Previously, we isolated several cDNA clones of the LIS1 gene implicated in Miller-Dieker syndrome. Analysis of the 5' end of one of the clones (8-1), which was originally thought to represent the 5' end of LIS1, indicates a striking similarity to mouse 14-3-3 epsilon. We have isolated a full-length cDNA of human 14-3-3 epsilon, for which sequence analysis reveals a strong nucleotide conservation with mouse 14-3-3 epsilon in both translated and untranslated regions (UTRs). Additionally, the predicted peptides of human, sheep, rat, and mouse 14-3-3 epsilon are identical. Using a 205-bp fragment common to LIS1 (8-1) and 14-3-3 epsilon as probe on adult and fetal multiple-tissue Northern blots, a -2-kb transcript is detected, identical to the pattern observed with a full-length 14-3-3 epsilon cDNA probe. LIS1-specific transcripts of approximately 7.5 and approximately 5 kb are not detected by the 0.2-kb probe, indicating that the similarity between the 5' sequence of LIS1 (8-1) and the 3' UTR of 14-3-3 epsilon is not the result of shared homology between the two genes. Instead, clone 8-1 is a chimera of 14-3-3 epsilon and LIS1 partial cDNAs, and therefore its 5' sequence does not represent the LIS1 5' end. Interestingly, we have mapped the 14-3-3 epsilon gene to the same chromosomal sub-band as LIS1 (17p13.3). However, 14-3-3 epsilon lies telomeric to LIS1 and outside the Miller-Dieker syndrome chromosome region but in a region frequently deleted in several types of cancer, and is a reasonable candidate tumor suppressor gene.