Current Microbiology

, Volume 56, Issue 4, pp 352–357

Construction of a Novel Pichia pastoris Cell-Surface Display System Based on the Cell Wall Protein Pir1

Authors

  • Qingjie Wang
    • State Key Laboratory of Microbial Technology, Life Science SchoolShandong University
  • Lei Li
    • State Key Laboratory of Microbial Technology, Life Science SchoolShandong University
  • Min Chen
    • State Key Laboratory of Microbial Technology, Life Science SchoolShandong University
  • Qingsheng Qi
    • State Key Laboratory of Microbial Technology, Life Science SchoolShandong University
    • State Key Laboratory of Microbial Technology, Life Science SchoolShandong University
Article

DOI: 10.1007/s00284-007-9089-1

Cite this article as:
Wang, Q., Li, L., Chen, M. et al. Curr Microbiol (2008) 56: 352. doi:10.1007/s00284-007-9089-1

Abstract

A novel system based on Pir1 from Saccharomyces cerevisiae was developed for cell-surface display of heterologous proteins in Pichia pastoris with the alpha-factor secretion signal sequence. As a model protein, enhanced green fluorescence protein (EGFP) was fused to the N-terminal of the mature peptide of Pir1 (Pir1-a). The expression of fusion protein EGFP-Pir1-a was irregular throughout the P. pastoris cell surface per detection by confocal laser scanning microscopy. A truncated sequence containing only the internal repetitive sequences of Pir1-a (Pir1-b) was used as a new anchor protein in further study. The fusion protein EGFP-Pir1-b was expressed uniformly on the cell surface. The fluorescence intensity of the whole yeast was measured by spectrofluorometer. Western blot confirmed that the fusion proteins were released from cell walls after mild alkaline treatment. The results indicate that a Pir1-based system can express proteins on the surface of P. pastoris and that the fusion proteins do not affect the manner in which Pir1 attaches to the cell wall. The repetitive sequences of Pir1 are required for cell wall retention, and the C-terminal sequence contributes to the irregular distribution of fusion proteins in P. pastoris.

Introduction

Cell-surface engineering in various microorganisms, especially in Saccharomyces cerevisiae, has been developed [6, 9, 20, 21]. Different groups of S. cerevisiae cell wall mannoproteins (CWPs) have been used as partners for protein-surface display: those both noncovalently [3, 14] and covalently linked to the cell wall [1, 23]. The latter has been classified as a glycosylphosphatidylinositol (GPI)-dependent CWP and part of the Pir (protein with internal repeats) CWP group [22].

Proteins expressed by S. cerevisiae are hyperglycosylated and have 50 to 150 mannoses/N-glycan, which may affect enzyme activity [8]. Compared with S. cerevisiae, Pichia pastoris—a methylotrophic yeast—sometimes contains hypermannosylated N-glycans together with shorter N-glycans (8 to14 mannoses) [5]. P. pastoris has been studied for protein-surface display based on the anchor systems from S. cerevisiae. Tanino et al. [24] immobilized the lipase from Rhizopus oryzae with a pro sequence on the P. pastoris cell surface using the flocculation functional domain of Flo1p (FS) from S. cerevisiae as an anchor [24]. This system was noncovalently linked to the cell wall and suitable for the enzymes possessing activities at the C-terminus [3]. Mergler et al. [15] displayed Kluyveromyces yellow enzyme on P. pastoris cell wall using the GPI anchor system from S. cerevisiae [15]. This system contains a C-terminal GPI anchor and is linked to the β-1,3-glucan of the cell wall by way of a short β-1,6-glucan bridge in S. cerevisiae, which is a better choice for enzymes that possess activities at the N-terminus [10, 12]. In this study, we successfully developed a novel P. pastoris display system, employing Pir1 from S. cerevisiae, to express the target protein on the yeast surface with the alpha-factor secretion signal sequence. Enhanced green fluorescence protein (EGFP) was used as a model protein. Pir1 is part of the CWP group, which attach directly to the β-1,3-glucan of the cell wall by way of alkali-sensitive linkages [11]. Pir proteins do not contain a C-terminal GPI anchor, and they are all processed by the Kex2 protease. Mature peptides have one or several units of an internal repetitive sequence at the N-terminus, which are integral to the ability of Pir CWP to bind to the cell wall [4, 7, 16]. The functions of different regions of Pir1 in P. pastoris were also investigated.

Materials and Methods

Strains and Media

Escherichia coli DH5α was used as the host for recombinant DNA manipulation. P. pastoris GS115 his4 (Invitrogen, Carlsbad, CA) was used as host for cell-surface display. E. coli was grown in Luria-Bertani medium at 37°C with 100 μg/mL ampicillin where necessary. Yeast was cultivated in buffered glycerol-complex medium (BMGY) and induced in buffered methanol-complex medium (BMMY) according the manufacturer’s directions (Pichia Expression Kit; Invitrogen).

Plasmid Construction and Transformation

The Pir1-a and Pir1-b genes were amplified by polymerase chain reaction (PCR) from the genomic DNA of S. cerevisiae EBY100, and the primers were derived from the published gene sequence of Pir1 (GenBank accession number D13740) (Fig. 1). The resulting fragments were digested at the EcoRI and NotI sites and then inserted into the EcoRI/NotI site of pPIC9k (Invitrogen), respectively, forming the plasmids of pPIC9k-Pir1-a and pPIC9k-Pir1-b. Using PCR, primers 4 and 5 were used to clone the EGFP gene from the plasmid pEGFPN1 (Clontech, Mountain View, CA) and inserted separately into the EcoRI/MluI site of the two reconstructed vectors (Fig. 1). The resulting plasmids were linearized by SalI and introduced into the host strain P. pastoris GS115 with the Gene Pulser Xcell Electroporation System (Bio-Rad, Hercules, CA). His+ transformants were selected on minimal dextrose medium (MD) plates and confirmed by PCR. The resulting strains were named GS115/Pir1-a-E and GS115/Pir1-b-E. Multicopy inserts of recombinant P. pastoris strains were selected by plating them onto YPD-G418 plates containing 4.0 mg/mL G418.
https://static-content.springer.com/image/art%3A10.1007%2Fs00284-007-9089-1/MediaObjects/284_2007_9089_Fig1_HTML.gif
Fig. 1

Construction of the cell-surface display vectors. (A) Vector pPIC9K. (B) Schematic representations of the mature Pir1 gene (Pir1-a) and the N-terminal repetitive sequences of Pir1-a (Pir1-b). (C) pPIC9K-Pir1-a and (D) pPIC9k-Pir1-b, novel constructed vectors for cell-surface display of heterologous protein in P. pastoris.(E) pPIC9K-Pir1-a-EGFP and (F) pPIC9k-Pir1-b-EGFP, plasmids for EGFP cell-surface display. S, alpha-factor secretion signal

Cultivation Conditions

The multicopy transformants were inoculated into 50 mL BMGY medium at 30°C to an optical density 600 (OD600) of 2 to 6. The cultures were centrifuged at 4,000 × g for 5 minutes and resuspended in BMMY medium to an OD600 of 1. Cells were induced at 20°C with 100% methanol every 24 hours to a final concentration of 1% [13]. P. pastoris GS115 was used as negative control.

Immunofluorescence Microscopy

After induction, the cells were collected and washed twice with phosphate-buffered saline (PBS). Resuspended cells were incubated first with primary rabbit immunoglobulin (Ig) G anti-EGFP antibody (eBioscience, San Diego, CA) for 1 hour at 1:125 dilution in PBS containing 1 mg/mL bovine serum albumin (BSA) at room temperature. They were then washed twice with PBS. The resulting cells were incubated with secondary rhodamine-conjugated goat antirabbit IgG antibody (KPL, Gaithersburg, MD) at 1:1000 dilution in PBS containing 1 mg/mL BSA at room temperature in darkness for 1.5 hours, washed twice, and suspended in PBS. Finally, cells were observed using the Leica TCS-SP2 confocal laser scanning microscopy (CLSM) system (Leica Microsystems, Heidelberg, Germany).

Whole-cell Fluorescence Measurements

The fluorescence of induced transformants was measured with a spectrofluorometer (FP-6500; Jasco, Tokyo, Japan) having 5-nm bandwidth. The excitation wavelength was 485 nm, and emission was 510 nm. The samples, which were collected at different time points (After 24 hours, 48 hours, 72 hours, 96 hours and 120 hours of induction), were washed twice with ice-cold PBS and adjusted to the same OD600 (i.e., 1) with PBS using an Heλios ultraviolet visible spectrophotometer (Thermo, Cambridge, UK) to avoid the “inner-filter” effect. In contrast, all of the EGFP fluorescence data were normalized using the maximal fluorescence signal [18].

Isolation of Pir1 Fusion Proteins from the Cell Wall and Study of the Localization of Fusion Proteins by Western Blot Analysis

After 96 hours of induction, growth medium was collected and concentrated by ultrafiltration, using a Millipore set-up with a membrane having a 3-kDa cutoff, according to the manufacturer’s instructions (Pall, Dreieich, Germany). The cells were centrifuged, and cell wall fractions were isolated according to Teparic et al. [25]. The remaining cell walls were incubated overnight in 30 mM NaOH at 4°C, followed by centrifugation at 10,000 × g for 5 minutes at 4°C. Extracts from the purified cell walls and the concentrated growth medium were analyzed by Western blot using primary rabbit IgG anti-EGFP antibody and secondary horseradish peroxidase-conjugated goat antirabbit IgG antibody (Xin Jing Ke Biotechnology, Beijing, PRC). The membrane was stained, in darkness, in a solution containing 100 mM Tris-HCl (pH 7.5), 0.8 mg/mL 3, 3’-diaminobenzidine, 0.4 mg/mL NiCl2, and 6 μL/mL H2O2.

Results

Construction of Universal Vectors for Cell-surface Display in P. pastoris

The intact mature Pir1 gene from S. cerevisiae (Pir1-a) (Fig. 1B) was inserted into the EcoRI/NotI site of pPIC9k (Fig. 1A), which is a secretory-expression plasmid with the alpha-factor prepro peptide sequence, to construct the plasmid pPIC9k-Pir1-a (Fig. 1C). An additional restriction site, MluI, was inserted before the N-terminus of Pir1-a for further N-terminal gene cloning (Fig. 1, primer 1). Pir1-a contains both the internal repetitive and C-terminal sequences. The resulting construct was confirmed by DNA sequencing using an ABI 3730 automatic sequencer (ABI, Foster City, CA).

Confirmation of EGFP Construction Displayed on the Cell Surface

To confirm the new cell-surface display system in P. pastoris, the plasmid pPIC9k-Pir1-a-EGFP (Fig. 1E) was constructed. After transformation, the recombinant GS115/Pir1-a-E transformants and GS115 (control) were induced in BMMY medium for 96 hours. The cells were immunostained and observed using CLSM. Figs. 2A and 2B show that fluorescence was detected on the cell surface of GS115/Pir1-a-E under both the fluorescein isothiocyanate (FITC) and rhodamine filters, whereas no fluorescence was observed from GS115 (Figs. 2D and 2E). The results strongly suggest that EGFP is expressed on the cell surface of P. pastoris, indicating that the construction of the new cell-surface display system was successful. However, detailed observation of the recombinant cells showed that fluorescence on the surface was irregular (Fig. 2A). Analysis of Pir1 in S. cerevisiae indicated that the C-terminus may be responsible for this irregular distribution [22]. To examine whether the same reason caused the irregular expression of EGFP on the surface of P. pastoris, a truncated sequence containing only the internal repetitive sequences of Pir1-a (Pir1-b) (Fig. 1B) was used as a new anchor gene and inserted into the EcoRI/NotI site of pPIC9k to construct the new plasmid pPIC9k-Pir1-b (Fig. 1D). The plasmid pPIC9k-Pir1-b-EGFP (Fig. 1F) was constructed for EGFP cell-surface display. As shown in Fig. 3A, fluorescence was uniformly distributed on the cell surface under the CLSM FITC filter after induction, indicating that the C-terminus of Pir1 plays a similar function in P. pastoris.
https://static-content.springer.com/image/art%3A10.1007%2Fs00284-007-9089-1/MediaObjects/284_2007_9089_Fig2_HTML.jpg
Fig. 2

Confocal laser scanning microscopy photographs of EGFP displayed on the cell surface. (A) GS115/Pir1-a-E under the FITC filter. (B) Immunofluorescence of GS115/Pir1-a-E under the rhodamine filter. (C) GS115/Pir1-a-E under normal white light. (D) GS115 under the FITC filter. (E) Immunofluorescence of GS115 under the rhodamine filter. (F) GS115 under normal white light

https://static-content.springer.com/image/art%3A10.1007%2Fs00284-007-9089-1/MediaObjects/284_2007_9089_Fig3_HTML.jpg
Fig. 3

CLSM images of GS115/Pir1-b-E cells. (A) Fluorescence micrograph under the FITC filter. (B) Contrast micrograph under normal white light

Fluorescence Intensity of EGFP Displayed on the Yeast Surface

To quantify the display procedure and optimize display time, the fluorescence intensity of whole yeast cells was measured after induction. Fig. 4 shows the time course of yeast cell fluorescence intensity after induction. The sum of the fluorescence of GS115/Pir1-a-E and GS115/Pir1-b-E increased with induction time and reached the highest value at 96 hours.
https://static-content.springer.com/image/art%3A10.1007%2Fs00284-007-9089-1/MediaObjects/284_2007_9089_Fig4_HTML.gif
Fig. 4

Evaluation of EGFP fluorescence intensity on the transformed cell surface. A relative fluorescence unit is an arbitrary unit measured with a spectrofluorometer under the conditions described in Materials and Methods. Error bars represent SDs of the data from three runs

Study of the Localization of EGFP-Pir1-a and -Pir1-b Fusion Proteins by Western Blot Analysis

To determine the correct localization of the fusion proteins, they were extracted using mild alkaline from the purified cell walls. The extracts from the purified cell walls of the different strains and concentrated samples of growth medium were analyzed by Western blot using an EGFP antibody. As shown in Fig. 5, both of the fusion proteins were detected in the extracts from the cell walls of GS115/Pir1-a-E and GS115/Pir1-b-E (Fig. 5, lanes 2 and 3), whereas neither was found in the supernatants after vigorously breaking yeast cells with glass beads (Fig. 5, lanes 5 and 6), indicating that they may attach to the cell wall in the same manner as native Pir1 in S. cerevisiae [17]. No specific polypeptides could be detected from the concentrated growth medium from GS115/Pir1-a-E and GS115/Pir1-b-E (Fig. 5, lanes 8 and 9), indicating that the fusion proteins were displayed mainly on the cell wall. The EGFP-Pir1-a fusion protein showed a band of approximately 120 kDa (Fig. 5, lane 2), and the EGFP-Pir1-b fusion protein showed a band >86 kDa (Fig. 5, lane 3), both of which were larger than the predicted protein sizes of 55 and 42 kDa, respectively.
https://static-content.springer.com/image/art%3A10.1007%2Fs00284-007-9089-1/MediaObjects/284_2007_9089_Fig5_HTML.gif
Fig. 5

Western blot analysis of the localization of the EGFP-Pir1-a and EGFP-Pir1-b fusion proteins. Lane 1 = GS115 (control). Lane 2 = EGFP-Pir1-a released from the cell walls of GS115/Pir1-a-E. Lane 3 = EGFP-Pir1-b released from the cell walls of GS115/Pir1-b-E. Lane M = Prestained protein molecular-weight marker (MBI, Lithuania). Lanes 4 to 6 = Supernatant of GS115, GS115/Pir1-a-E, and GS115/Pir1-b-E after vigorously breaking yeast cells with glass beads. Lanes 7 to 9 = GS115, GS115/Pir1-a-E, and GS115/Pir1-b-E concentrated growth media

Discussion

We succeeded in developing a new P. pastoris cell-surface display system based on Pir1 from S. cerevisiae. The Pir1 protein belongs to the Pir group of CWPs, which contain propeptides and are processed by the Kex2 protease at the Golgi [2]. The intact mature Pir1 protein (Pir1-a) starts downstream of the Kex2p cleavage site and contains internal repetitive and the C-terminal sequences. Recently, P. pastoris cell-surface display systems have been constructed using the GPI-anchor system and the FS anchor from S. cerevisiae [15, 24]. Given the development of genetic engineering, more cell-surface display systems based on S. cerevisiae are expected to be developed in P. pastoris to use the limited resources of the yeast surface.

In our study, fusion protein EGFP-Pir1-a was successfully but irregularly expressed throughout the yeast cell surface (Fig. 2). Previous reports have shown that the display of glycosyltransferases on the surface of S. cerevisiae with the Pir1 system was also irregular [1]. The irregular immobilization of the functional protein may affect its action with the substrate [19]. Analysis of Pir1 in S. cerevisiae indicated that the C-terminus may be responsible for its irregular distribution [22]. Therefore, a new cell surface-display system with only internal repetitive sequences of Pir1-a as an anchor protein was employed. As shown in Fig. 3, the new system displayed EGFP uniformly on the cell surface, indicating that the C-terminal sequence of Pir1 has a similar function in P. pastoris.

Fluorescence-intensity measurement of the whole yeast cell showed that the display efficiency of both constructs is similar (Fig. 4), indicating that the repetitive sequences are important for the ability of Pir1 to bind to the cell wall in P. pastoris. The EGFP-Pir1-a and EGFP-Pir1-b fusion proteins were released by mild alkaline treatment (Fig. 5), indicating that they may be covalently linked to the cell wall as in S. cerevisiae [16]. No specific polypeptides could be detected from the concentrated growth medium, indicating that the fusion proteins were retained mainly at the cell wall. The molecular mass of the fusion proteins was larger than those of the predicted values, implying that they may be highly glycosylated. Tanino et al. [24] also reported that the protein size of the fusion protein in P. pastoris using the FS anchor was much larger than the calculated value because of high O-glycosylation [24].

In a summary, we constructed a novel P. pastoris cell-surface display system based on the Pir1 CWP of S. cerevisiae. Employing EGFP as a model protein, the system was confirmed to be successful and efficient. Meanwhile, our results also indicated that the C-terminal sequence of Pir1 is responsible for its irregular distribution in P. pastoris.

Acknowledgments

This project was supported by a key fund from the Chinese Education Ministry (2005) and a grant from the National Natural Science Foundation of China (Grant No. 30470399).

Copyright information

© Springer Science+Business Media, LLC 2008