To explore the scope and significance of alternate promoter usage and its putative inter-relationship to alternative splicing, we searched expression sequence tags for the 5 region of acetylcholinesterase (ACHE) genes. Three and five novel first exons were identified in human and mouse ACHE genes, respectively. Reverse transcription-PCR and in situ hybridization validated most of the predicted transcripts, and sequence analyses of the corresponding genomic DNA regions suggest evolutionarily conserved promoters for each of the novel exons identified. Distinct tissue specificity and stress-related expression patterns of these exons predict combinatorial complexity with known 3 alternative AChE mRNA transcripts. Unexpectedly one of the 5 exons encodes an extended N terminus in-frame with the known AChE sequence, extending the increased complexity to the protein level. The resultant membrane variant(s), designated N-AChE, is developmentally regulated in human brain neurons and blood mononuclear cells. Alternative promoter usage combined with alternative splicing may thus lead to stress-dependent combinatorial complexity of AChE mRNA transcripts and their protein products.Alternative splicing and alternate promoter usage both expand the complexity of gene products. While the massive contribution of alternative splicing to such expansion is widely recognized (1), less is known about the scope and significance of alternate promoter usage. Moreover the directionality of transcription processes raises the yet unresolved possibility that these two phenomena are inter-related, namely that the choice of the first exon determines downstream splice choices (2). The recent accumulation of genomic and gene expression data bases together with the development of sophisticated bioinformatic tools makes these questions amenable for experimental analysis as multiple gene products display both alternate promoter usage and alternative splicing variations. By alignment of expression sequence tags (ESTs) 1 against genomic sequences, for example, it is possible to explore the different alternatively spliced products of a single gene (3, 4). However, EST data bases are biased toward the 3Ј end of mRNAs and occasionally contain genomic contaminations that may cause misinterpretation of the genomic information (5). To correctly evaluate the inter-relationship between alternate promoter usage and alternative splicing, it is therefore necessary to characterize the identified variants using traditional molecular biology tools at the RNA level and, if applicable, at the protein level as well.The acetylcholine-hydrolyzing enzyme acetylcholinesterase (AChE) provides an adequate example for such a study. AChE pre-mRNA is subject to stimulus-induced 3Ј alternative splicing (6), and previous evidence has suggested that it is also subject to alternate promoter usage (7,8). We analyzed the genomic regions flanking the human and mouse ACHE genes and found an unexpected, evolutionarily conserved diversity of alternate exons at their 5Ј end. The newly id...