Introduction

Recently, the development of yeast surface display system heralds a new era of combinatorial bioengineering by virtue of characteristics of well developed cultural technique, post-translational modification as well as the industrial application in the field of biotechnology (Kondo and Ueda 2004). In yeast surface display technology, proteins or enzymes are immobilized on the cell surface by genetic operation and could be used for a wide variety of biotechnological applications (Tanaka et al. 2010), such as the delivery of live vaccines (Georgiou et al. 1997; Liljeqvist et al. 1997), immobilized biocatalysts used bioconversions (Li et al. 2008; Lee et al. 2004; Tanino et al. 2009), bioremediation of heavy metals (Samuelson et al. 2000; Kuroda and Ueda 2006), and as biosensors(Li et al. 2008). As a common host, Pichia pastoris proved its competitive ability for appropriate N-glycosylation for heterologous protein, inducible expression regulated tightly by alcohol oxidase 1 promoter (pAOX1) as well as little own protein secretion in the supernatant when compared to other common yeasts such as Saccharomyces cerevisiae (Wang et al. 2010; Tanino et al. 2006).

To some extent, the efficiency of protein display or enzyme anchoring on the yeast surface which has not been accurately evaluated hinders the practical application of cell surface display systems. Shibasaki et al. (2001) calculated the number of enhanced green fluorescent protein (EGFP) displayed as α-agglutinin fused protein on the S. cerevisiae cell surface by measuring the fluorescence intensity of the cell pellet suspended in PBS buffer (pH 7.4). But few reports have described the quantification of foreign proteins/enzyme displayed on the P. pastoris cell surface for any anchor. In addition, due to differences between the two hosts, the method suggested by Shibasaki et al. (2001) was insufficiently accurate for quantifying surface displaying proteins.

FLO1p from S. cerevisiae is a lectin-like anchor protein that plays an important role in flocculation. FLO1p contains 1,200 amino acids and is composed of the following four domains: (1) secretion signal, (2) flocculation functional domain, (3) glycosylphosphatidylinositol (GPI)-anchor attachment signal, and (4) membrane-anchoring domain (FS anchor) (Kondo and Ueda 2004).The FS anchor of approx. 874 amino acids here contains the N-terminal flocculation functional domain (FFD) which can recognize and adhere noncovalently to cell-wall components such as mannan carbohydrates (Han et al. 2009; Kondo and Ueda 2004). Moreover, the FS anchor is especially suitable for displaying the enzymes which has their active sites near the C terminus such as Rhizopus oryzae lipase (ROL), Rhizomucor miehei lipase, RML (Han et al. 2010), so that it is becoming a popular anchor for yeast surface display (Matsumoto et al. 2002; Han et al. 2010). Tanino et al. (2006) displayed the Rhizopus oryzae lipase including a pro sequence (ProROL) on P. pastoris with FS as anchor, and demonstrated that the FS anchor system in P. pastoris is effective in the same manner as in S. cerevisiae or ever more competitive when taking account of the factor that P. pastoris can be cultivated at a higher cell density.

In this study, we used the FS anchor to display a fusion protein of EGFP and Candida antarctica lipase B (CALB) on the surface of P. pastoris. CALB catalyzes many esterification and trans esterification reactions (Jaeger and Eggert 2002). Here we examined the relationship between the fluorescence intensity of leaked fusion protein and its protein concentration in the culture medium of recombinant EGFP-CALB/CALB-EGFP displaying P. pastoris. Using the relationship between fluorescence intensity and protein concentration, we quantified the number of fusion displayed on the P. pastoris cell surface. This method can be extend to other CALB yeast surface displaying systems involving other anchor proteins, thus facilitating more efficient CALB displaying systems and enabling accurate evaluation of displaying efficiency of single host cell.

Materials and methods

Strains and media

Escherichia coli strain Top 10 was used as a host for recombinant DNA manipulation. P. pastoris strain GS115 was used to display the fuse protein of EGFP and CALB. E. coli cells were grown in LB medium (1 % tryptone, 0.5 % yeast extract, 0.5 % NaCl and 0.1 % glucose). P. pastoris was grown in BMGY medium(1 % yeast extract, 2 % peptone, 100 mM KH2PO4/K2HPO4 buffer (pH 6.0), 1.34 % yeast nitrogen base without amino acids (YNB), 4 × 10−5 % biotin, and 1 % v/v glycerol), and BMMY medium (1 % yeast extract, 2 % peptone, 100 mM potassium phosphate buffer (pH 6.0), 1.34 % yeast nitrogen base, 4 × 10−5 % biotin, and 0.5–2 % v/v methanol).

Construction of expression plasmids

The fluorescence positive control plasmid pKFS-EGFP was constructed as follows: a MluI–EcoRI fragment of the EGFP gene was prepared using PCR (primers 5′-ATGAACGCGTATGGTGAGCAAGGGCGAGG-3′ and 5′-CTGAGAATTCCTTGTACAGCTCGTCCATGC-3′) using pEGFP-N1 (GenBank Accession No.U55762) as a template, and inserted into the MluI–EcoRI site of the integration-type yeast cell-surface-displaying cassette vector pKFSR (stored in our laboratory; Han et al. 2009). The resulting plasmid was referred to as pKFS-EGFP. The pKFS-EGFP-CALB plasmid was constructed as follows: a MluI-ApaI fragment of the EGFP gene was prepared using PCR (primers 5′-ATGAACGCGTATGGTGAGCAAGGGCGAGG-3′ and 5′- TAGTGGGCCCCTTGTACAGCTCGTCCATGC-3′) using pEGFP-N1 (GenBank Accession No. U55762) as a template, and an ApaI–EcoRI fragment of the CALB gene was prepared using PCR (primers,5′-ATGTGGGCCCCTACCTTCCGGTTCG-3′and 5′- GTATGAATTCGGGGGTGACGATGC-3′) using pKNS-CALB (stored in our laboratory; Su et al. 2010) as a template. The two genes were linked through ApaI and inserted into the MluI–EcoRI site of the integration-type yeast cell-surface-displaying cassette vector pKFSR. The resulting plasmid was referred to as pKFS-EGFP-CALB. A pKFS-CALB-EGFP plasmid was constructed as follows: an ApaI–EcoRI fragment of the EGFP gene was prepared using PCR (primers, 5′-ACTTGGGCCCATGGTGAGCAAGGGCGAG-3′ and 5′-CTGAGAATTCCTTGTACAGCTCGTCCATGC-3′) using pEGFP-N1 (GenBank Accession No.U55762) as a template, and an MluI-ApaI fragment of the CALB gene was prepared using PCR (primers 5′-ATTGACGCGTCTACCTTCCGGTTCG-3′ and 5′-ATTAGGGCCCGGGGGTGACGATGC-3′) using pKNS-CALB as a template. The two genes were linked through ApaI and inserted into the MluI–EcoRI site of the integration-type yeast cell-surface-displaying cassette vector pKFSR. The resulting plasmid was referred to as pKFS-CALB-EGFP. The plasmids pKFS-EGFP, pKFS-EGFP-CALB, and pKFS-CALB-EGFP were transformed into P. pastoris using the lithium chloride method.

Fluorometric, microscopy and flow cytometry (FLC) assays

A fluorescence spectrophotometer (Tecan Infinite M200, San Jose, CA, USA) was used to measure the fluorescence of samples in Tris–HCl buffer (50 mM, pH8.0) on a 96-well plate. A filter with excitation at 480 nm and a filter for emission at 520 nm were used to detect fluorescence intensity. Fluorescence microscopy was carried out using an L710 confocal laser scanning microscope (CLSM) (Shibasaki et al. 2001).The 488 nm line of an argon/krypton ion laser was used for excitation. To measure the fluorescent signal of EGFP, the beam splitter and filter were set to 500–530 nm. The flow cytometer used for the analysis was a Flow Cytometry Quanta SC equipped with an argon ion laser (488 nm wavelength) capable of GFP excitation (Bahl et al. 2004).

Analysis of lipase activity

Lipase activity was measured using 4-nitrophenyl caprylate (pNPC; Sigma) and following the release of p-nitrophenol, at 405 nm. Free lipases and cell surface-displayed lipases were dissolved in Tris–HCl buffer (50 mM, pH 8.0), mixed with substrate solution (0.0625 mM pNPC, 0.1 % Triton X-100 in lipase assay buffer), and reacted for 5 min before measuring the A405 using a kinetic microplate reader. After centrifugation at 10,000×g for 1 min, 200 μl of the supernatant was added to a 96-well plate and absorbance was measured. Product yields were determined based on a standard curve prepared using p-nitrophenol as a standard. One unit of activity was defined as the amount of enzyme that released l μmol pNP per min from pNPC under the assay conditions. Each sample was assayed in triplicate and the average value was determined (Han et al. 2009).

Western blot analysis and protein analysis

SDS-PAGE using an 8 % (v/v) gel was carried out. The protein extracts separated on the gel were electroblotted on a nitrocellulose (NC) membrane (PALL Corporation, Port Washington, NY, USA) and allowed to react with the anti-GFP HRP-DirecT (MBL International, Woburn, MA, USA).A Bradford protein assay kit was used to determine protein concentration (Tanino et al. 2006; MBL International; Afanassiev et al. 2000).

Results

Detection the EGFP fluorescence intensity on cell surface

Using CLSM and FLC, the fluorescence emitted from the transformants GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP cultivated in BMGY and BMMY liquid medium were investigated. As a control, cells harboring the plasmid pKFS-EGFP (GS115/pKFS-EGFP) were used. After cultivation, all three strains showed obvious fluorescence when evaluate dusing CLSM and FLC. The light spots of GS115/pKFS-EGFP were much denser and brighter than the other two strains, GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP detected using CLSM (Fig. 1—2a, 2b, 2c). Additionally, the intensity of the light spots from GS115/pKFS-EGFP-CALB was similar to that of GS115/pKFS-CALB-EGFP (Fig. 1—2c), and the distribution of the light spots from the two strains were identical. Upon further analysis using FLC, the fluorescence histograms show the same trend as observed using CLSM; the signal of GS115/pKFS-EGFP was much stronger that for the other two plasmids (approx. 10 times), and the signal of the other two strains were similar to each other (Fig. 1—3a, 3b, 3c).

Fig. 1
figure 1

Confocal laser scanning microscopy and flow cytometry assay of P. pastoris recombinants strains (cultivated for 120 h): GS115/KFS-EGFP (a), GS115/KFS-EGFP-CALB (b) and GS115/KFS-CALB-EGFP (c). In FLC results, the left histogram shows the negative control of the flow cytometry assay is GS115 and the right histogram represents the P. pastoris recombinants strains

Full size image

Detection of CALB activity

The CALB displayed on recombinant yeasts GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP were analyzed. As a negative control, GS115/PKFS-EGFP showed no CALB activity (date not shown) and the CALB activity displayed on the cell surface of GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP were similar (Table 1). The highest activity of the two strains obtained at 120 h induced by methanol was consistent (approx. 500 U/g). These results agree with those of Tanino et al. (2006) who showed that some protein leaked into the culture medium. Notably, leakage into the medium for both the GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP appeared to approach the same extent since the CALB activities in the supernatants were similar (Table 1, approx. 14 U/mL). Moreover, the OD600 of GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP revealed similar growth rates.

Table 1 Lipase activity of the two fused EGFP-CALB strains
Full size table

These results suggest that the position between EGFP and CALB in the fusion protein does not affect significantly the activity of CALB displayed on the cell surface or in the medium. A correlation between fluorescence intensity of the cells and lipase activity on the cell wall were expressed as Fig. 2. The fluorescence intensity and lipase activity went up at the same slope with the culture time increased. Besides, when the fluorescence intensity of the displaying yeast increased within the range of 4 × 107 to 108 RFU/g, the corresponding lipase activity displayed on the cell surface showed a coincident rise from 150 to 500 U/g dry cell wt.

Fig. 2
figure 2

Correlation between fluorescence intensity and lipase activity (dry cell) in the cell wall

Full size image

SDS-PAGE and Western blot analysis

Proteins of interest were extracted from the yeast cell wall and medium and analyzed using SDS-PAGE and Western blot (Fig. 3). The result in SDS-PAGE showed that only one band appeared in the medium and it is of the same molecular size as the main band of extracts from recombinant yeast cell wall. When performed with Western Blot, both these proteins of same molecular mass recognized anti-EGFP antibody as well. As mentioned in section of ‘Detection of CALB activity’, we also detected certain activity of CALB and certain fluorescence of EGFP in the medium and on the cell wall. Since no other protein existed in the medium, the protein in the medium must have the peptides or structures of both CALB and EGFP. Taking the other factors that FS has been fused with CALB and EGFP by genetic engineering and FS is non-covalently adherent to the cell wall as well as the leakage in the medium based on FS surface displaying, the protein appeared in medium could be referred as same as that immobilized on the cell wall. The molecular weights of all fusion proteins displayed on the cell surface or leaked into the medium were similar (approx. 220 kDa) regardless of what strains were used; GS115/pKFS-EGFP, GS115/pKFS-EGFP-CALB as well as GS115/pKFS-CALB-EGFP. This high molecular weight may have been due to glycosylation events in P. pastoris (Ardo et al. 2012). Proteins displayed by FS often have O-glycosylated or have high-mannose and/or hybrid structures linked to FS fused protein so the molecular mass of the fusion protein is larger than the calculated value (Goossens et al. 2011; Tanino et al. 2006).

Fig. 3
figure 3

a SDS-PAGE of the protein displayed on the cell surface and leaked in the medium; b Western blot results (M marker (TaKaRa, Shiga, Japan; D532A), 1 GS115, 2 GS115/KFS-EGFP, 3 GS115/KFS-EGFP-CALB, 4 GS115/KFS-CALB-EGFP, S protein from the supernatant)

Full size image

Fluorescence and quantification of protein in the medium

Since no other protein detected, this protein concentration in the supernatant was attributed to the fusion protein. The protein concentration of GS115/pKFS-EGFP-CALB always kept accordant to that of GS115/pKFS-CALB-EGFP during the regular culture process (Table 2). Additionally, fluorescence spectrophotometry indicated that the increased fluorescence intensity of the two strains were similar as well. In addition, the fluorescence intensity of the two strains increased proportionately and respectively with increased protein concentration.

Table 2 Fluorescence intensity and protein concentration in the medium of two fused EGFP-CALB strains
Full size table

Quantification of molecular number displayed on the cell surface

From the result indicated by Fig. 2, the relationship between the fluorescence intensity and protein concentration for corresponding strain could be described as the following linear formula. The linear formula of GS115/pKFS-EGFP-CALB is y = 50.85 + 8.44 × 10−5x, the linear formula of GS115/pKFS-CALB-EGFP is y = 48.53 + 8.76 × 10−5x, where x is fluorescence intensity (RFU/mL) and y is protein concentration (μg/mL).

In view of the alignment between the fusion protein anchored on the cell wall and the leakage into the medium, we extrapolated the formula above to quantify the protein quantification and calculate the number of molecules on the cell surface.

Paus et al. (2002) indicated that 1OD600 = ∼5 × 107 P. pastoris cells. Therefore from Avogadro’s number, one dalton (Da) has a value of 1.660538921 × 10−27 kg; thus, the number of fused protein molecules displayed per cell is \( y \times 6.02 \times 10^{23} /\left( {M/N} \right) \), where M is molecule weight (kDa) of fusion protein which is deemed to be 220 kDa because of glycosylation, N is cell number and y is protein concentration (μg/mL).

The numbers and trend of the fusion protein displayed on the P. pastoris cell surface during the fermentation process are indicated in Fig. 4. Compared with the number of displayed EGFP molecules per cell calculated by Shibasaki et al. (2001), an approx. order of magnitude as much as 104 per cell for the fusion protein displayed on P. pastoris surface based on FS anchor system could be obtained.

Fig. 4
figure 4

Number of fusion protein molecules displayed on P. pastoris per cell

Full size image

Discussion

Currently, yeast cell surface display systems have become effective tools for protein evolution and enzyme catalyst development. Many yeast surface systems have been developed using P. pastoris as a host (Kondo and Ueda 2004). However, there is no suitable method for quantifying displayed protein; thus, the efficiency of display systems cannot be enhanced and better enzyme display systems using yeast whole cell catalysts cannot be explored. Shibasaki et al. (2001) used S. cerevisiae as a host and a GPI-anchor system such as α-agglutinin. However, a new method for quantifying other displayed proteins in other display systems, particularly for P. pastoris, is necessary. In this work, we proposed a method for quantifying molecular numbers of a model enzyme such as CALB displayed on the cell surface of P. pastoris based on the FS anchor system.

In this study, we constructed strains GS115/pKFS-EGFP, GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP which were then used to display proteins on the P. pastoris cell surface. EGFP was used as a monitoring protein. The results from CLSM and FCL showed that the fluorescence of GS115/pKFS-EGFP is much brighter than that of GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP, but the fluorescence of GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP were similar. CALB activity and protein concentration analysis showed a similar trend for both GS115/pKFS-EGFP-CALB and GS115/pKFS-CALB-EGFP, which reached the same maximum value. Thus, the location of EGFP and CALB in the fused gene does not result in a distinct difference in CALB display.

SDS-PAGE and western blot results demonstrated that the protein displayed on the cell surface and that leaked into the medium had the same molecular weight of approx. 220 kDa and could recognize the EGFP antibody. Thus, the protein displayed on the cell surface and that leaked into the medium were the same and should have the same properties. Additionally, the results of SDS-PAGE combined with those of western blot analysis showed that only the target protein existed in the culture medium, so the fluorescence and protein concentration detected in the medium were due to EGFP-CALB leaked into the medium. During extraction and purification, the EGFP-CALB protein anchored to the cell wall may degenerate and emit green fluorescence. Thus, quantification of CALB displayed on the P. pastoris directly is quite difficult.

The relationship between fluorescence and the protein concentration of the fusion protein leaked into culture medium was linear. Using our designs and results above, the correlation between fluorescence intensity and protein concentration in the medium was found to be also appropriate for fused EGFP-CALB displayed on the cell surface. Based on our linear formula, we developed a method for quantifying protein displayed on the P. pastoris cell surface.

Compared to the results of Shibasaki et al. (2001), the number of the fusion proteins displayed on the surface of P. pastoris is at least in the same order of magnitude as the number of EGFP displayed on the surface of S. cerevisiae). Shibasaki et al. (2001) only observed a single EGFP protein without regard to whether this molecule was part of a fusion protein. Additionally, compared with S. cerevisiae, P. pastoris has more advantages, such as ease of growth to very high cell density and limited presence of other endogenous proteins in the culture medium (Monika et al. 2009). P. pastoris surface display systems appear more promising for industrial application. Quantification of proteins displayed on the P. pastoris surface may provide insight into the development of more efficient yeast surface display systems.