Significance of myocytes with positive DNA in situ nick end-labeling (TUNEL) in hearts with dilated cardiomyopathy.

Situ Nick End-Labeling (TUNEL) in Hearts With Dilated Cardiomyopathy To the Editor: Recently, Kanoh et al1 seriously questioned the reliability of the DNA in situ nick end-labeling (TUNEL) assay as a method of detecting cardiomyocyte apoptosis (CA) in dilated cardiomyopathy. Our experience with both failing2 and infarcted3 human hearts is remarkably different. We agree that TUNEL positivity is not 100% specific to what is morphologically defined as apoptosis. However, if one accepts double-strand DNA breaks as one of the hallmarks of apoptotic cell death (which Kanoh et al must even question), our experience is that the detection of this preferred substrate by the TUNEL assay results in reasonable (0.1% to 1.0%) and reproducible (r50.88) estimates of CA frequency in relevant samples. DNA ladders are demonstrable in the TUNEL-positive areas when the amount of positive cells exceeds '0.04%. We think that the key reason for the findings by Kanoh et al is their lack of appropriate standardization of the TUNEL assay. As we have pointed out previously,2–5 one must go beyond the manufacturer’s instructions to avoid erroneously false-positive and false-negative results. This can be done by using adjacent tissue sections treated with DNase I as a positive control of apoptosis and by interrupting the staining reaction on the appearance of positive signal in these sections. This procedure confirms the optimal sensitivity of the assay and normalizes it for differences in tissue permeability.2– 4 Using this approach, TUNEL positivity is never zero; rather, it is in the range of 0.003% to 0.01% in normal myocardium. Furthermore, the very high numbers of labeled cells in the positive samples of Kanoh et al (7.9% and 5.8% using the electron microscopic immunogold assay and light microscopic TUNEL staining, respectively) point to problems in standardization (or the presence of artefacts) when methods based on DNA fragmentation are used. We and others have repeatedly observed features of apoptotic morphology, such as condensed instead of hypertrophied nuclei, in TUNEL-positive cardiomyocytes.2,4 Why Kanoh et al failed to find morphological evidence of CA could be due to the very small number of cells studied per tissue sample (10% of the '500 cells that would be required to find 1 truly apoptotic cardiomyocyte).2,5 Because only 15% (n56) of the biopsies were positive, how many of the 20 patients actually contributed to their positive data? The value of the TUNEL assay lies in its excellent signal-tonoise ratio and, hence, suitability for the quantification of very low amounts of positive cells. Although potentially more specific tests, such as the Taq polymerase assay, should be rigorously tested for the quantification of CA, we think that the TUNEL assay is currently the method of choice for this purpose, provided that it is allowed to perform at its best.