Correction for Tomschik et al., Fast, long-range, reversible conformational fluctuations in nucleosomes revealed by single-pair fluorescence resonance energy transfer

BIOPHYSICS. For the article “Fast, long-range, reversible conformational fluctuations in nucleosomes revealed by single-pair fluorescence resonance energy transfer,” by Miroslav Tomschik, Haocheng Zheng, Ken van Holde, Jordanka Zlatanova, and Sanford H. Leuba, which appeared in issue 9, March 1, 2005, of Proc Natl Acad Sci USA (102:3278–3283; first published February 22, 2005; 10.1073/pnas.0500189102), the authors wish to note the following: “In this article, we used single-pair fluorescence resonance energy transfer (spFRET) experiments to investigate conformational fluctuations in nucleosomal DNA. Based on the reported results, we concluded that there were spontaneous, long-range, reversible fluctuations in nucleosomal DNA, leading to opening of the nucleosome for protein binding. A potential artifact in such studies is the photoblinking of the acceptor fluorophore (1), which was not conclusively ruled out in our 2005 article. Subsequent investigations make it clear that the photoblinking artifact impacted the values reported in Fig. 4D and in the last three columns of Table 1. We (M.T., K.v.H., and J.Z.) have now used formaldehyde-crosslinked nucleosomes and imaging in the presence of Trolox—an efficient triplet-state quencher that helps suppress photoblinking (2)—to reevaluate our previous conclusions. Analysis of the data indicates that most of the events previously characterized as nucleosome ‘opening’ must have corresponded to photoblinking. There is, nevertheless, evidence for the existence of infrequent, rapid opening events. We would like to point out that the very low frequency of opening events deduced from the present measurements is in agreement with the earlier biochemical measurements from Widom's laboratory (3). A manuscript reporting on these data has now been published (4). We emphasize that the data in the 2005 article are wholly reproducible; the problem is one of data interpretation caused by the existence of photoblinking. We note that the original data published in 2005 did contain a clear indication of long-range opening events. Thus, for example, the time trajectories in Fig. 4A contain events in which the fluorescence intensity of the acceptor dye in low-FRET states is significantly higher than that in the photobleached state, indicating true molecular fluctuations. However, these events are now confirmed to be rarer than put forward in the original report.”