Naunyn-Schmied Arch Pharmacol (2009) 380:11–24 DOI 10.1007/s00210-009-0411-2
ORIGINAL ARTICLE
The anti-arrhythmic peptide AAP10 remodels Cx43 and Cx40 expression and function Jennifer A. Easton & Jorgen S. Petersen & Patricia E. M. Martin
Received: 6 August 2008 / Accepted: 4 March 2009 / Published online: 27 March 2009 # Springer-Verlag 2009
Abstract The anti-arrhythmic peptide AAP10 has previously been shown to acutely upregulate electrical cell-tocell coupling mediated via connexin 43 gap junctions. In the present work, we have further examined the connexin (Cx) specificity and mechanism of action of this peptide in HeLa cells expressing Cx43, Cx40 or Cx26. The ability of cells to transfer the small fluorescent dyes Alexa 488 (MW 570) or Alexa 594 (MW 759), as markers for metabolic coupling mediated via gap junctions, before and after exposure to AAP10 and/or the protein kinase C inhibitor chelerythrine for 5 h was determined by microinjection analysis. Immunofluorescence analysis assessed the effect of AAP10 on the spatial localisation of each Cx sub-type. Cell extracts were isolated for Western blot and reverse transcription polymerase chain reaction analysis at 0, 5, 10, 18 and 24 h following exposure to AAP10 and the relative Cx expression profiles determined. AAP10 enhanced the J. A. Easton : P. E. M. Martin (*) Department of Biomedical and Biological Sciences, School of Life Sciences, Glasgow Caledonian University, Glasgow G4 OBA Scotland, UK e-mail:
[email protected] J. S. Petersen Discovery Research, Zealand Pharma A/S, Glostrup, Denmark Present Address: J. A. Easton Department of Physiology, University of Liverpool, Liverpool L69 3BX, UK Present Address: J. S. Petersen Merck Serono International SA, 1202 Geneva, Switzerland
ability of Cx43 and, to a lesser extent, Cx40 to transfer Alexa 488. It also enhanced the ability of Cx43 to transfer Alexa 594 but not Cx40. Inhibition of protein kinase C blocked this enhanced response in both Cx sub-types. Western blot analysis determined that AAP10 induced Cx40 protein expression over periods of up to 24 h with an associated increase in the localisation of Cx40 at points of cell-to-cell contact following 24-h exposure. Cx43 expression was transiently induced following exposure to the peptide for 5–10 h, with an associated increase in Cx43 at points of cell-to-cell contact, returning to control levels by 18–24 h, via a post-translational mechanism independent of chelerythrine. A transient increase in Cx40 mRNA expression but not Cx43 mRNA expression was also observed. By contrast, AAP10 had no effect on the ability of Cx26 gap junctions to transfer the dyes or on the level of Cx26 expression. We propose that AAP10 is a versatile peptide that remodels metabolic coupling via Cx43 and to a lesser extent Cx40 gap junction channels via an initial protein-kinase-C-dependent pathway modifying local responses at the plasma membrane. This is followed by enhanced Cx43 or Cx40 protein expression. Keywords Connexin . Gap junction . Intercellular communication . Anti-arrhythmic peptide . AAP10, protein kinase C Abbreviations Cx connexin DMEM Dulbecco’s modified essential medium ECL enhanced chemiluminescence P/S penicillin/streptomycin PBS phosphate-buffered saline PKC protein kinase C
12
Introduction Gap junction intercellular communication channels provide a coordinated intercellular network that has far-reaching consequences on cellular physiology by permitting the passive exchange of small regulatory molecules and ions that are central to intercellular signalling events (Mese et al. 2007; Laird 2006; Evans et al. 2006). These dodecameric channels are composed of two connexons or hemichannels, one contributed from each cell, that are comprised of six connexin protein sub-units arranged around a central pore. Up to 21 connexin sub-types with different tissue expression profiles and unique functional properties have been identified to date (Mese et al. 2007; Laird 2006; Evans et al. 2006). These channels are dynamically regulated and changes in intercellular communication and remodelling of gap junctions are increasingly associated with pathophysiological conditions including heart conditions such as ischaemia (Beardslee et al. 2000; Turner et al. 2004), cardiac arrhythmia (Kanagaratnam et al. 2006; Gutstein et al. 2001) and atherosclerosis (Blackburn et al. 1995; Chadjichristos et al. 2006). The importance of connexinmediated communication in health and disease is further highlighted as changes in connexin expression are associated with cellular proliferation and differentiation events including tumourigenesis (Mesnil et al. 2005), woundhealing events in endothelial (Chanson et al. 2005) and epithelia tissue (Brandner et al. 2004; Richards et al. 2004), as well as an increasing number of genetically inherited connexin-associated channelopathies (Laird 2006; LaiCheong et al. 2007). To help unravel the role of connexins in this diverse range of conditions, a panel of tools and reagents that selectively inhibit or enhance connexinmediated communication are increasingly required and the underlying mechanistic properties and connexin specificity of these agents need to be dissected. Furthermore, by understanding the integral role that connexins play in such a variety of cellular responses, these agents become attractive therapeutic targets. A family of anti-arrhythmic peptides (AAPs) have been identified as agents that specifically enhance connexin (Cx)-mediated communication between cardiac myocytes (Muller et al. 1997; Kjolbye et al. 2007) and have also been shown to function in non-cardiac tissue including human osteoblasts (Jorgensen et al. 2005) and in HeLa cells transfected to express Cx43-GFP but not in those transfected to express Cx26-GFP (Clarke et al. 2006). The first well-characterised AAP molecule, AAP10 (H-Gly-Ala-Gly4Hyp-Pro-Tyr-NH2), is suggested to improve cardiac gap junctional coupling via a protein kinase Cα pathway (Weng et al. 2002; Dhein et al. 2001). This peptide is less stable than rotigaptide, due to its composition being of L amino acids rather than D amino acids (Kjolbye et al. 2003) yet
Naunyn-Schmied Arch Pharmacol (2009) 380:11–24
may offer a versatile tool to study the effect of upregulation of connexin-mediated communication in a variety of cellular events (Axelsen et al. 2007). In the present study, we have extended the analysis of connexin-specific effects of AAP10 from acute studies focussed on electrical coupling of cells to extended time periods and assessed the impact of the peptide both at the transcriptional and post-translational levels in HeLa cells selected to express Cx43, Cx40 or Cx26. The results show that AAP10 enhances metabolic coupling via Cx43 and to a lesser extent Cx40 gap junctions, but not Cx26, by a protein kinase C (PKC) pathway and that this enhanced metabolic coupling is associated with increased levels of Cx expression.
Materials and methods Cell culture HeLa wild-type (wt) cells were maintained in Dulbecco’s modified Eagle’s Media (DMEM) supplemented with 10% v/v foetal bovine serum, 100 µg/ml penicillin/streptomycin and 2 mM L-glutamine (cDMEM; Lonza, UK). HeLa cells stably expressing mouse Cx26, Cx40 and Cx43 (named HeLa26, HeLa40 and HeLa43, respectively) were maintained in cDMEM supplemented with 0.5 µg/ml puromycin-selective antibiotic (Sigma-Aldrich, UK; Mesnil et al. 1995). To examine the effect of AAP10 on Cxmediated communication, transfected HeLa cells were incubated with 50 nM AAP10 for 5 h prior to further analysis (Clarke et al. 2006). AAP10 was supplied by Zealand Pharma and was high-performance-liquid-chromatography-purified with a purity of 99%. The effect of PKC inhibition on Cx functionality and expression was examined following treatment of HeLa43 or HeLa40 cells with 5 μM chelerythrine or 300 nM staurosporine (Calbiochem) and 50 nM AAP10 for 5 h. To control for cell viability, cells were stained with trypan blue (Sigma) as required. Cx functionality assay Cells (1×106) were seeded onto 60-mm2 culture dishes, grown to 80% confluency and treated as required. Individual cells were microinjected with either Alexa 488 (charge 2, Mw 570 Da) or Alexa 594 dyes (charge −2, Mw 759 Da (Invitrogen, UK)), using an Eppendorf 5120 femtojet system linked to a Cairns monochromator with ~30 cells injected per plate for each experimental group. Cells were incubated at room temperature for 5 min to allow transfer to occur, washed in phosphate-buffered saline (PBS) and fixed in 3.75% formaldehyde. Cells were viewed under ×20 lens on a Zeiss Axiovert 200 microscope and appropriate filter
Naunyn-Schmied Arch Pharmacol (2009) 380:11–24
sets to record the extent of dye transfer from each successfully injected cell, as a marker for ‘metabolic coupling’. Initial observations indicated that for Cx43expressing cells dye transfer was extensively upregulated following AAP10 treatment with dye frequently spreading to more than 20–30 cells, where it became difficult to count the actual spread accurately. To distil the complexity of the dye transfer data and to enable accurate quantification, a statistically robust (and conservative) method is to group the data into categories (e.g. Harris et al. 2000; Chaytor et al. 2001; Clarke et al. 2006). In the present experiments, we grouped the data into three ‘arbitrary’ units of dye transfer (transfer to zero, one to nine or more than ten neighbouring cells) that were sufficient to identify changes. The number of cells each injected cell transferred dye to was recorded into the three categories. For individual plates, the percentage of cells transferring dye to zero, one to nine or more than ten cells was subsequently calculated, as the number of cells actually injected ranged from 25 to 35 cells depending on injection efficiency. Each plate was thus regarded as an independent ‘observation unit’. Experiments were repeated in triplicate at one setting and on three further occasions, giving n=9. The data were collated in prism and presented as percentage of injected cells per plate transferring dye to zero, one to nine or more than ten neighbouring cells ± SE as previously described (Harris et al. 2000; Clarke et al. 2006; Fig. 1). All data were subject to rigorous statistical analysis using Student’s t test or Dunnett’s multiple-comparison analysis as required. This enabled statistical differences between each relevant control and treatment group to be made where P
