The transcriptional enhancer of a chicken Ul small nuclear RNA gene has been shown to extend over approximately 50 base pairs of DNA sequence located 180 to 230 base pairs upstream of the Ul transcription initiation site. It is composed of multiple functional motifs, including a GC box, an octamer motif, and a novel SPH motif. The contributions of these three distinct sequence motifs to enhancer function were studied with an oocyte expression assay. Under noncompetitive conditions in oocytes, the SPH motif is capable of stimulating Ul RNA transcription in the absence of the other functional motifs, whereas the octamer motif by itself lacks this ability. However, to form a transcription complex that is stable to challenge by a second competing small nuclear RNA transcription unit, both the octamer and SPH motifs are required. The GC box, although required for full enhancer activity, is not essential for stable complex formation in oocytes. Site-directed mutagenesis was used to study the DNA sequence requirements of the SPH motif. Functional activity of the SPH motif is spread throughout a 24-base-pair region 3' of the octamer but is particularly dependent upon sequences near an SphI restriction site located at the center of the SPH motif. Using embryonic chicken tissue as a source material, we identified and partially purified a factor, termed SBF, that binds sequence specifically to the SPH motif of the Ul enhancer. The ability of this factor to recognize and bind to mutant enhancer DNA fragments in vitro correlates with the functional activity of the corresponding enhancer sequences in vivo.The small nuclear RNAs (snRNAs) of the U family are evolutionarily conserved and metabolically stable RNA molecules present in the nuclei of eucaryotic cells. Considerable evidence exists that the Ul, U2, U4, U5, and U6 snRNAs are involved in the splicing of mRNA precursors (37). With the exception of U6, the snRNAs are synthesized by RNA polymerase II (6) and have a distinctive N2,N2,N7-trimethylguanosine cap structure (34).The genes that code for the vertebrate snRNAs are usually present in multiple copies per genome and have several features that distinguish them from protein-coding genes. The promoter regions of snRNA genes lack transcription signals normally found upstream of genes transcribed by RNA polymerase II (e.g., they lack TATA and CCAAT boxes). Nevertheless, they contain two distinct and evolutionarily conserved cis-acting regulatory regions within their 5'-flanking DNA sequences that are important for snRNA gene expression: a proximal region centered near position -55 relative to the transcription initiation site and a distal region located near position -200 (reviewed in references 6 and 32). Moreover, the formation of the 3' ends of the snRNAs is dependent on the initiation of transcription from an snRNA gene promoter (4,13,14,28). Because of the unique properties of snRNA transcription complexes, the promoter elements of snRNA genes are not functionally interchangeable with comparable elements of mRNA genes...