The quest for an atrium-specific biomarker.

Dear Sir, With great interest we read the article by Maass et al. [1] who started a ‘quest for an atrium-specific biomarker’ using oligodeoxyribonucleotide (ODN) microarrays to identify differences in gene expression profile between atrial and ventricular myocardium. One of the ‘atrium-enriched genes’ is Nppa that has a ‘fold A/V microarray’ of 3.33 [1], whereas in hearts of healthy human adults, the concentration of Nppa transcripts, which encode preproANP (a precursor of atrial natriuretic peptide or ANP), is >100-fold higher in atrial than in ventricular myocardium [2, 3]. The explanation for this apparent discrepancy is that gene expression profiling was done with RNA extracted from neonatal rat atrial and ventricular cardiomyocytes. This particular choice of tissue is highly peculiar, assuming that the biomarker, once found, will be used in material originating from adult human patients. Foetal (and neonatal) rat (and human) atrial and ventricular myocardium possess roughly equal proANP concentrations [2, 3]. In adults, proANP is no longer found in healthy rat and human ventricular myocardium [2, 3] but reappears under pathological conditions such as heart failure and hypertension [3, 4]. Furthermore, the heart of newborns changes from a mainly glucose-metabolising tissue to a primarily fatty acid-oxidising tissue, thereby upregulating the expression of genes involved in fatty acid transport and β-oxidation [5]. Another cautionary note concerns the tendency to equate transcript/RNA levels as determined by ODN microarray analysis with protein levels. This may not be justified because: (1) Many genetic loci specify all kinds of non-protein-coding transcripts besides the classical polyadenylated protein-coding mRNAs that might contribute to the signals observed on ‘DNA chips’ [6]; (2) Depending on the stringency of the hybridisation and washing conditions, the same DNA probe may be bound by related nucleic acid sequences with different coding capacity; and (3) Cells possess multiple mechanisms to control gene expression at the post-transcriptional level [7], many of which affect the ratio between a specific mRNA and the protein(s) it encodes. In conclusion, to find heart chamber type-specific proteins that will potentially allow the identification of atrium- or ventricle-specific biomarkers in human adults, the use of adult human atrial and ventricular RNA for gene expression profiling is recommended, provided any novel candidate is tested at the protein level, e.g. by Western blot analysis.