Glucocerebrosidase gene has an alternative upstream promoter, which has features and expression characteristic of housekeeping genes

Database searches have shown that a part of glucocerebrosidase (GBA) transcripts may originate at an alternative upstream promoter (P2) located 2.6 kb upstream of the known (P1) GBA promoter. The putative alternative transcripts contained one or two extra exons (exon -2 or exons -2, -1, respectively), but the first ATG codon and predicted amino-acid sequence are the same as in the transcript from P1. Luciferase assays confirmed promoter activity of both sites in HepG2 cells: the P1 construct exhibited the highest activity of luciferase (17.82±1.10 relative luciferase units), while the P2 construct reached 3.01±0.43 relative luciferase units. Serial 5' deletions of P2 led to changes in reporter activity, the most prominent decreases were observed in deletion constructs carrying bases -353 to -658, and -353 to -920 (numbered as in NM_001005750.1), respectively. This suggests that the P2 core promoter is contained within the region of -920bp to -1311bp. Three P2 transcription initiation sites were found by 5' RACE at positions 347, 380, and 413bp upstream of the +1 ATG. The expression stability of transcripts from P2, P1 was studied in 20 human tissues and was higher than that of GAPDH and ACTB, which are commonly used as reference housekeeping genes. The P2 contains an unmethylated CpG island, multiple Sp-1 consensus binding sites and, unlike P1, does not contain a TATA box, features all common to the majority of housekeeping gene promoters. We have examined DNA samples from a phenotypically diverse group of twenty Ashkenazi Jewish Gaucher patients homozygous for the common mild mutation N370S. Both P1 and P2, as well as exons -2 and -1, did not contain any sequence variations, with the exception of the known polymorphism rs10908459 found on one allele. The phenotypical differences in the patients were thus not explained by nucleotide variations in both promoters.