Gap junctions are present between most animal cell-types. They allow different molecules and ions, mostly small intracellular signalling molecules (e.g. IP3), to pass freely between cells. Thus, gap junctions connect the cytoplasm of cells and in cardiac muscle, allow the transmission of electrical impulse that trigger the precisely tuned muscle contraction. One gap junction channel is composed of two connexons (or hemichannels) which connect across the intercellular space. Each connexon contains 6 molecules of connexins (Cx). Gap junctions channels formed from two identical connexons are called homotypic, while those with differing connexon are heterotypic. In turn, connexon of uniform Cx composition are called homomeric, while those with differing connexins are heteromeric. Several gap junctions channels (hundreds) assemble into a macromolecular complex called a junctional plaque. As in most tissues and organs, multiple Cx types are expressed in the heart (Cx43, Cx40 and Cx45) and are found in distinctive combinations and relative quantities in different, functionally-specialised subsets of cardiac myocyte.

Mutations in genes that encode Cx have only rarely been identified as being a cause of human cardiac disease, but remodelling of Cx expression and gap junction organisation are well documented in acquired adult heart disease, notably ischemic heart disease and heart failure (Figure 1). Dr. Dupont has focused his research on the description of the channel composition and how it influences the function of the gap junctions. The first aim has been to establish a cell model (RLE cells) with regulated expression of a second Cx (i.e. Cx40 and Cx45) in addition to the endogenous Cx43 to examine the structure and function of the resulting gap junctions. Transfectants display homogenous distribution of connexins. However, when Cx40 and Cx43 are co-expressed, heterogeneous distribution of putative heteromeric connexons are formed as is observed in vivo. Mixed Cx40/Cx43 connexons and pure Cx40 connexons are not compatible with pure Cx43 connexons (Figure 2.) and cells expressing Cx40 and Cx43 communicate through pure Cx43 channels (homomers/homotypes). Cx45 is less selective, thus, when Cx45 and Cx43 are co-expressed, heteromeric Cx45/Cx43 and homomeric Cx45 connexons are compatible with homomeric Cx43 connexons. The stoichiometries of connexins expression appear to impose a selectivity of compatibility that likely influences the functionality. Lucifer yellow dye transfer is inhibited by around 50% in both cell lines. Since only Cx43 homomers/homotypes are permeable to this dye, the predicted 96% heteromer formation (random association, binomial distribution) is not compatible with only 50% reduction in dye transfer. Co-immunoprecipitation experiments shows that mostly homomers (connexon made of a single connexin) are formed in both cell lines. In the Cx40/Cx43 expressing cells, this result in a 50% loss of Cx43 channels because heterotypes are undetectable by immunofluorescence (Figure 2) and the number of channels remains constant for a stoichiometry of ~1:1 resulting in ~50% pure Cx40 channels (impermeable to the dye) and ~50% pure Cx43 channels (permeable to the dye). In the Cx43/Cx45 expressing cells, the number of channels doubles (100% increase) for a stoichiometry of ~1:1 but since homotypes (Cx43:~50%, Cx45:~50%) and heterotypes (cx43/Cx45:~100%) can and do form, this results in 50% loss of pure Cx43 channels and therefore 50% decrease in dye transfer.

Since, RLE cells do not produce action potential, Dr. Dupont has cloned out stable cell lines from HL-1 atrial myocytes which are more similar to primary myocytes. He has characterised 5 clones for their pacemaker activity, caffeine-releasable Ca2+ storage, contractability and myofilament expression. All HL-1 clones co-express Cx40, Cx43 and Cx45 which colocalise but with a heterogeneous distribution (Figure 3). Less than 20% of each Cx is in the form of heteromers. Different profiles are observed, for instance, clone#6, has 100% contractility, expresses high levels of Cxs and has fast propagation; whereas clone#2 has little if any contractility, expresses poor levels of Cxs and has slow conduction. Clone#6 contains 4X more junction Cxs than clone#2, but conduction velocity is 6-8X faster than clone#2 (Figure 4). In conclusion, other factors are involved to explain this difference.

The future studies of Dr. Dupont will focus in the correlation of Cx co-expression (using siRNA and inducible plasmids) and conduction velocities in immortal HL-1 clones, in order to explain how gap junctions form the cell-to-cell pathways for propagation of the precisely orchestrated patterns of current flow that govern the regular rhythm of the healthy heart.