Dear colleagues,

Cathepsin A (Cath A) and b-galactosidase (b-gal) form the high molecular weight protein complex (Mw 680 kDa). This core complex can be extended to 1.28 MDa with sialidase (Sial) and other lysosomal proteins. The formation of complexes protects b-gal and activates Sial in the lysosome. On the other hand, mutations in the complexes cause of at least three genetic diseases: galactosialidosis (mutations in Cath A), sialidosis (mutations in Sial) and b-galactosialidosis (mutations in b-gal). However, it is not clear yet how the complexes function in the cell.

The goal of my project was to express human lysosomal b-Gal and Cath A in the yeast Pichia pastoris and reconstruct the complex between recombinant the proteins.

For better reviewing we could separate the project at four major parts: cloning in E. Coli (1), transfection of methylotrophic yeast with b-gal or Cath A (2), co-transfection of a positive clone expressing Cath A with b-gal (3) and reconstruction of complex between the expressed recombinant proteins (4). Moreover, each transformation we could separate at few smaller steps like selection of recombinant clones on the selective medium, detection of the gene of interest with PCR and selection of most efficient positive clone with Western blot analysis and the enzyme assay (Fig. 1).

Before the cloning in E. coli (step 1), we modified the cDNAs. We added the cleavage site of bovine factor 10 and polyHis-tag to the sequence of the genes. The modification of cDNA with cleavage site allowed us to cut the modified region after protein purification and polyHis-tag simplified the protein purification to one-step by affinity chromatography. Before the transfection, we also sequenced junction region of ready plasmids.

Then we transformed the yeast by electroporation with linerized plasmid. For transfection we used histidinol deficient yeast strain G115 and vectors pPIC-9 or pPIC9k that were commercially available from Invitrogen. The first vector contained a copy of histidinol oxydase gene. The second vector contained a copy of kanamycin resistance factor. Thus, after transfection with b-gal (step 2), we could select the recombinant clones on histidinol deficient medium, and then after transfection or co-transfection with Cath A (step 2 and 3, correspondently), we could perform the selection of clones with antibiotic. As a MOC control, we used a linerized construct without insert.

After the selection of clones, we tested them by PCR for the presence of insert. For b-gal (Fig. 2A) we amplified yeast genomic DNA with polyHis- and Gal-antisense-primers. The primers were complimentary to the sequences of polyHis-tag and b-gal, correspondently. The positive clones amplified 243 b.p. DNA fragment. This band was absent in the MOC control. However, we could see it in the positive control, which was the plasmid.

According to the results of enzyme assay, the positive clones (Fig. 2B) also secreted recombinant b-gal into the medium. Other wise, no enzyme activity was found in the MOC control as well as non-transformed yeast. Finally, the expressed recombinant protein reacted with anti b-gal antibody (Fig 2C). It also bound to the affinity column (Fig. 2D).

After the transformation of yeast with Cath A, we followed the same strategy. First, we confirmed the presence of the Cath A cDNA in the yeast genome by PCR amplification. Because of small size of Cath A cDNA (2.1 kB) we were able to use primers provided by the vendor. On the gel, we saw two bands (Fig. 3A). One of them (2.1 kB) corresponded to the gene of alcohol oxydase. This band was also present in the MOC-control. Another band (1.9 kB) had the same size that we expected for the Cath A cDNA. This band was absent in the MOC-control as well as some of recombinant clones. Western blot analysis (Fig 3C) demonstrated that yeast had not secreted Cath A into the cell culture medium. However, we detected the presence of Cath A in cell homogenates (Fig. 3C) Moreover, the immunoreactive protein was absent in the MOC-control. We also performed the enzyme assay for Cath A (Fig 3B). According to our results, the positive clones had sufficiently higher level of the enzyme activity. However, the presence of active protein in cell homogenates of control and negative clones directed us that yeast might contain an enzyme that could hydrolyse substrate of human Cath A. Finally, the affinity chromatography demonstrated that the proposed recombinant Cath A as early expressed recombinant b-gal also had polyHis-tag (Fig. 3D).

For co-transfection with b-gal we chose preselected clone CA7 that had maximal level of the Cath A activity. The PCR analysis (Fig 4A) showed that both genes (b-gal and Cath A) were present in the yeast genome. Moreover, the positive clones secreted b-gal in the cell culture medium (Fig 4B). However, the level of enzyme activity was 10 times less than we detected in b-gal monotransfected clones (Fig 2B). Moreover, Cath A activity was detected in cell homogenates of positive clones. Despite, no activity was found in the cell culture medium (Data not shown). The same result we obtained with Western blot (Fig 4C): yeast homogenates contained both enzymes and the yeast culture medium contained b-gal.

In our previous experiments with monotransfected clones, we found that secreted b-gal degraded in the cell culture medium (Data not shown). After co-transfection with Cath A, we found that something improved the enzyme stability. We speculated that Cath A might protect the b-gal in the yeast culture medium of co-transfected clones as it did in the lysosome. We proposed that expressed recombinant proteins could make a complex in the yeast. To check this hypothesis, we tried to precipitate b-gal with anti Cath A antibody. We prepared different dilutions of Cath A antibody and added them to the yeast homogenates (pH 5.2). The homogenates of monotransfected clones with b-gal served us as a negative control. After 4 h incubation at 37^oC, we precipitated the pellet with Pansorbin cells and measured the remained b-gal activity in the supernatants.

According to our data (Fig. 5), anti- Cath A antibody also precipitated b-gal from cell homogenates of co-transfected clones. However they were sufficiently less effective against b-gal in the negative control. Thus, we found that co-expressed proteins might interact and form a complex.

To determine the Mw of the complex we loaded yeast homogenates of mono and co-transfected clone onto a gel filtration column (Superose 6B, Amersham-Pharmacia Biotech). Our experiment showed, that presence of recombinant Cath A in the cell homogenates dramatically changed the profile of b-gal activity (Fig. 6). The peak fraction of b-gal expressed in monotransfected clone corresponded the Mw 260 kDa. Same Mw had the enzyme isolated from human placenta. However, we detected an extra peak on the chromatogram of b-gal expressed in co-transfected clone (Mw 460 kDa). Thus, the recombinant complex had different steochiometry than the binary complex of b-gal and Cath A purified from human placenta (680 kDa).

Conclusion:

1. Human recombinant b-gal (65 kDa) and Cath A (50 kDa) expressed and co-expressed in methylotrophic yeast Pichia pastoris.

2. Enzymes co-expressed in the yeasts formed showed immunochemical (b-gal and Cath A) and enzymatic (b-gal) identity to corresponding human proteins.

3. As native lysosomal enzymes, co-expressed recombinant proteins formed binary core-like complex that had different molecular weight (~460 kDa).