Dear Sir,
Thank you for the opportunity to address the reader’s very important question regarding the roles of various connexin isoforms in the complex function of the human AV node.1 These isoforms can serve as a rate-dependent AV conduction axis during normal sinus rhythm or a lifesaving filter of high-frequency excitation produced by atrial tachyarrhythmias or as a backup junctional pacemaker.
Cx43, Cx40 and Cx45 are the major connexin isoforms expressed in the human AV junction. Cx43 is the most well characterised and anatomically-mapped isoform, especially in humans.2 Several important studies3–5 have reported the mRNA and protein expression of Cx40 and Cx45 in different regions of the AV junction, which appear to play an important role in the complex function of the human AV node. Cx31.9 remains controversial: its mRNA was reported to be present in the human AV node but its protein expression was undetectable in humans, unlike evidence of its ortholog Cx302 in the mouse sinoatrial and AV nodes.6 This is a good example of how mRNA expression data do not necessarily correlate with protein expression and function.
Though the mRNA and protein expression data are robust and clearly address important questions, to investigate this particular issue it would be necessary to immunolable serial sections of the human AV junction, as was done with Cx43 in Hucker et al.2 This method is more advantageous as it allows us to directly compare histological data with protein expression in the region of interest. The availability of histological data confirms the location and boundaries of the various regions of the AV junction (which have morphological differences) that are otherwise not easy to define.
The serial sectioning protocol is technically challenging and strongly dependent on the quality of antibodies. For more than a decade we have tried to anatomically map Cx45 and Cx40 protein expression using immunofluorescence microscopy. Unfortunately only Cx43 mapping was possible due to the low quality of antibodies available on the market and those produced by our collaborators and us. Despite numerous attempts to immunolable serial sections of the human AV and sinoatrial nodes, we have been unable to obtain the consistent high-quality signals from either Cx40 or Cx45 antibodies that are necessary for 3D reconstruction of their distribution in the AV junction.
Over the past few years, we have tested and validated several new antibodies that might work well for this purpose, especially with the advent of the novel CLARITY method, which allows protein mapping in 3D anatomical structures without the need for tissue sectioning.7 This topic is an on-going collaborative project in our laboratory and we hope to have data that directly answer this question in the near future.
Igor R Efimov, George Washington University, Washington, DC, USA
Sharon George, George Washington University, Washington, DC, USA