Discussion

Mut⁺ strains show metabolite methanol more efficient than Mut^s strains. P.

pastoris X33 is a wild strain belongs to Mut⁺ genotype, therefore the advantages of using X33 is mainly possessing a good growth rate when facing a methanol induction.

In this study, the Mut⁺ selection was conducted before Zoecin selections (for higher

copy number strains) in order to avoid choosing a higher copy number but with a Mut^s

genotype. During Mut⁺ selections, 35 colonies were chosen from 50 colonies for

growing well in MM plates and these strains were further screening in

Zeocin-containing plates. Our results showed all strain grew well on plates with lower

Zeocin concentrations (50 μg/mL and 100 μg/mL), but only few could grow on plates

with higher Zeocin concentrations (500 μg/mL and 1000 μg/mL). It suggested that

multiple recombination sites were happened after electroporation of vectors.

During Zeocin selections, 9 strains tolerant to 1000 μg/mL Zeocin were

discovered (Fig. 3). After ELISA quantification of the fusion allergen, 3 strains showed

higher productivity than 10 μg/mL after 24 h of methanol induction, and 5 strains

showed the productivity lower than 10 μg/mL. Among these strains, No. 16 was found

without the ability of fusion allergen expression. It suggests that strain 16 might be a

Mut^s strain.

43

Our ELISA and Western blotting results indicated that the fusion allergen was

secreted to the medium and the α-helix linker did work. Via a high cell density

cultured in a fermentor (Fig. 9), we observed that before adaption to methanol, for P.

pastoris methanol was used for induction of fusion allergen production and P. pastoris

grew slowly. Once P. pastoris adapted to methanol, methanol would be used as the

energy source to expand the biomass and secreted the Pichia-derived protease to

decompose the expressed fusion allergen as the nutrients. When the biomass came into

a saturation stage, the methanol was used for fusion allergen production again. In order

to solve the disadvantage of co-expression of protease, P. pastoris strain such as

SMD1168 and KM71 might be good choices. These strains are protease-deficient and

therefore the recombinant proteins could be preserved during the methanol induction.

In addition, to elongate the stage of glycerol-fed to achieve a higher biomass yield

before methanol feeding is an alternative way.

The reports regarding productivity of Der p 1 or Der p 2 expressed in different

hosts were summarized in Table 4. Reports show that crucial commitment of proteolytic

activity of a pichia-derived rDer p1 to sensitization toward IgE and IgG responses

(Takai et al., 2002; van Oort et al., 2004; Kikuchi et al., 2006). Der p 1 was first

expressed in E. coli in 1988 however produced in the form of inclusion body that rDer p

44

1 was hardly to be determined. During 2000-2001, rDer p 1 was reported to be

successfully expressed in Drosophila cells (20 μg/mL) and mammalian cells (34 μg/mL)

with a low productivity. Glycan test showed that N-glycosylation was not happened in

natural Der p 1, but were observed in Drosophila cells and mammalian cells (Jacquet et

al., 2000; Massaer et al., 2001). Natural Der p 1 showed a sharp band on SDS-PAGE

instead of a smear bands (Jacquet et al., 2002). In 2002, an expression of rDer p 1 in P.

pastoris was reported and rDer p 1 in cultured medium was 70 μg/mL. However, the

expressed rDer p 1 was found to be hyperglycosylated. Compared with the natural Der p

1, hyperglycosylation may lead rDer p 1 to be difficultly analyzed in vivo for losing its

histamine release activity, proteolytic activity and the inhibition of IgE binding ability.

After treated with N-Glycosidase, rDer p 1 recovered the similar activities to natural

Der p 1 (Jacquet et al., 2002). N-Glycosidase therefore can be a way to remove

glycosylation on the fusion allergen when we produce the rDer p 1 with a glycosylation.

And another strategy was to take a site-direct mutation of proDer p 1 (N52Q) which

optimized the rDer p 1 into a more similar character as the natural Der p 1 (van Oort et

al., 2004).

In addition to the effect of glycosylation on its performance of protease activity,

existence of the prosequence of natural Der p 1 is also thought to be an important issue.

45

Reports shows that rDer p 1 expressed with proDer p 1 domain design achieved a

similar activity to natural Der p 1(Jacquet et al., 2000; Massaer et al., 2001). We realize

that protease activity of Der p 1 was directly associated with the prodomain that would

function as an intramolecular chaperon to help Der p 1 correctly folded into an active

form, and it suggests that prosequence of Der p1 may be necessary for its immuno-

efficacy in clinic trial. After the construction of full-length Der p 1 fused with Der p 2,

this fusion allergen may have ability to facilitate itself to pass through epithelium cells

and enhance the immune response. Besides the above reports, rDer p 1 had also already

expressed in the rice as an edible vaccine and achieved a content of 50μg/grains.

Der p 2 was also initially expressed in E. coli as an inclusion body. Expressed in

eukaryotic system of S. cerevisiae (7 μg/mL) however was mostly to be degraded.

Expression of rDer p 2 in tobacco showed a good activity in oral delivery in murine

model of asthma (Ho, 2002) however with a low productivity. Co-expression of Der p 1

and 2 and in the form of a fusion allergen by P. pastoris is now first reported in this

study. The productivity of fusion allergen achieved 203 μg/mL in 60 hours of the

culture. This is also the highest yield among the related articles which had been

reported.

In addition to the previous issues, switching the linker sequence into other linkers

46

with different characters is also an interesting work. Overall, to establish an economic,

efficient, and to get an active form of fusion allergen, the choice of an adequate link is

absolutely important.