as the mean ± SEM. Bar = 200 μm. *P < 0.05; ***P < 0.001. Control-KC, control keratinocyte; KID-KC, keratitis-ichthyosis-deafness syndrome-derived keratinocyte; NB, neurobiotin; SEM, standard error of the mean; SLDT, scrape-loading dye transfer.

Next, the activity of hemichannels in KID-KCs was measured using whole-cell patch clamp and neurobiotin uptake assay. Moderate membrane currents were recorded from control-KCs at both depolarizing and hyperpolarizing membrane voltages, whereas large currents were elicited from KID-KCs at all tested membrane voltages, more prominently at depolarizing voltages between +30 mV and +110 mV (Figure 2c). The maximum current density recorded from KID-KCs was 80% greater than that in control-KCs (9.0 ± 1.3 pA/pF, n = 21 cells vs. 5.0 ± 0.6 pA/pF, n = 14 cells, measured at +110 mV, P < 0.05) (Figure 2d). Consistent with the patch clamp results, KID-KCs showed a marked increase in uptake of neurobiotin tracer compared with control-KCs (n = 29 and 34 cells, respectively, P < 0.001) (Figure 2e). These results suggested enhanced membrane conductivity and neurobiotin permeability in KID-KCs, indicating hyperactive hemichannel behavior conferred by the mutation. Our in vitro findings indicated that, despite expressing C×26 at a relatively low level in culture, the KID-KCs displayed an aberrant cellular phenotype that has been reported previously in other KID syndrome disease models (Arita et al., 2006; García et al., 2016; Lee et al., 2009).

To confirm that the immortalized KID-KCs remained capable of proliferation and differentiation, characteristics of primary KCs in vivo, these cells were tested in a human-murine chimeric skin graft model (Di et al., 2011). Histological examination of the skin graft regenerated from immortalized KID-KCs showed similar features seen in the skin of a patient with KID syndrome, including hyperkeratosis and spongiosis (Figure 3). Collectively, the immortalized KID-KCs are a suitable model for evaluating therapeutic efficacy of AS-siRNA for KID syndrome.

Figure 3. Epidermal morphology of grafted skin in human-murine chimeric skin graft model.

Figure 3.

Open in a new tab

Primary fibroblasts and immortalized keratinocytes derived from the patient with KID syndrome harboring a heterozygous GJB2 c.148G>A mutation or a healthy donor were used to generate bioengineered skin sheets, which were grafted onto NOD-severe combined immunodeficiency mice (NSG mice). Eight weeks after grafting, regenerated skin grafts were harvested. (a, d) Macroscopic examination showed fine, dry scales in the graft generated from (d) patient cells compared with that generated from (a) control cells. (c, f) Histological examination showed hyperkeratosis and spongiosis in (f) the patient skin graft, resembling that seen in (c) the patient skin. (b, e) The skin architecture of (e) the control skin graft was also similar to (b) healthy donor skin. Bar = 100 μm. KID, keratitis-ichthyosis-deafness.

AS-siRNA selectively inhibited the c.148G>A mutation in KID-KCs

Nineteen candidate AS-siRNAs with a targeting sequence complementary to the c.148G>A mutation (S1–S19) were screened at a concentration of 50 nM in HeLa cell lines stably expressing WT or mutant GJB2 fused with the GFP reporter gene. Both cell lines were transfected with each of 19 AS-siRNAs, followed by flow cytometry analysis for reduction of GFP intensity (Supplementary Figure S1b). The knockdown efficiency of S7 and S10 in cells expressing the mutant GJB2-GFP was approximately 50%. Because S7 inhibited mutant GJB2 specifically and reproducibly from three independent screening experiments (Supplementary Figure S2), this siRNA was selected for further study.

The allele-specific action of S7 at 50 nM was tested in the KID-KCs harboring the heterozygous mutation. KID-KCs and control-KCs were treated with S7, and the mRNA expression of GJB2 in treated cells was examined 24 hours after treatment by qRT-PCR. The treatment resulted in a significant decrease of 63% in total GJB2 mRNA in KID-KCs compared with untreated cells (n = 3, P = 0.0065), but the decrease was not detected in control-KCs with the same treatment (n = 3, P = 0.84) (Figure 4a). Further investigation using allele-specific primers showed no difference in mRNA expression of the WT allele between untreated and S7-treated cells (n = 3, P = 0.51 for KID-KCs and P = 0.60 for control-KCs), whereas mRNA expression of the mutant allele in KID-KCs was significantly inhibited by 43% following S7 treatment (n = 3, P = 0.0065). At the protein level, C×26 expression showed an average decrease of 56% (range, 52–64%, n = 3) in total endogenous C×26 expression in KID-KCs following S7 treatment, compared with those treated with an irrelevant control siRNA (si-cont; Figure 4b and c). This change was not detected in S7-treated control-KCs, which showed a slight increase in C×26 expression (range, 5–16%, n = 3). All results indicated that S7 had strong selectivity for the mutant GJB2 c.148G>A allele over the WT allele in the patient KCs, and it had little effect on the normal KCs where only the WT allele was present.

Figure 4. Allele-specific GJB2 knockdown by S7.

Figure 4.

Open in a new tab

(a) The mRNA expression of total GJB2, WT, and MUT GJB2 alleles in control-KCs (Control) and KID-KCs (KID) treated with AS-siRNA S7 was examined by qRT-PCR and compared with UT cells. (b) Total C×26 protein expression in S7-treated cells from three independent IP/IB experiments. (c) The expression was quantified using densitometry. β-actin was used as a loading control and HeLa cells were used as a negative control. A reduction in total C×26 expression was detected in KID-KCs after S7 treatment, but such a change was not detected in control-KCs. **P < 0.01. AS-siRNA, allele-specific small interfering RNA; control-KC, control keratinocyte; IB, immunoblotting; IP, immunoprecipitation; KID-KC, keratitis-ichthyosis-deafness syndrome-derived keratinocyte; MUT, mutant; N.S., not significant; UT, untreated; WT, wild-type.

AS-siRNA treatment reversed aberrant GJ and hemichannel functions in KID-KCs

KID-KCs treated with S7 or si-cont at 50 nM were further analyzed for GJ intercellular communication and hemichannel activity. GJ-mediated intercellular diffusion of neurobiotin tracer was analyzed 24 hours after treatment, using the scrape-loading dye transfer assay. The results showed a 24% increase in neurobiotin diffusion in KID-KCs treated with S7 compared with those treated with si-cont (n = 34 and 37 images, respectively, P < 0.01) (Figure 5b and d), whereas no significant difference in neurobiotin transfer was observed in control-KCs treated with either S7 or si-cont (n = 30 and 35 images, respectively, P > 0.05) (Figure 5a and c). The hemichannel activity in the treated cells was assessed by whole-cell patch clamp. The results showed a decrease of 35% in membrane current density in KID-KCs treated with S7 compared with those treated with si-cont (9.02 ± 1.16 pA/pF, n = 20 cells vs. 5.86 ± 0.43 pA/pF, n = 22 cells, measured +110 mV, P < 0.05), whereas no statistical differences were found in control-KCs following S7 or si-cont treatment (n = 10 cells each group, P > 0.05) (Figure 5f–h). Notably, the current density level in S7-treated KID-KCs was comparable with that in control-KCs. The activity of hemichannels was further examined by neurobiotin uptake assay, which showed a significant decrease in neurobiotin uptake in KID-KCs after S7 treatment (n = 21 and 27 images, P < 0.001), in line with the patch clamp results (Figure 5i and j).

Figure 5. S7 treatment corrected abnormal GJ and hemichannel functions in KID-KCs.

Figure 5.

Open in a new tab

(a, b) SLDT was performed in siRNA-treated KID-KCs (KID) and control-KCs (Cont) to assess GJ activity, (c, d) Analysis of NB transfer (red). Each dot in panels (c) and (d) represents the average NB transfer from a single image. Three independent experiments were carried out and at least 10 images were analyzed from each experiment. Restoration of GJ activity was detected in KID-KCs following S7 treatment. Whole-cell patch clamp and NB uptake were carried out to examine hemichannel activity. (e–g) Records of currents from single cells under the voltage step protocol. (h) The plot of current density against membrane voltage shows correction of hyperactive hemichannels in KID-KCs after S7 treatment. (i) Representative NB (red) uptake images. The nuclei were stained with DAPI (blue). (j) Data analysis shows reversal of aberrantly enhanced NB uptake in KID-KCs. Data are presented as the mean ± SEM. Bar = 200 μm. *P < 0.05; **P < 0.01; ***P < 0.001. Control-KC, control keratinocyte; KID-KC, keratitis-ichthyosis-deafness syndrome-derived keratinocyte; NB, neurobiotin; N.S., not significant; SEM, standard error of the mean; si-cont, control small interfering RNA; SLDT, scrape-loading dye transfer.

These findings suggested functional recovery following the inhibition of mutant GJB2 allele by S7, namely improvement of the defective GJ-mediated cell coupling and reversal of the aberrant nonjunctional hemichannel behavior, including electrical conduction and permeability.

Low-level off-target effects of AS-siRNA S7

Despite promising efficacy data obtained from S7, a general concern in preclinical AS-siRNA studies is off-target effects that may cause unintended alteration in unrelated gene expression (Trochet et al., 2018). To explore comprehensively the specificity of S7, RNA sequencing was carried out S7-treated or nontreated KID-KCs. 26,485 genes from the libraries were mapped to the reference human genome (with 15,802 null- or low-expressed genes), among which only six genes were found to be differentially expressed in S7-treated KID-KCs compared with the nontreated cells (range of fold change: 2.01–2.32), indicating that S7 resulted in mild global effects on the KID-KC transcriptome. To validate the results, the top five upregulated (MMP1, MMP10, MMP9, ANGPTL4, and CXCL5) and downregulated genes (GLB1L2, NSA2, AFAP1L1, GPR137, and TMEM109) were further analyzed by qRT-PCR (Table 1). Control-KCs with or without S7 treatment were run in parallel as additional controls. The results confirmed the upregulation of the MMPs and the downregulation of GPR137 in KID-KCs with comparable levels of fold change (Supplementary Figure S4); however, the MMPs were also found upregulated in S7-treated control-KCs.

Table 1.

Top Five Up- and Downregulated Genes from RNA-Seq Experiments

Top five upregulated genes (KID-KCs, S7 vs. untreated)
Gene Protein Fold Change *P-adj
MMP1 Matrix metalloproteinase-1 2.24, up 9.50E-22
MMP10 Matrix metalloproteinase-10 2.06, up 4.19E-30
MMP9 Matrix metalloproteinase-9 1.95, up 4.90E-65
ANGPTL4 Angiopoietin-like 4 1.93, up 3.57E-28
CXCL5 C-X-C motif chemokine 5 1.74, up 6.26E-45
Top five downregulated genes (KID-KCs, S7 vs. untreated)
Gene Protein Fold Change *P-adj
GLB1L2 Galactosidase Beta 1 Like 2 2.31, down 2.17E-14
NSA2 Ribosome biogenesis homolog 2.16, down 1.99E-68
AFAP1L1 Actin filament-associated protein 1-like 1 2.01, down 1.99E-27
GPR137 G protein-coupled receptor 137 2.00, down 3.96E-09
TMEM109 J Transmembrane Protein 109 1.94, down 3.92E-47
Open in a new tab

Abbreviations: KID-KC, keratitis-ichthyosis-deafness syndrome-derived keratinocyte; RNA-Seq, RNA sequencing;

*

P-adj, adjusted P-value for multiple statistical testing (Benjamini-Hochberg method).

A common cause of siRNA-mediated gene alteration is sequence matching between the seed region of siRNA and the target mRNA (Jackson and Linsley, 2010; Yilmazel et al., 2014). To investigate whether this was an underlying mechanism for the differential expression of the aforementioned genes, we employed the online tool, genome-wide enrichment of seed sequences, to analyze the sequence of S7. A total of 180 genes were found to have sequence matches with S7 (Supplementary Table S2), among which only GPR137 was found, with its 3′-untranslated region and coding sequence complementary to the seed region of either of the S7 strands. This implied that downregulation of GPR137 may have resulted from seed regionedependent off-target effects. The mechanism of alteration of the other genes remains to be clarified.

DISCUSSION

We studied AS-siRNA using KID-KC, an immortalized, patient-derived KC cell line that harbors a heterozygous c.148G>A mutation, representing the genetic state in patients with KID syndrome. Although the use of primary KCs from patient skin biopsy would have been ideal for this study, these cells have a limited lifespan that restricts us from performing multiple experiments. To bypass the restrictions, we used immortalized patient KCs because of their indefinite lifespan and capability of proliferation and differentiation, features of primary KCs (Choi et al., 2017). This was further confirmed in our in vivo skin graft experiments, where regenerated skin grafts from immortalized KID-KCs recapitulated the epidermal architecture of the skin of patients with KID syndrome.

Our immunostaining found lower C×26 expression level in cultured immortalized cells than that in skin tissues. This is not surprising, as monolayer culture contains a dominating proportion of proliferating KCs with low-level C×26 expression (Martin et al., 2014). Despite the lower expression, we were able to show aberrant hemichannel and GJ behavior in KID-KCs, which is in line with data generated from previous in vitro models such as Xenopus oocytes (Lee et al., 2009; Sanchez et al., 2013), HeLa cells (Di et al., 2005), and corneal epithelial cells (Shurman et al., 2005) ectopically expressing the c.148G>A mutant. This suggested that the reduced C×26 expression in vitro is unlikely to influence the interpretation of our results. Collectively, the immortalized, patient-derived model, which recapitulates the genetics, cellular, and histological phenotypes of the condition, has significant advantages over the previous models and hence can serve as a good preclinical model for translational development of new therapeutic approaches.

The action of previously reported approaches to inhibit connexins, including monoclonal antibodies (Xu et al., 2017) and synthetic peptide mimetics (Becker et al., 2012), is mediated by either altering the biophysical property of target connexin channels or modulating interaction between target connexins and their binding partners. In contrast to those approaches, AS-siRNA silences target gene expression by degrading mRNA based on perfect sequence matching, thereby blocking the translation of target protein (Jackson and Linsley, 2010). As AS-siRNA is able to discriminate between mutant and WT mRNA sequences differing by even a single base, we harnessed this technology to develop a targeted therapy for KID syndrome. Our results have confirmed that the lead AS-siRNA, S7, targeted the mutant GJB2 allele in a potent and specific manner while maintaining expression of the WT allele and its protein function, providing strong basis for future translation of the AS-siRNA.

In human KCs, C×26 forms heteromeric and heterotypic channels with other compatible types of connexins (Di et al., 2001). These heterogenous channels have biophysical properties differing from their homogenous counterparts, providing dynamic regulation in response to different stimuli. Recent studies proposed that aberrant interaction with C×43 is an emerging mechanism by which certain C×26 mutants can cause diseases through heteromeric channels (García et al., 2015; Shuja et al., 2016). We performed double immunofluorescence staining in our patient skin tissues and cultured KID-KCs, and the results did not show clear colocalization between C×26 and C×43 (data not shown). As this is our preliminary data, further investigation is required in the future.

The low endogenous expression of C×26 in human KCs (Di et al., 2001; Richard et al., 2002) posed a challenge in our initial attempts of immunoblotting to detect C×26, which showed multiple bands. Issues regarding the presence of multiple bands were also reported by others (Gassmann et al., 2009) and were considered to result from oligomers and protein aggregates of C×26. Our optimized immunoprecipitation and immunoblotting approach allowed enrichment of low-abundant C×26 in cultured patient KCs, leading to successful quantification of siRNA-mediated C×26 knockdown.

Our work provides proof-of-concept for the use of AS-siRNA in targeted therapy for KID syndrome. In the context of patient skin, the AS-siRNAemediated reversal of connexin channel function may possibly improve the disturbed epidermal Ca2+ gradient that is contributed by homomeric or heteromeric channels formed by mutant C×26 (Bosen et al., 2015), thereby leading to improved hyperkeratotic phenotype. Also, because enhanced hemichannel activity has been linked to release of inflammatory cytokines when exposing c.148G>A-expressing KCs to peptidoglycans from an opportunistic pathogen, Staphylococcus aureus (Donnelly et al., 2012), AS-siRNA is also likely to contribute to control of skin infection and inflammation. Furthermore, the c.148G>A mutation has been found in most patients with KID syndrome. Thus, the mutation-targeted AS-siRNA would serve as a potentially effective and safe therapeutic intervention for KID syndrome, the debilitating condition that has no effective specific treatment options at present. Although in vivo delivery of siRNA remains challenging, strategies including nanoparticles (Zheng et al., 2012), penetration enhancers (Hegde et al., 2014), microneedles (Chong et al., 2013), and electroporation (Broderick et al., 2012) have shown promise in topical siRNA delivery into the skin in a noninvasive or minimally invasive manner, causing silencing of target genes. We are currently optimizing a topical delivery platform, and the therapeutic efficacy of the AS-siRNA will be tested in our in vivo human-murine chimeric skin graft model generated using immortalized KID-KCs. If successful, our strategy potentially could be adapted to other skin conditions with dominant mutations.

MATERIALS AND METHODS

KID-KCs

Under a protocol approved by the local ethics committee (12/LO/ 1522), 3-mm punch biopsies from a patient with KID syndrome with the heterozygous c.148G>A mutation and a healthy volunteer donor were obtained with informed written consent. The epidermis was isolated freshly from the biopsies as described previously (Di et al., 2011). The primary KCs were immortalized by transduction with a second-generation, replication-deficient, self-inactivating HIV-1 lentiviral vector (Yáñez-Muñoz et al., 2006) constructed with human papilloma virus type 16 E6/E7 cDNA (Supplementary Figure S3). Immortalized cells were established following serial propagation and thereafter were cultured in the KC culture medium without feeder cells.

Statistical analysis

All data were expressed as the mean ± standard error of the mean. Comparisons of data from qRT-PCR, patch clamp, neurobiotin uptake, and scrape-loading dye transfer experiments were made by Student’s t-test using GraphPad Prism v6.01 (GraphPad Software, San Diego, CA). Differences with a P-value less than 0.05 were considered statistically significant.

Detailed methods for immunostaining, siRNA design, qRT-PCR, immunoprecipitation and immunoblotting, patch clamp, neurobiotin uptake, scrape-loading dye transfer, RNA sequencing, and in vivo skin graft experiments are described in Supplementary Materials and Methods.

Supplementary Material

1

ACKNOWLEDGMENTS

The authors gratefully acknowledge the participation of all patients and families in this study, as well as the National Institute for Health Research Biomedical Research Centre based at Great Ormond Street Hospital for Children NHS Foundation Trust and UCL Institute of Child Health. The views expressed are those of the author(s) and not necessarily those of the NHS, the National Institute for Health Research, or the Department of Health. This work was also supported by Newlife Foundation for Disabled Children and the UCL Bogue Fellowship. The authors are indebted to Ayad Eddaoudi and Dale Moulding (ICH Core Facility) for their expertise in flow cytometry and image analysis. WLD is a Great Ormond Street Hospital Children’s Charity Senior Lecturer. TWW is funded by the National Institutes of Health (grant numbers EY013163 and EY026911). VAK is funded by the Wellcome Trust (grant number WT104076MA).

Abbreviations:

AS-siRNA

allele-specific small interfering RNA

GJ

gap junction

KC

keratinocyte

KID

keratitis-ichthyosis-deafness

KID-KC

keratitis-ichthyosis-deafness syndrome-derived keratinocyte

qRT-PCR

quantitative reverse transcriptase-PCR

si-cont,

control small interfering RNA

siRNA

small interfering RNA

WT

wild-type

Footnotes

CONFLICT OF INTEREST

The authors state no conflict of interest.

SUPPLEMENTARY MATERIAL

Supplementary material is linked to the online version of the paper at www.jidonline.org, and at https://doi.org/10.1016/j.jid.2019.09.022.

Data availability statement

Data sets related to this article can be found at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE131709, an open-source online data repository hosted at Gene Expression Omnibus (GEO).

REFERENCES

  1. Arita K, Akiyama M, Aizawa T, Umetsu Y, Segawa I, Goto M, et al. A novel N14Y mutation in connexin26 in keratitis-ichthyosis-deafness syndrome: analyses of altered gap junctional communication and molecular structure of N terminus of mutated connexin26. Am J Pathol 2006;169:416–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Becker DL, Thrasivoulou C, Phillips AR. Connexins in wound healing; perspectives in diabetic patients. Biochim Biophys Acta 2012;1818: 2068–75. [DOI] [PubMed] [Google Scholar]
  3. Bosen F, Celli A, Crumrine D, Vom Dorp K, Ebel P, Jastrow H, et al. Altered epidermal lipid processing and calcium distribution in the KID syndrome mouse model C×26S17F. FEBS Lett 2015;589:1904–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Broderick KE, Chan A, Lin F, Shen X, Kichaev G, Khan AS, et al. Optimized in vivo transfer of small interfering RNA targeting dermal tissue using in vivo surface electroporation. Mol Ther Nucleic Acids 2012;1: e11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Burns FS. A case of generalized congenital keratoderma with unusual involvement of eyes, ears and nasal and buccal mucous membranes. J Cutan Dis 1915;33:255–60. [Google Scholar]
  6. Caceres-Rios H, Tamayo-Sanchez L, Duran-Mckinster C, De La Luz Orozco M, Ruiz-Maldonado R. Keratitis, ichthyosis, and deafness (KID syndrome): review of the literature and proposal of a new terminology. Pediatr Dermatol 1996;13:105–13. [DOI] [PubMed] [Google Scholar]
  7. Choi M, Park M, Lee S, Lee JW, Cho MC, Noh M, et al. Establishment of immortalized primary human foreskin keratinocytes and their application to toxicity assessment and three dimensional skin culture construction. Biomol Ther (Seoul) 2017;25:296–307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chong RH, Gonzalez-Gonzalez E, Lara MF, Speaker TJ, Contag CH, Kaspar RL, et al. Gene silencing following siRNA delivery to skin via coated steel microneedles: in vitro and in vivo proof-of-concept. J Control Release 2013;166:211–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Churko JM, Laird DW. Gap junction remodeling in skin repair following wounding and disease. Physiology 2013;28:190–8. [DOI] [PubMed] [Google Scholar]
  10. Coggshall K, Farsani T, Ruben B, Mccalmont TH, Berger TG, Fox LP, et al. Keratitis, ichthyosis, and deafness syndrome: a review of infectious and neoplastic complications. J Am Acad Dermatol 2013;69:127–34. [DOI] [PubMed] [Google Scholar]
  11. Di WL, Gu Y, Common JE, Aasen T, O’toole EA, Kelsell DP, et al. Connexin interaction patterns in keratinocytes revealed morphologically and by FRET analysis. J Cell Sci 2005;118:1505–14. [DOI] [PubMed] [Google Scholar]
  12. Di WL, Larcher F, Semenova E, Talbot GE, Harper JI, Del Rio M, et al. Ex-vivo gene therapy restores LEKTI activity and corrects the architecture of netherton syndrome-derived skin grafts. Mol Ther 2011;19:408–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Di WL, Rugg EL, Leigh IM, Kelsell DP. Multiple epidermal connexins are expressed in different keratinocyte subpopulations including connexin 31. J Invest Dermatol 2001;117:958–64. [DOI] [PubMed] [Google Scholar]
  14. Donnelly S, English G, de Zwart-Storm EA, Lang S, van Steensel MA, Martin PE. Differential susceptibility of C×26 mutations associated with epidermal dysplasias to peptidoglycan derived from Staphylococcus aureus and Staphylococcus epidermidis. Exp Dermatol 2012;21:592–8. [DOI] [PubMed] [Google Scholar]
  15. Elfgang C, Eckert R, Lichtenberg-Fraté H, Butterweck A, Traub O, Klein RA, et al. Specific permeability and selective formation of gap junction channels in connexin-transfected HeLa cells. J Cell Biol 1995;129:805–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. García IE, Bosen F, Mujica P, Pupo A, Flores-Muñoz C, Jara O, et al. From hyperactive connexin26 hemichannels to impairments in epidermal calcium gradient and permeability barrier in the keratitis-ichthyosis-deafness syndrome. J Invest Dermatol 2016;136:574–83. [DOI] [PubMed] [Google Scholar]
  17. García IE, Maripillán J, Jara O, Ceriani R, Palacios-Muñoz A, Ramachandran J, et al. Keratitis-ichthyosis-deafness syndrome-associated C×26 mutants produce nonfunctional gap junctions but hyperactive hemichannels when co-expressed with wild type Cx43. J Invest Dermatol 2015;135:1338–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gassmann O, Kreir M, Ambrosi C, Pranskevich J, Oshima A, Röling C, et al. The M34A mutant of connexin26 reveals active conductance states in pore-suspending membranes. J Struct Biol 2009;168:168–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hegde V, Hickerson RP, Nainamalai S, Campbell PA, Smith FJ, Mclean WH, et al. In vivo gene silencing following non-invasive siRNA delivery into the skin using a novel topical formulation. J Control Release 2014;196:355–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jackson AL, Linsley PS. Recognizing and avoiding sirna off-target effects for target identification and therapeutic application. Nat Rev Drug Discov 2010;9:57–67. [DOI] [PubMed] [Google Scholar]
  21. Laird DW. Life cycle of connexins in health and disease. Biochem J 2006;394: 527–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lee JR, Derosa AM, White TW. Connexin mutations causing skin disease and deafness increase hemichannel activity and cell death when expressed in Xenopus oocytes. J Invest Dermatol 2009;129:870–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Levit NA, Sellitto C, Wang HZ, Li L, Srinivas M, Brink PR, et al. Aberrant connexin26 hemichannels underlying keratitis-ichthyosis-deafness syndrome are potently inhibited by mefloquine. J Invest Dermatol 2015;135:1033–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Lopez W, Gonzalez J, Liu Y, Harris AL, Contreras JE. Insights on the mechanisms of Ca(2þ) regulation of connexin26 hemichannels revealed by human pathogenic mutations (D50N/Y). J Gen Physiol 2013;142:23–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Martin PE, Easton JA, Hodgins MB, Wright CS. Connexins: sensors of epidermal integrity that are therapeutic targets. FEBS Lett 2014;588: 1304–14. [DOI] [PubMed] [Google Scholar]
  26. Mazereeuw-Hautier J, Bitoun E, Chevrant-Breton J, Man SY, Bodemer C, Prins C, et al. Keratitis-ichthyosis-deafness syndrome: disease expression and spectrum of connexin 26 (GJB2) mutations in 14 patients. Br J Dermatol 2007;156:1015–9. [DOI] [PubMed] [Google Scholar]
  27. Mese G, Sellitto C, Li L, Wang HZ, Valiunas V, Richard G, et al. The C×26-G45E mutation displays increased hemichannel activity in a mouse model of the lethal form of keratitis-ichthyosis-deafness syndrome. Mol Biol Cell 2011;22:4776–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Patel V, Sun G, Dickman M, Khuu P, Teng JM. Treatment of keratitisichthyosis- deafness (KID) syndrome in children: a case report and review of the literature. Dermatol Ther 2015;28:89–93. [DOI] [PubMed] [Google Scholar]
  29. Press ER, Shao Q, Kelly JJ, Chin K, Alaga A, Laird DW. Induction of cell death and gain-of-function properties of connexin26 mutants predict severity of skin disorders and hearing loss. J Biol Chem 2017;292: 9721–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Richard G, Rouan F, Willoughby CE, Brown N, Chung P, Ryynänen M, et al. Missense mutations in GJB2 encoding connexin-26 cause the ectodermal dysplasia keratitis-ichthyosis-deafness syndrome. Am J Hum Genet 2002;70:1341–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Sanchez HA, Villone K, Srinivas M, Verselis VK. The D50N mutation and syndromic deafness: altered C×26 hemichannel properties caused by effects on the pore and intersubunit interactions. J Gen Physiol 2013;142: 3–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Shuja Z, Li L, Gupta S, Mes̗e G, White TW. Connexin26 mutations causing palmoplantar keratoderma and deafness interact with connexin43, modifying gap junction and hemichannel properties. J Invest Dermatol 2016;136:225–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Shurman DL, Glazewski L, Gumpert A, Zieske JD, Richard G. In vivo and in vitro expression of connexins in the human corneal epithelium. Invest Ophthalmol Vis Sci 2005;46:1957–65. [DOI] [PubMed] [Google Scholar]
  34. Trochet D, Prudhon B, Beuvin M, Peccate C, Lorain S, Julien L, et al. Allele-specific silencing therapy for dynamin 2-related dominant centronuclear myopathy. EMBO Mol Med 2018;10:239–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Trochet D, Prudhon B, Vassilopoulos S, Bitoun M. Therapy for dominant inherited diseases by allele-specific RNA interference: successes and pitfalls. Curr Gene Ther 2015;15:503–10. [DOI] [PubMed] [Google Scholar]
  36. van Steensel MA, van Geel M, Nahuys M, Smitt JH, Steijlen PM. A novel connexin 26 mutation in a patient diagnosed with keratitis-ichthyosis-deafness syndrome. J Invest Dermatol 2002;118:724–7. [DOI] [PubMed] [Google Scholar]
  37. Xu L, Carrer A, Zonta F, Qu Z, Ma P, Li S, et al. Design and characterization of a human monoclonal antibody that modulates mutant connexin 26 hemichannels implicated in deafness and skin disorders. Front Mol Neurosci 2017;10:298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Yáñez-Muñoz RJ, Balaggan KS, Macneil A, Howe SJ, Schmidt M, Smith AJ, et al. Effective gene therapy with nonintegrating lentiviral vectors. Nat Med 2006;12:348–53. [DOI] [PubMed] [Google Scholar]
  39. Yilmazel B, Hu Y, Sigoillot F, Smith JA, Shamu CE, Perrimon N, et al. Online GESS: prediction of miRNA-like off-target effects in large-scale rnai screen data by seed region analysis. BMC Bioinformatics 2014;15:192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Yum SW, Zhang J, Valiunas V, Kanaporis G, Brink PR, White TW, et al. Human connexin26 and connexin30 form functional heteromeric and heterotypic channels. Am J Physiol Cell Physiol 2007;293: C1032–48. [DOI] [PubMed] [Google Scholar]
  41. Zheng D, Giljohann DA, Chen DL, Massich MD, Wang XQ, Iordanov H, et al. Topical delivery of siRNA-based spherical nucleic acid nanoparticle conjugates for gene regulation. Proc Natl Acad Sci USA 2012;109: 11975–80. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1613077-supplement-1.pdf (968KB, pdf)

RESOURCES