We amplified the flag epitope GLB1 from Genscript GLB1 cDNA ORF clone (NM_000404.3) using PCR, then cloned it into pSB1C3 (BBa_K2807014) and a mammalian expression vector, pEGFP-C1. Subsequently, we carried out point mutations to the wild-type GLB1 gene to generate GLB1 TCT mutant (BBa_K2807016) and GLB1 CCG mutant (BBa_K2807015) which are point mutations that are reported to reduce GLB1 enzymatic activity by more than 95% in patients suffered from GM1 gangliosidosis (Hofer et al., 2010). To allow normalization of transfection and expression efficiency of GLB1 in cells, we retained EGFP in C1 vector. To eliminate the possibility of the fusion protein to affect enzymatic activity, we separated EGFP and GLB1 by a stop codon and frameshift linker sequence. We also included a Kozak sequence in front of GLB1 to allow effective ribosome binding to translate GLB1.

To verify the expression of GLB1 WT and mutants, we expressed both plasmids in HEK293T cells and then tested for the protein expression levels by Western blot. Cell lysates were separated on an SDS-PAGE gel, and transferred onto a nitrocellulose membrane for blotting. The plasmids with GLB1 had a Flag epitope tag as well as an EGFP tag preceding the GLB1 coding sequence. We probed the membrane with anti-EGFP, anti-Flag and anti-Actin antibodies. Figure 1 showed that GLB1 protein is expressed in both GLB1 WT and the GLB mutants at 84kDa, the expected size of GLB1 precursor, with the non-transfected lane as the negative control. In addition, a 27kDa band corresponding to EGFP was seen for both GLB1 WT and the GLB1 mutants, but not the non-transfected control, confirming our transfection efficiency. Therefore, we can conclude that our WT and mutant EGFP-GLB1 construct are able to express both EGFP and precursor GLB1 at the correct length. Moreover, we confirmed that the reported single nucleotide mutation in GLB1 does not change gene translation level or degradation rate. As such, any change in apparent enzymatic activity of the mutant GLB1 is likely due to a change in the catalytic efficiency of the enzyme but not the amount of enzyme.

Next, we wanted to investigate if the precursor form of GLB1 is enzymatically active by itself. We transfected EGFP-GLB1 plasmid into HEK293T cells and carried out β-galactosidase functional assay using cell lysate. To optimise the method of cell lysis for obtaining the β-galactosidase protein, we tried three methods, namely Promega Passive Lysis Buffer, 1% Triton and freeze-Thaw method using dry ice. Upon running a Western blot with the lysates from the varying methods, we could see that the Promega passive lysis buffer produced the least number of non-specific bands as well as the correct band size for GLB1 at about 84kDa as seen in Figure 2. As such, this lysis buffer was used for subsequent enzymatic assays.

We carried out two β-galactosidase assays, using galactoside analog o-nitrophenylβ- D-galactopyranoside (ONPG) and 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal) as the substrate respectively. Functional β-galactosidase converts the colourless galactoside analog o-nitrophenylβ-D-galactopyranoside (ONPG) into galactose and chromophore o-nitrophenol, producing a bright yellow product. The final product can then be quantified using a spectrophotometer at 420 nm to determine the amount of substrate converted into the yellow chromophore. On the other hand, GLB1 cleaves X-gal to produce bright blue 5-bromo-4 -chloro-3-hydroxyindole and galactose as products. The appearance of the blue colour indicates that the sample contains a functional β-galactosidase enzyme, while a lack of blue colouration will be expected for the GLB1 mutants.

As shown in Figure 3, we incubated ONPG with cell lysates from pGEMT expressing DH5α with 6hr or 24hr IPTG induction (Tubes A and B) as positive control, HEK293T transfected with empty vector (Tube C), HEK293 cells transfected with GLB1 WT (Tubes D) and GLB1 TCT and GLB1 CCG mutants (Tubes E and F). The absorbance reading at 420nm is shown in Table 1. Tubes A and B (Figure 3), our positive controls, had significant β-galactosidase activity. However, for Tubes C to F (Figure 3), no colour change was observed. Taking into consideration that GLB1 is a lysosomal enzyme with optimal pH at 4, we repeated the experiment at pH=4.5 and there was no observable colour change. Similar results were observed in the dot activity assay using X-gal, where both non-transfected cells, wild-type and mutant GLB1 produced very faint blue colouration (Figure 4C-F) as compared to the positive control (Figure 4A-B).

The above results confirmed that precursor β-galactosidase is not functional, as post-translational modification via proteolytic cleavage at the C-terminal of the GLB1 precursor protein in the acidic environment of the lysosome is required for maturation of beta-galactosidase (Kreutzer et al., 2008). Therefore, in the future, we will engineer EGFP-GLB1 plasmid to include the cleavage signal to produce 64kDa and 22kDa proteolytic fragment that makes up the mature enzyme.