Nature Genet. 14, 415– 420 (1996).

Due to a cloning error, the sequence reported for ING1 was incorrect. The error appears to have been a result of a compression introducing a frameshift and of the ING1 gene encoding several differentially spliced isoforms that contain a common 3′ exon, one of which is of a size very similar to that reported in the publication above. The original (ING1a′, previously called ING1) and the corrected (ING1a) sequences of the isoform of ING1 that was first reported are shown at right ( a). The sequence that we reported was frameshifted and truncated at the 5′ end, and we now know that it encodes a protein with a predicted mass of 46,751 rather than the predicted 33,253 daltons. A truncated ING1a message also encodes an expressed protein of 23,656 daltons that results from initiation at the first internal ATG in the conserved exon. An alternatively spliced isoform, ING1b, shares a common 3′ exon with ING1a and encodes a protein with a predicted mass of 31,843 daltons. We now refer to the proteins encoded by this gene as p47ING1a, p33ING1b and p24ING1c due to their relative electrophoretic mobilities. These polypeptides are found in the majority of primary and established tissue culture cells examined, including those of fibroblast, epithelial and glial origin. A schematic diagram of the proteins encoded by the two major transcripts of ING1 are shown (b). All three proteins contain a region with a high degree of homology to PHD fingers that are implicated in transcriptional regulation.