In silico and functional studies of the regulation of the glucocerebrosidase gene

Abstract

In Gaucher disease (GD), the inherited deficiency of glucocerebrosidase results in the accumulation of glucocerebroside within lysosomes. Although almost 300 mutations in the glucocerebrosidase gene (GBA) have been identified, the ability to predict phenotype from genotype is quite limited. In this study, we sought to examine potential GBA transcriptional regulatory elements for variants that contribute to phenotypic diversity. Specifically, we generated the genomic sequence for the orthologous genomic region ( approximately 39.4kb) encompassing GBA in eight non-human mammals. Computational comparisons of the resulting sequences, using human sequence as the reference, allowed the identification of multi-species conserved sequences (MCSs). Further analyses predicted the presence of two putative clusters of transcriptional regulatory elements upstream and downstream of GBA, containing five and three transcription factor-binding sites (TFBSs), respectively. A firefly luciferase (Fluc) reporter construct containing sequence flanking the GBA gene was used to test the functional consequences of altering these conserved sequences. The predicted TFBSs were individually altered by targeted mutagenesis, resulting in enhanced Fluc expression for one site and decreased expression for seven others sites. Gel-shift assays confirmed the loss of nuclear-protein binding for several of the mutated constructs. These identified conserved non-coding sequences flanking GBA could play a role in the transcriptional regulation of the gene contributing to the complexity underlying the phenotypic diversity seen in GD.

Figure 1

The GBA locus. Local topology of the genomic sequence surrounding the GBA gene (in humans). GBA and GBAP are both on the reverse strand, antiparallel to both MTX1 and MTX1P. Following the early primate duplication, MTX1P in primates occupies the spot previously held by MTX1 (earlier in evolution). Coordinates reflect NCBI Build 36.1 (hg18).

Figure 2

Predicted transcription factor-binding sites (TFBSs) identified by Genomatix Frameworker. The “Upstream Module” (left) and the “Downstream Module” (right): a cartoon shows the location and orientation of different predicted TFBSs, and a schematic representation of the genomic region including exons 1 and 2 of GBA (upstream) and exons 3 and 4 of MTX1/MTX1P (downstream) illustrate the relative location and composition of analogous clusters identified in various species and human pseudogene.

Figure 3

Multi-sequence alignment of the GBA promoter region (A) and intron 3 of the MTX1P (B). The specific TFBSs and the nucleotides altered are annotated.

Figure 4

The design of the luciferase construct (pGL4.10-hGBAloci) is shown. Using a PCR-based insertion strategy, 8.1 kb of the GBA locus encompassing both upstream and downstream sequence was subcloned stepwise into pGL4.10.

Figure 5

Luciferase expression in transfected COS-7 cells. Luciferase expression, as represented as firefly luciferase (Fluc) normalized to Renilla luciferase (Rluc), is shown for eight transfected reporter constructs. Five upstream and three downstream mutated TFBSs are shown. Inserts show a schematic representation of the upstream and downstream modules, showing the strand orientation of the sites.

Figure 6

Gel-shift assays were performed to validate the loss of binding to the predicted sequence alterations in the TFBSs. For the V$MOKF (A) and V$CMYB (B) sites, protein binding to the wild-type sequence was confirmed, while a loss of binding was observed with the mutant sequence. Probe concentrations (126 fmol) are the same in all lanes; unless stated otherwise, 10 μg protein extract was used.