Skip Navigation


Human Molecular Genetics Advance Access originally published online on December 1, 2006
Human Molecular Genetics 2007 16(2):165-172; doi:10.1093/hmg/ddl452
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
16/2/165    most recent
ddl452v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Unsworth, H. C.
Right arrow Articles by Kelsell, D. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Unsworth, H. C.
Right arrow Articles by Kelsell, D. P.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Tissue-specific effects of wild-type and mutant connexin 31: a role in neurite outgrowth

Harriet C. Unsworth1,{dagger}, Trond Aasen1,{dagger}, Suzanne McElwaine2 and David P. Kelsell1,*

1 Centre for Cutaneous Research and 2 Department of Experimental Haematology, Institute of Cell and Molecular Science, Queen Mary's School of Medicine and Dentistry, University of London, Whitechapel, E1 4AT, UK

* To whom correspondence should be addressed at: Centre for Cutaneous Research, Institute of Cell and Molecular Science, Queen Mary's School of Medicine & Dentistry, University of London, 4 Newark Street, Whitechapel, London, UK. Fax: , +44 2078827171; Email: d.p.kelsell{at}qmul.ac.uk

Received September 28, 2006; Revised November 16, 2006; Accepted November 24, 2006


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Channels formed by connexins (Cx), the major protein subunits of gap junctions, allow passage of ions and molecular messengers between cells to provide a mechanism of synchronized cellular response. Twenty human Cx isoforms have been identified and mutations in the gene GJB3 encoding the 31 kDa isoform, Cx31, can cause dominant or recessive skin disease, dominant or recessive deafness or dominant neuropathy with deafness. Cx31 is expressed in differentiating keratinocytes in skin. Here, we also demonstrate endogenous Cx31 expression in human neuronal cell lines, particularly in differentiated neurones. Exogenous Cx31 expression induced neurite outgrowth in human neuronal cell lines, but not differentiation in primary human keratinocytes. Though neither the neuropathy and hearing loss mutation (66delD)Cx31 nor the skin disease associated mutation (R42P)Cx31 is able to traffic to the plasma membrane, the R42P mutant induced neurite outgrowth to a level equal to wild-type Cx31. In contrast, there was significantly reduced neurite outgrowth after (66delD)Cx31 expression. In addition to indicating a potential disease mechanism for the neuropathy/deafness mutation, this work demonstrates a tissue-specific function for Cx31.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
The ability of cells to communicate with each other is of vital importance in maintenance of normal tissue homeostasis. Gap junctions composed of connexins can provide this mechanism of synchronized cellular response facilitating the metabolic and ion exhange functions of the cell. The association of human genetic disease with specific connexin mutations including progressive neuropathy, sensorineural hearing loss and hyperproliferative skin disease has established a major role for gap junctions in a diverse range of physiological processes (1,2). Dominant and recessive mutations in the gene GJB3 encoding Cx31 can cause the skin disease erythrokeratoderma variabilis (EKV) (3,4). Dominant GJB3 mutations have also been described associated with progressive hearing loss with no epidermal manifestations (5). To add further complexity to the biology of Cx31 and its role in disease is association of another dominant GJB3 mutation, 66delD, with peripheral neuropathy and sensorineural hearing loss (6). These genetic studies suggest that different mutations in the Cx31 protein have distinct effects on epidermal differentiation, auditory transduction and peripheral neuronal function. For example, transfection of skin disease associated Cx31 mutations into keratinocytes induces cell death, whereas deafness and neuropathy mutations do not (7). In this study, we have investigated the cellular functions of wild-type and mutant Cx31 in neuronal cells.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
The SH-SY5Y cell line is a human epithelial cell derived neuroblastoma cell line, consisting of neuroblastic cells with multiple, fine, short cell processes (8). These cells were differentiated into human neuron-like cells, with sequential treatment of all-trans retinoic acid and brain-derived neurotrophic factor (RA/BDNF) as described previously (9). Differentiation in these cells was accompanied by a characteristic change in cell morphology, with the growth of long, branched neuronal processes from the cell body [a widely used marker for neuronal differentiation (9–11)]. This was accompanied by a decrease of cytoplasmic medium neurofilament (NF-M) as previously reported (9) with a redistribution of NF-M from the cytoplasm to the perinuclear area and the base of growing neuronal processes. The expression of human Cx isoforms was examined by immunofluorescence in the SH-SY5Y cell line, pre- and post-RA/BDNF differentiation treatment (Fig. 1). Low or absent expression of Cx26 (Fig. 1A and B), Cx30.3, Cx32, Cx40, Cx43 and Cx45 (data not shown) was seen in differentiated and undifferentiated neuronal cells. In undifferentiated cells, Cx30 and Cx31 were expressed in the cytoplasm and Cx31 was seen at low levels in the plasma membrane (Fig. 1C and E). Following induced differentiation treatment, Cx30 could be seen in the cell membrane and in the processes that extended from the cell body (Fig. 1D). Cx31 was strongly expressed both in the cytoplasm and at the cell membrane where bright, punctate GJ plaque-like protein aggregates were clearly visible at the plasma membrane and between the cell processes at points of cell–cell contact (Fig. 1F).


Figure 4521
View larger version (104K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. Immunofluorescence analysis of connexin expression (green) in undifferentiated (A, C, E) and differentiated (B, D, F) SH-SY5Y neuronal cells. No expression of Cx26 was observed (A and B). Low levels of Cx30 (C) and Cx31 (E) were expressed in undifferentiated SH-SY5Y cells and the expression of these proteins is increased upon differentiation with more punctate staining at the membrane for both Cx30 (D) and Cx31 (F). This is accompanied by a characteristic change in morphology, with the growth of neurites from the cell body (40x).

 
The increase in Cx31 and Cx30 expression in SH-SY5Y cells after differentiation suggested that one or both of these Cx may play an important role in neuronal differentiation. To test this hypothesis, pSin-(WT)Cx31–EGFP and pSin-(WT)Cx30–EGFP constructs were transfected into SH-SY5Y cells (Fig. 2A–C). Cells transfected with the empty pSin vector had little effect on the cell morphology (Fig. 2A). The cells transfected with (WT)Cx30–EGFP maintained the more rounded cell shape of untransfected cells (Fig. 2B). In contrast, SH-SY5Y cells transfected with (WT)Cx31–EGFP showed the characteristic cell shape of differentiated cells similar to that of the differentiated morphology of neurite processes following RA/BDNF treatment (Fig. 2B). We also transfected rat PC12 neuronal and saw a similar pattern, although not as extensive (data not shown). PC12 cells transfected with (WT)Cx31–EGFP showed changes in cell shape, with longer cell processes that showed some branching. The cell body also changed, becoming less rounded and more irregular in shape. Our results indicate that the expression of Cx31–EGFP induce neurite outgrowth in SH-SY5Y cells (and to a lesser extent in PC12 cells), a key feature of neuronal differentiation. Previously, Cx32 and Cx43 have been shown to effect neurite outgrowth in PC12 cells (11). Although neurite outgrowth is a common measure of neuronal outgrowth, we also looked at several markers of neuronal differentiation. We saw little change in expression by western blotting but this may, in part, be due to the low level of transfection efficiency (data not shown). One marker, NF-M, is reduced upon neuronal differentiation. Using immunofluorescence, we observed a reduction in NF-M expression in cells expressing Cx31–EGFP post-transfection (Fig. 3), consistent with neurite outgrowth and neuronal differentiation. It is thus likely that the expression of Cx31 mediates at least certain aspects of neuronal differentiation.


Figure 4522
View larger version (66K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2. Connexin transfection of Cx31 but not Cx30 induces neurite outgrowth. Green denotes pSin construct expression, red denotes Tubulin expression and blue represents nuclear DAPI staining. Transfection of SH-SY5Y cells with Cx31–EGFP induced increased spontaneous neurite outgrowth (C) compared with EGFP (A) and Cx30 (B). Transfection of (66delD)Cx31–EGFP (D), (R42P)Cx31–EGFP (E) and (C86S)Cx31–EGFP (F) in SH-SY5Y neuronal cells shows that the skin disease mutants are equally effective at inducing neurite outgrowth as wild-type Cx31, whereas there is a reduction in spontaneous neurite outgrowth in cells expressing the neuropathy mutant 66delD compared with EGFP or Cx30, summarized in Figure 5 (40x).

 


Figure 4523
View larger version (36K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 3. SH-SY5Y cells were transfected with Cx31–EGFP (shown in green in merged image, A) and immunofluorescence was performed to detect NF-M expression (red in merged image, A). Cells expressing Cx31–EGFP (A and B: 1–4, arrowed) all have very little NF-M expression (A and C) consistent with the finding that this neurofilament is reduced (and replaced by others) during neuronal differentiation (40x).

 
We have previously shown in skin that Cx31 is expressed at low levels in basal and spinous layers of the epidermis and at high levels in differentiated keratinocytes in the granular layer (12). In order to see whether exogenous expression of Cx31 in the basal layer of the epidermis would affect keratinocyte differentiation, we infected primary human keratinocytes (passage 1 or 2) with retroviral construct driving Cx31–EGFP chimeric constructs from a cytomegalovirus (CMV) promoter (80–90% transduction efficiency). Cells were seeded on a de-epidermalized dermis and raised to the air–liquid interphase for 3–12 days in order to induce epidermal differentiation similar to that seen in vivo (13). In contrast to the effect seen in neuronal cells, no striking difference in keratinocyte differentiation upon induced Cx31 expression was observed by haematoxylin and eosin (H&E) staining nor using markers of differentiation, such as involucrin, K1, K10 and loricrin (Fig. 4). This suggests that Cx31 may have tissue-specific functions.


Figure 4524
View larger version (106K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 4. Primary human keratinocytes (passage 1) were retrovirally transduced with pSin–EGFP (A and B) and pSin-Cx31–EGFP (C and D) vectors and cultured on de-epidermilized dermis raised at the air–liquid interphase for 8 days to induce differentiation. No obvious difference was observed by H&E staining (A and C) or differentiation markers such as Fillagrin (B and D in red) (20x).

 
Three dominant disease-associated Cx31 mutants were transfected into SH-SY5Y cells to investigate their influence on Cx31-induced neuronal differentiation: (66delD)Cx31 mutation is associated with neuropathy and hearing loss and (R42P)Cx31 and (C86S)Cx31 are associated with the skin disease EKV (6,14) (Fig. 2D–F). The degree of neuronal differentiation associated with transfection of the mutant and wild-type constructs was measured using a neurite outgrowth assay (15) (Fig. 5). The results of the neuronal outgrowth assay confirms previous work (13) that non-transfected, naive SH-SY5Y cells show a degree of spontaneous differentiation, with 43% showing differentiation-like neurite growth. Control transfected cells carrying the empty pSin–EGFP vector showed a similar level of differentiation (36.2%), as did cells transfected with pSin(WT)Cx30–EGFP (36.6%). As expected from previous transfection experiments, pSin(WT)Cx31–EGFP-transfected cells showed higher numbers of differentiated cells (60%). pSin(66delD)Cx31–EGFP inhibited differentiation to less than half of that in non-transfected cells, with only 20% of transfected cells showing the differentiated morphology. Surprisingly, the similar non-trafficking mutant pSin(R42P)Cx31–EGFP transfected cells showed a higher level of differentiation than cells transfected with the (WT)Cx31 construct at 63% differentiation. Another non-trafficking skin disease (C86S)Cx31 also induced neurite outgrowth (Fig. 2F). In conclusion, this suggested that Cx31 induced neuronal differentiation via a non-gap junction associated mechanism associated with a cytoplasmic pool of Cx31. Another observation was that (R42P)Cx31 and (C86S)Cx31 did not induce neuronal cell death even though a cell death phenotype occurs in epithelial cells further supporting tissue-specific processing of mutant Cx31 (7).


Figure 4525
View larger version (32K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 5. Graphic display of neurite outgrowth in SH-SY5Y cells transfected with various constructs. Only cells expressing a construct where counted in each field. Expression of (WT)Cx31–EGFP and (R42P)Cx31–EGFP significantly increased spontaneous neurite outgrowth compared with control constructs, whereas the neuropathy mutant 66delD caused a reduction in spontaneous neurite outgrowth.

 
SH-SY5Y cells had very low endogenous Cx expression, but cell differentiation (induced by treatment with RA and BDNF) brought about an increase in the expression of both Cx30 and Cx31. Gap junctional coupling in these cells was measured in untreated and pSin-(WT)Cx31–EGFP retrovirally transduced SH-SY5Y cells by dye transfer assay (Fig. 6). Coupling was found to be similar and relatively low in these cells in either their undifferentiated or Cx31-induced differentiated state, lending further support to a gap junction-independent role for Cx31 in neuronal differentiation.


Figure 4526
View larger version (113K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 6. SH-SY5Y cells were retrovirally transduced with pSin–EGFP (A and C) and pSin-Cx31–EGFP (B and D) vectors. At 100% cell confluence, a solution containing 100µM Alexa A-350 dye (350 kDa, blue) and Alexa A-568 dye (731 kDa, red) was added and cells were scraped with a scalpel and incubated for 5 min before three times PBS washes and fixation in 4% paraformaldehyde (A and B). Both cells were relatively poorly coupled (compared with other neuronal and keratinocyte cells tested). Some A-568 dye transfer was noted, whereas the smaller A-350 dye spread to a further few cells. No transfer was seen in scrapeloadings using control fluorescent dextran (MW 10 000). (C) and (D) represents phase contrast overlay of (A) and (B) to demonstrate scrape and 100% confluency (20x).

 
To see whether these cell morphology changes correlated with changes in molecular gene expression signatures, microarray gene expression was performed comparing the gene expression profiles of (WT)Cx31 and (66delD)Cx31 expressed in SH-SY5Y cells by retroviral transduction (estimated to infect ~90% of cells). No analysis was performed with (R42P)Cx31 or (C86S)Cx31 as these constructs induced cell death in the Phoenix packaging cell line used to generate the retrovirus. Following analysis of the Hu133A arrays using Genespring, a number of genes (103) were identified as being differentially expressed in cells expressing the pSin(66delD)Cx31–EGFP construct when compared with wild-type and EGFP controls (Fig. 7). Supervized clustering of the samples using the differentially expressed gene list highlights a group of genes (29 genes, coloured red) with a >2-fold increase in gene expression. In addition, genes with decreased expression in 66delD were also identified (74 genes, coloured blue).


Figure 4527
View larger version (61K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 7. With the use of Genespring analysis software, data analysis was performed on human 133A microarray data (Affymetrix). A number of genes were identified as being differentially expressed, by at least 2-fold, between the differentiated and undifferentiated phenotypes. Statistical analysis of these differentially expressed genes (one-way ANOVA) resulted in 103 genes that exhibited differential expression in 66delD on comparison with Cx31 and EGFP control (yellow, similar expression levels; red, increased expression; blue, decreased expression relative to Cx31 and EGFP). A number of the differentially expressed genes are associated with neuronal differentiation. Supervized clustering was performed using Genespring. First, an experiment tree was generated using a Spearman correlation and subsequently a gene tree was obtained using a Pearson correlation—the resulting tree is illustrated above.

 
Additionally, we performed pathway architect analysis of (WT)Cx31–EGFP versus EGFP and (66delD)Cx31–EGFP versus EGFP in order to identify hypothetical pathways regulated by wild-type versus mutant Cx31. As seen in Figure 8 and Table 1, a number of pathways that were shared or expressed uniquely by WT/EGFP and 66delD/EGFP were identified.


Figure 4528
View larger version (22K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 8. Pathway architect analysis of microarray data. Genes deregulated 2-fold or more in SH-SY5Y cells expressing wild-type Cx31–EGFP versus EGFP alone, or from cells expressing 66DelD-Cx31–EGFP versus EGFP, were imported into Pathway Architect (Stratagene) for analysis. Using a relevance interaction algorithm, pathways were generated, whereby it is thought likely that genes deregulated in the microarray analysis significantly associates with this pathway. This example shows the pathways generated by a relevance interaction network of binding factors. Some pathways (in blue) are shared between wild-type and mutant Cx31, for example, HNF4A. Other pathways are unique (in black) such as TAF1 where more than 40 genes deregulated ≥2-fold are associated with this binding factor in Cx31 versus EGFP cells, whereas this pathway is not present at all in 66delD versus EGFP cells. Genes in green are downregulated in the microarray analysis, whereas red denotes genes upregulated (colour is gradient-dependent).

 


View this table:
[in this window]
[in a new window]

  		Table 1. Pathway architect analysis of microarray data

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
The data presented in this study demonstrate a role for Cx31 in enhancing neurite outgrowth, a key feature of neuronal differentiation, related to both endogenous and exogenous Cx31 expression. In this study, we provide evidence that Cx31 has a function beyond intercellular channel formation to encompass multifunctional cellular roles in a tissue-specific manner. The expression of Cx31 is unlikely to be essential for neuronal differentiation, however, as recessive deafness Cx31 mutations do not cause neuropathy and Cx31 KO mice show no abnormal neuronal development or behaviour (16,17). Furthermore, we have unpublished data using shRNA-mediated silencing of Cx31 where no difference in neurite outgrowth was observed in SH-SY5Y cells. Although we did not observe a significant increase in neurite outgrowth after expression of Cx30, other connexins may compensate for Cx31 in its absence, such as Cx32 and Cx36, as seen with some connexins in various tissues including skin. The reduction in neurite outgrowth after expression of (66delD)Cx31 may be due to a transdominant interaction with other connexins or indeed other Cx31-interacting proteins regulating this process.

Towards identifying the molecular changes associated with Cx31 expression, gene array analyses were performed and revealed a change in global gene expression between (WT)Cx31 and the deafness/neuropathy-associated mutation, (66delD)Cx31. Further studies are required to fully characterize the signalling pathways and cell processes mediated by Cx31. A number of genes involved in actin re-organization were decreased in the 66delD-transduced cells, such as cofilin-1 (3-fold decreased) and nebulette (2.2-fold decreased), indicating potential de-regulation of components of the cytoskeletal pathway which may account for the block in differentiation. Furthermore, genes such as neurofilament light peptide (NEFL) and growth differentiation factor 11 (bone morphogenic protein-11) were at least 5-fold decreased in 66delD-transduced cells. The former is involved in filament formation in neuronal cells and the latter has a major role in mesoderm differentiation and axial skeleton patterning.

Neuronal differentiation is associated with major remodelling of the actin cytoskeleton in order to enable the growth of neurites from the cell body (18–20). The major morphological change associated with neuronal differentiation requires not only changes to the actin cytoskeleton, but also an increase in plasma membrane area to accommodate the growth of neurites. This rapid expansion has been estimated to be in the order of 0.5 µm2/min per mammalian neurite (21). A role for actin in the formation of functional Cx31 GJ channels has recently been suggested; chemical disruption of actin filaments in HeLa cells transfected with Cx31 prevents the formation of functional channels at the plasma membrane (22). In addition, Cx31 could be able to bind actin via a protein intermediate. Cx43 has been shown to be able to bind to drebrin, an actin-binding protein that could serve to link gap junctions to the actin cytoskeleton, and drebrin depletion by small interfering RNA was shown to reduce Cx43 GJ expression and coupling (23).

Previously, functional studies on connexins have focused on their role in forming functional gap junctions and later as hemichannels (24). The association of human diseases associated with specific Cxs have provided useful molecular tools to dissect the function of connexins in different tissues. Specifically, mutations in Cx31 have revealed tissue-specific processing and distinct cell phenotypes of skin disease or deafness–neuropathy-associated mutations (7,25,26). In this study, we have extended these studies and identified a non-gap junction function of Cx31 in neurite outgrowth.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Cell culture
SH-SY5Y cells were cultured in 5% CO2 at 37°C in 1:1 Ham's F12:DMEM media supplemented with 10% FCS. PC12 cells were cultured in 5% CO2 at 37°C in E4 media supplemented with 10% FCS and 1 x streptomycin/penicillin.

Transfection of neuronal cells
Cells were seeded at 2 x 105 per 13 mM cover slip in 24 well plates, 24 h prior to transfection. Cells were transfected with the following constructs, using fugene lipofection: pSin(WT) Cx30–EGFP, pSin(WT)Cx31–EGFP, pSin(del66D)Cx31–EGFP, pSin(R42P)Cx31–EGFP and pSin(C86S)Cx31–EGFP. For each cover slip, transfection mix containing 0.6 µl Fugene and 0.2 µg DNA construct and made up to 20 µl with serum-free culture medium was prepared. Cells were incubated for 48 h, then fixed in 4% paraformaldehyde for 90 min. Cells were stained using mouse {alpha}-NF-M antibody (Zymed, USA) and Alexa Fluor 546 rabbit {alpha}-mouse (Molecular Probes, USA) and counterstained with DAPI [2-(4-amidinophenyl)-6-indolecarbamidine dihydrochloride, Sigma] at 1 µg/ml to visualize cell nuclei. Similar expression levels (driven from CMV promoter) between the constructs were noted.

Neurite outgrowth assay
Differentiation was assessed in SH-SY5Y cells using the previously reported neurite outgrowth assay (9). Cells with one or more neurite longer than the length of the cell body were counted as differentiated. Differentiation is expressed as percentage transfected cells showing differentiated morphology of total tranfected cells per field of view. Fifteen fields of view were counted from each of three separate transfection experiments.

Immunofluorescence
Endogenous Cx expression was examined in SH-SY5Y cells using in-house rabbit anti-Cx31 and rabbit anti-Cx30 antibodies, with Alexa-488 labelled goat anti-rabbit fluorescent secondary antibodies (Molecular Probes). Cells were fixed for 90 min in 4% paraformaldehyde, then washed in phosphate-buffered saline (PBS) before 15 min permeabilization with 0.1% Triton X-100 (Sigma) and blocking in 0.2% fish-skin gelatine. Cells were incubated with primary antibodies diluted to 1:100 for 1 h at 37°C, then washed three times for 10 min with PBS and incubated with secondary antibody diluted to 1:200 for 1 h at room temperature in darkness. Cells were washed three times with PBS before 5 min incubation with DAPI, followed by three further PBS washes. Cover slips were mounted onto slides using immunomount before examination by fluorescent microscopy.

Retroviral transduction of cells
In order to produce retrovirus containing the pSin–Cx constructs, phoenix packaging cells were transfected with the relevant constructs (as described above), and at 24 h after transfection, the cells were moved to a 32°C incubator, where retrovirus was released into the culture medium. The resulting retrovirus-containing media were incubated with the target cells for 48 h at 37°C in order to achieve retroviral transduction of the target cells. Efficiency of transduction was estimated to be 80–90% after 48 h of incubation.

Dye transfer assay
Cell coupling was measured using scrape loading. At 100% cell confluence, a solution containing 100 µM Alexa A-350 dye (350 kDa, blue) and 100 µM Alexa A-568 dye (731 kDa, red) was added and cells were scraped with a scalpel and incubated for 5 min before three times PBS washes and fixation in 4% paraformaldehyde. No transfer was seen in scrapeloadings using control fluorescent dextran (MW 10 000). Extensive dye transfer was seen in control cells (HaCaT keratinocytes). All probes were obtained from Invitrogen and images were taken on a Nikon inverted fluorescent microscope using Metamorph software.

Microarray analysis
Total RNA was isolated from retrovirally transduced SH-SY5Y cells, transduced with either Cx31-66delD–EGFP, Cx31–EGFP or EGFP alone. Extraction was performed using the QIAGEN RNAeasy mini kit. RNA quantity and purity were assessed using a Biophotometer (Eppendorf). Sample preparation for Affymetrix analysis was performed according to Manufacturer's protocol. Five micrograms of total RNA was reverse transcribed to double-stranded cDNA and subsequently an in vitro transcription was performed at 37°C for 16 h (biotinylated cRNA). Fragmentation of biotinylated cRNA (20 µg) was performed at 94°C for 35 min and hybridization cocktails were prepared according to manufacturer's instructions. Hybridizations were performed at 45°C for 16 h, and subsequently the arrays were washed and stained using streptavidin phycoerythrin and the EukGE Ws2v4 protocol. Samples were hybridized to TEST3 chips (Affymetrix) to assess the quality of starting material. The 3'/5' GAPDH ratios for all samples were <3.0 (mean 0.91, range 0.84–0.96) and sensitivity controls (BIO B–D) were detected in all samples. Having rigorously assessed the quality of the RNA samples, they were hybridized to human 133A arrays (Affymetrix). These arrays have more than 22 000 genes represented on the chip surface. All scanned images were analysed using Microarray suite (MAS 5.0) and the present/marginal/absent calls were assigned. A target intensity value of 100 was used for all samples and scaling factors were within 3-fold of each other as per manufacturer's instructions. MIAME guidelines were adhered to wherever possible.

All data files were imported into and normalized (Per Chip, Per Gene) within Genespring 6.1. The normalized data were analysed according to phenotype, i.e. differentiated (Cx31, GFP alone) versus undifferentiated (66delD). Data files from the 133A arrays were ‘filtered on flags’ in two ways: first, by removing all genes called ABSENT in all samples and secondly by identifying genes called Present in at least 30% of samples. This latter list of ‘Changing genes’ contained 10 852 genes and was subsequently used in fold change interpretations and statistical analysis in Genespring (Welch's t-test). Supervized clustering was performed in Genespring using the 103 genes that were a minimum of 2-fold differentially expressed and statistically different between the ‘undifferentiated’ and ‘differentiated’ phenotype (experiment tree: Spearman correlation, gene tree: Pearson correlation; Fig. 7).

For pathway architect analysis (http://www.stratagene.com), genes with a minimum of 2-fold differential expression were imported into the software (493 genes in Cx31 versus EGFP, and 448 genes in del66D versus EGFP). All genes were selected and standard relevance interaction pathways were performed to generate hypothetical pathways that may be significantly different between wild-type and mutant Cx31 (Table 1 and Fig. 8).


   ACKNOWLEDGEMENTS
 
This work was supported by a BBSRC CASE studentship and a BBSRC project grant (D.P.K.).

Conflict of Interest statement. None declared.


   FOOTNOTES
 
{dagger} The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Kelsell D.P., Dunlop J., Hodgins M.B. (2001) Human diseases: clues to cracking the connexin code? Trends Cell Biol. 11:2–6.[CrossRef][Web of Science][Medline]

  2. Laird D.W. (2006) Life cycle of connexins in health and disease. Biochem. J. 394:527–543.[CrossRef][Web of Science][Medline]

  3. Richard G., Smith L.E., Bailey R.A., Itin P., Hohl D., Epstein E.H. Jr, DiGiovanna J.J., Compton J.G., Bale S.J. (1998) Mutations in the human connexin gene GJB3 cause erythrokeratodermia variabilis. Nat. Genet. 20:366–369.[CrossRef][Web of Science][Medline]

  4. Gottfried I., Landau M., Glaser F., Di W.L., Ophir J., Mevorah B., Ben-Tal N., Kelsell D.P., Avraham K.B. (2002) A mutation in GJB3 is associated with recessive erythrokeratodermia variabilis (EKV) and leads to defective trafficking of the connexin 31 protein. Hum. Mol. Genet. 11:1311–1316.[Abstract/Free Full Text]

  5. Xia J.H., Liu C.Y., Tang B.S., Pan Q., Huang L., Dai H.P., Zhang B.R., Xie W., Hu D.X., Zheng D., et al. (1998) Mutations in the gene encoding gap junction protein beta-3 associated with autosomal dominant hearing impairment. Nat. Genet. 20:370–373.[CrossRef][Web of Science][Medline]

  6. Lopez-Bigas N., Olive M., Rabionet R., Ben-David O., Martinez-Matos J.A., Bravo O., Banchs I., Volpini V., Gasparini P., Avraham K.B., et al. (2001) Connexin 31 (GJB3) is expressed in the peripheral and auditory nerves and causes neuropathy and hearing impairment. Hum. Mol. Genet. 10:947–952.[Abstract/Free Full Text]

  7. Di W.L., Monypenny J., Common J.E., Kennedy C.T., Holland K.A., Leigh I.M., Rugg E.L., Zicha D., Kelsell D.P. (2002) Defective trafficking and cell death is characteristic of skin disease- associated connexin 31 mutations. Hum. Mol. Genet. 11:2005–2014.[Abstract/Free Full Text]

  8. Biedler J.L., Helson L., Spengler B.A. (1973) Morphology and growth, tumorigenicity, and cytogenetics of human neuroblastoma cells in continuous culture. Cancer Res. 33:2643–2652.[Abstract/Free Full Text]

  9. Mario Encinas I.M., Liu Y., Wang H., Ashraf M., Cena V., Gallego C., Comella J.X. (2000) Sequential treatment of SH-SY5Y cells with retinoic acid and brain-derived neurotrophic factor gives rise to fully differentiated, neurotrophic factor dependent, human neuron-like cells. J. Neurochem. 75:991–1003.[CrossRef][Web of Science][Medline]

  10. Heo K., Ha S.H., Chae Y.C., Lee S., Oh Y.S., Kim Y.H., Kim S.H., Kim J.H., Mizoguchi A., Itoh T.J., et al. (2006) RGS2 promotes formation of neurites by stimulating microtubule polymerization. Cell Signal. 18:2182–2192.[CrossRef][Web of Science][Medline]

  11. Belliveau D.J., Bani-Yaghoub M., McGirr B., Naus C.C., Rushlow W.J. (2006) Enhanced neurite outgrowth in PC12 cells mediated by connexin hemichannels and ATP. J. Biol. Chem. 281:20920–20931.[Abstract/Free Full Text]

  12. Di W.L., Rugg E.L., Leigh I.M., Kelsell D.P. (2001) Multiple epidermal connexins are expressed in different keratinocyte subpopulations including connexin 31. J. Invest. Dermatol. 117:958–964.[CrossRef][Web of Science][Medline]

  13. Ojeh N.O., Frame J.D., Navsaria H.A. (2001) In vitro characterization of an artificial dermal scaffold. Tissue Eng. 7:457–472.[CrossRef][Web of Science][Medline]

  14. Wilgoss A., Leigh I.M., Barnes M.R., Dopping-Hepenstal P., Eady R.A., Walter J.M., Kennedy C.T., Kelsell D.P. (1999) Identification of a novel mutation R42P in the gap junction protein beta-3 associated with autosomal dominant erythrokeratoderma variabilis. J. Invest. Dermatol. 113:1119–1122.[CrossRef][Web of Science][Medline]

  15. Canals M., Angulo E., Casado V., Canela E.I., Mallol J., Vinals F., Staines W., Tinner B., Hillion J., Agnati L., et al. (2005) Molecular mechanisms involved in the adenosine A and a receptor-induced neuronal differentiation in neuroblastoma cells and striatal primary cultures. J. Neurochem. 92:337–348.[CrossRef][Web of Science][Medline]

  16. Liu X.Z., Xia X.J., Xu L.R., Pandya A., Liang C.Y., Blanton S.H., Brown S.D., Steel K.P., Nance W.E. (2000) Mutations in connexin31 underlie recessive as well as dominant non-syndromic hearing loss. Hum. Mol. Genet. 9:63–67.[Abstract/Free Full Text]

  17. Plum A., Winterhager E., Pesch J., Lautermann J., Hallas G., Rosentreter B., Traub O., Herberhold C., Willecke K. (2001) Connexin31-deficiency in mice causes transient placental dysmorphogenesis but does not impair hearing and skin differentiation. Dev. Biol. 231:334–347.[CrossRef][Web of Science][Medline]

  18. Tojima Takuro E.I. (2004) Signal transduction cascades underlying de novo protein synthesis required for neuronal morphogenesis in differentiating neurons. Prog. Neurobiol. 72:183–193.[CrossRef][Web of Science][Medline]

  19. Cotrina M.L., Lin J.H., Alves-Rodrigues A., Liu S., Li J., Azmi-Ghadimi H., Kang J., Naus C.C., Nedergaard M. (1998) Connexins regulate calcium signaling by controlling ATP release. Proc. Natl Acad. Sci. USA 95:15735–15740.[Abstract/Free Full Text]

  20. Cotrina M.L., Lin J.H., Nedergaard M. (1998) Cytoskeletal assembly and ATP release regulate astrocytic calcium signaling. J. Neurosci. 18:8794–8804.[Abstract/Free Full Text]

  21. Pfenninger K.H.F. (1981) Lectin labeling of sprouting neurons II. Relative movement and appearance of glycoconjugates during plasmalemmal expansion. J. Cell Biol. 89:547–559.[Abstract/Free Full Text]

  22. Qiang He Li C.F., Yu Liu, Jun Liu Mu, Ping Tan Zhi, qian Pan, Yan Fang Fai, Sheng Liang De, Qian Wu Ling, Gao Long Zhi, Ping Dai He, et al. (2005) Cx31 is assembled and trafficked to cell surface by ER–Golgi pathway and degraded by proteosomal or lysosomal pathways. Cell Res. 15:455–464.[CrossRef][Web of Science][Medline]

  23. Butkevich Eugenia H.S., Dirk Wenxel, Tomoaki Shirao, Rainer Duden, Irina Majoul. (2004) Drebrin is a novel connexin-43 binding partner that links gap junctions to the submembrane cytoskeleton. Curr. Biol. 14:650–658.[CrossRef][Web of Science][Medline]

  24. Goodenough D.A. and Paul D.L. (2003) Beyond the gap: functions of unpaired connexon channels. Nat. Rev. Mol. Cell Biol. 4:285–294.[CrossRef][Web of Science][Medline]

  25. Diestel S., Richard G., Doring B., Traub O. (2002) Expression of a connexin31 mutation causing erythrokeratodermia variabilis is lethal for HeLa cells. Biochem. Biophys. Res. Commun. 296:721–728.[CrossRef][Web of Science][Medline]

  26. Abrams C.K., Freidin M.M., Verselis V.K., Bargiello T.A., Kelsell D.P., Richard G., Bennett M.V., Bukauskas F.F. (2006) Properties of human connexin 31, which is implicated in hereditary dermatological disease and deafness. Proc. Natl Acad. Sci. USA 103:5213–5218.[Abstract/Free Full Text]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Hum Mol GenetHome page
D. Tattersall, C. A. Scott, C. Gray, D. Zicha, and D. P. Kelsell
EKV mutant connexin 31 associated cell death is mediated by ER stress
Hum. Mol. Genet., December 15, 2009; 18(24): 4734 - 4745.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
16/2/165    most recent
ddl452v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Unsworth, H. C.
Right arrow Articles by Kelsell, D. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Unsworth, H. C.
Right arrow Articles by Kelsell, D. P.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?