C H A P T E R T H I R T E E N

Expression in the Yeast

Pichia pastoris
James M. Cregg,*,† Ilya Tolstorukov,*,† Anasua Kusari,*
Jay Sunga,* Knut Madden,† and Thomas Chappell†

Contents
1.Introduction 170
2.Other Fungal Expression Systems 170
3.Culture Media and Microbial Manipulation Techniques 171
4.Genetic Strain Construction 172
4.1. Mating, creating diploids 172
4.2. Random spore analysis 173
5. Gene Preparation and Vector Selection 174
6. Transformation by Electroporation 176
7. DNA Preparation 176
8. Examination of Strains for Recombinant Protein Production 178
9. Assay Development—The Yeastern Blot 182
10. Posttranslational Modification of the Recombinant Protein
(Proteinases and Glycosylation) 184
11. Selection for Multiple Copies of an Expression Cassette 185
References 187

Abstract
The yeast Pichia pastoris has become the premier example of yeast species
used for the production of recombinant proteins. Advantages of this yeast for
expression include tightly regulated and efficient promoters and a strong
tendency for respiratory growth as opposed to fermentative growth. This chapter
assumes the reader is proficient in molecular biology and details the more
yeast specific procedures involved in utilizing the P. pastoris system for gene
expression. Procedures to be found here include: strain construction by classical
yeast genetics, the logic in selection of a vector and strain, preparation of
electrocompetent yeast cells and transformation by electroporation, and the
yeast colony western blot or Yeastern blot method for visualizing secreted
proteins around yeast colonies.

* Keck Graduate Institute of Applied Life Sciences, Claremont, California, USA

{
Biogrammatics, Inc., Carlsbad, California, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63013-5 All rights reserved.

169
170 James M. Cregg et al.

1. Introduction
Relative to other expression systems, yeast got off to a slow start in the
early 1980s, primarily due to poor results with several proteins in baker’s
yeast Saccharomyces cerevisiae (Romanos et al., 1992). Promising results in
shake flask culture, more often than not, led to disappointing yields when
scaled up in fermentor cultures. In addition, yeast were not likely to be of
value in producing human proteins containing N-linked carbohydrates as
injectable pharmaceutical drugs, a major goal of many biotechnology and
pharmaceutical companies, because their sugars were of a high mannose
type in composition and configuration, quite different from that of humans.
As a result, recombinant glycoproteins made in a yeast system have a typical
fungal-like N-glycosylation pattern, which a human immune system recog-
nizes as foreign and rejects. This results in the rapid clearance of yeast
products from the blood and a strong immune response from a patient
that can possibly result in death. Since the 1980s, these problems have been
addressed and several yeasts have become productive alternative systems for
recombinant protein production. As a eukaryotic microbial expression
system, yeast are a good alternative for proteins for which expression in a
bacterial system leads to the synthesis of improperly folded, and inactive
protein aggregates or inclusion bodies.
This review will focus primarily on the most popular of these new yeast
expression systems, Pichia pastoris. Only details of procedures that are specific
or peculiar to expression in P. pastoris will be covered; for more common
methods (e.g., agarose gel electrophoresis, immunoblotting, general recom-
binant DNA methodology, etc.), readers are referred to the many excellent
books describing these methods including: (Sambrook et al., 1989) or the
series by (Ausubel et al., 2001).

2. Other Fungal Expression Systems

In addition to P. pastoris, several other yeast species have been devel-
oped for expression; among them, Hansenula polymorpha, Pichia methanolica,
Kluyveromyces lactis, Arxula adeninivorans, and Yarrowia lipolytica are the best
known and developed (Gellissen, 2005). Many of the reasons for using one
of these alternative yeast expression systems, as well as the methods needed
to construct recombinant strains, are similar to each other and to P. pastoris.
Like P. pastoris, H. polymorpha and P. methanolica are methylotrophic
yeast and virtually all of the advantages cited for P. pastoris are also true for
these other species (Gellissen, 2002). In particular, all three have the potent
promoter regulating the expression of the alcohol oxidase gene from their
Expression in the Yeast Pichia pastoris 171

respective species available for controlling recombinant protein expression.

However, both H. polymorpha and P. methanolica are different from
P. pastoris in that the construction of expression strains is often somewhat
more laborious. With H. polymorpha, this extra labor can be rewarded with
strains harboring 100 or more copies of an expression cassette. In addition,
H. polymorpha can readily grow at temperatures up to 50
C providing the
potential of significantly decreasing bioprocess times and thereby reducing
costs.
K. lactis was the first yeast species after S. cerevisiae to be developed for
recombinant protein expression and is well known for its use in production
of rennin for cheese processing (van Ooyen et al., 2006). An advantage of
K. lactis for some purposes is that, like S. cerevisiae, it is officially on the
generally recognized as safe (‘‘GRAS’’) list of microorganisms.
A. adeninivorans is a dimorphic yeast with advantages for secreting recom-
binant proteins (Boer et al., 2005). This yeast has both a vegetative yeast like
growth state and a mycelial growth state in which the cells send out mycelia
like filamentous fungi. During the mycelial growth phase, the synthesis
of secreted proteins is enhanced relative to its vegetative growth phase.
In addition, A. adeninivorans does not O-glycosylate secreted proteins in its
mycelial growth phase. Finally, A. adenininvorans has a relatively high growth
temperature of up to 48
C and osmotolerance up to 3.4 M (10%) NaCl.
Y. lipolytica, like A. adeninivorans is a dimorphic yeast (Madzak et al.,
2005). Most recombinant genes are expressed in Y. lipolytica off the XPR2
promoter which, in its native state, expresses the gene for alkaline extracellular
protease. However, several other promoters, all of which are constitutive, are
also available for the yeast.
Finally, certain filamentous fungi, such as Neurospora crassa, various
Aspergillus species and Sordaria macrospora, have proved to be effective
expression systems for certain recombinant products, particularly secreted
proteins. However, the techniques for dealing with filamentous fungi
are very different from yeast and will not be dealt with here. Readers
interested in expression in filamentous fungi are referred to one of several
excellent reviews on these systems (Heerikhuisen et al., 2005; Kuck and
Poggeler, 2005).

3. Culture Media and Microbial

Manipulation Techniques
Techniques for culturing P. pastoris (and most other yeast species) at
the bench level are identical to those used for Escherichia coli and S. cerevisiae.
The most common rich medium for cultivation is YPD (1% yeast extract,
172 James M. Cregg et al.

2% peptone, 2% dextrose) and defined medium is YNB (0.67% yeast

nitrogen base with ammonium sulfate and without amino acids, 2% dex-
trose plus any amino acids or nucleotides required for growth at  50 ug/ml
each). Growth of P. pastoris on methanol requires that the dextrose is
replaced with methanol to 0.5%. The recipe for mating/sporulation
medium is 0.5% sodium acetate, 1% KCl 1% glucose. For Petri plates the
media are prepared with 2% agar. Incubations are typically done at 30
C. In
liquid YPD, P. pastoris has a generation time of approximately 90 min and a
generation time of approximately 3 h in defined medium. With methanol as
sole carbon source and a defined culture medium, the generation time is
around 5 h.

4. Genetic Strain Construction

Although a number of markers of various types exist for P. pastoris, the
right combination for your purposes may not exist. Therefore, it may be
necessary to construct a new host strain with the optimal set of genetic
markers. The first step in strain construction is the mating and selection of
diploid strains (Tolstorukov and Cregg, 2008). Because P. pastoris is func-
tionally homothallic, the mating type of a strain is not a consideration in
planning a genetic cross as cells of the same strain will also mate; However,
the mating efficiency between P. pastoris cells is low. Therefore, it is
essential that strains to be crossed contain complementary markers that
allow for selective growth of crossed diploids, and against the growth of
self-mated diploids and parental strains. Auxotrophic markers are generally
most convenient for this purpose, but mutations in any gene that affect the
growth or other phenotype of P. pastoris such as genes required for utiliza-
tion of methanol or a nitrogen source (e.g., methylamine) can be used as
well. Following are the steps needed to mate P. pastoris:

4.1. Mating, creating diploids

1. To begin a mating experiment, select a fresh colony from each strain to
be mated from YPD plate (no more than one week old) and streak each
across the length of an independent YPD plate.
2. After overnight incubation at 30
C, transfer the cell streaks from both
plates onto a single sterile velvet such that the streaks from one plate are
perpendicular to those on the other.
3. Transfer the cross streaks from the velvet to a mating/sporulation plate
and incubate for 2–3 days at room temperature to initiate mating.
4. After incubation, replica plate to an appropriate agar medium for the
selection of complementing diploid cells. Diploid colonies will grow at
Expression in the Yeast Pichia pastoris 173

the junctions of the streaks after approximately 2–3 days of incubation at

30
C. Diploid P. pastoris cells are approximately twice as large as haploid
cells and easily distinguished by examination under a light microscope by
their size and their high efficiency of sporulation.
5. Purify diploid colonies by streaking at least once for single colonies on
diploid selection medium.
6. Diploid P. pastoris strains efficiently undergo meiosis and sporulation in
response to nitrogen limitation. To initiate this phase of the life cycle,
transfer freshly grown diploid colonies from a YPD or YNB plus glucose
plate to a mating/sporulation plate either by replica-plating or with an
inoculation loop; incubate the plate for 3–4 days at room temperature.
Sporulation in all Pichia species correlates with accumulation of a red
pigment in the ascus; therefore, sporulated diploid samples are easily
distinguished by their tan color relative to the white color of haploid
cultures. Diploids can also be distinguished by a high number of asci in
the cell culture as observed by normal or phase-contrast microscopy.

4.2. Random spore analysis

P. pastoris spores are small and adhere to one another making tetrad dissec-
tion via micromanipulation difficult. Therefore, spore products are analyzed
using a random spore analysis (RSA), as follows:
1. Transfer an inoculation loop full of sporulated P. pastoris cells (from Step 6
above) to a 1.5 ml microcentrifuge tube containing 0.25 ml of sterile
water, vortex the mixture.
2. In a fume hood, add 0.25 ml of diethyl ether to the spore preparation and
vortex thoroughly for approximately 5 min at room temperature. The
ether treatment selectively kills vegetative cells remaining in spore
preparations.
3. While still in the hood, centrifuge the cells for 2 min, remove the ether
(upper phase) and resuspend the pellet in the remaining water. Spread
10 and 100 ul of the suspension onto a nonselective YPD plate.
4. After 4–5 days incubation at 30
C, streak out single colonies onto a
fresh YPD plate as a master plate for further analysis.
5. Replica plate the master plate onto a set of plates containing suitable
diagnostic media. Alternatively/additionally, the initial plates with
100–600 colonies can also be replica plated. For example, spore products
from a his4 and arg4 cross would be analyzed on YNB plus glucose
supplemented with:
a. No amino acids
b. Arginine
c. Histidine
d. Arginine and Histidine
174 James M. Cregg et al.

6. Compare the phenotype of individual colonies on each of the diagnostic

plates to identify strains with the desired phenotype(s).
7. Select several colonies that appear to have the appropriate phenotype
and streak for single colonies onto a nonselective medium plate such as
YPD; then, retest a single colony from each streak on the same set of
diagnostic plates. This step is important since P. pastoris spores adhere
tightly to one another and colonies resulting from spore germination
frequently contain cells derived from more than one spore. Another
consequence of spore clumping is that markers appear not to segregate
1:1 but to be biased toward the dominant or wild-type phenotype. For
example, in ahis4  arg4 cross as described above, more HisþArgþ spore
products will be apparent than the 25% expected in the population and
HisArg spore products will appear to be underrepresented.

5. Gene Preparation and Vector Selection

The first consideration in the selection of a P. pastoris expression vector
is whether you intend to secrete a protein product or produce it intracellu-
larly. A general rule of thumb is to produce a recombinant protein in the
same way it is expressed in its native host: if a protein is produced intracel-
lularly by its native host, one should also produce it intracellularly in the
yeast host; if the protein is secreted from its native host, secrete it from the
yeast system. Although there have been exceptions to this general rule, it is
generally best to follow it since the intracellular and secretory environments
are very different from each other and synthesis of a protein in the wrong
compartment may result in a protein that is improperly folded and inactive.
A number of vectors have been constructed for P. pastoris expression; a list
and detailed discussion of these vectors can be found in (Lin-Cereghino and
Lin Cereghino, 2008) and at http://www.biogrammatics.com (Fig. 13.1).
To clone a gene into a P. pastoris expression vector, an available template
can be amplified with appropriate primers by the polymerase chain reaction
(PCR), or the gene can be synthesized. De novo synthesis can facilitate cases
where DNA optimization or gene modification is required. In any case,
suitable restriction sites at the termini can be generated to facilitate the
cloning. For example, an EcoRI site can be added to the 50 (ATG-contain-
ing) end of the gene and a NotI site to the 30 end of the gene, to facilitate
cloning into the pPICZ series of vectors (Invitrogen, Carlsbad, CA), as long
as sites for these enzymes do not exist within the sequence of the gene. For
P. pastoris expression vectors from Biogrammatics, Inc., an appropriate type
IIS restriction enzyme site and ‘‘seamless’’ cloning sequences can be added
to flank a gene for cloning into an optimal expression context.
Expression in the Yeast Pichia pastoris 175

A Bsa I Bsa I AOX1 TT

Alpha pre- pro-

A. gossypii TEF promoter

AOX1 promoter
Pme I
ZeoR

pJAZ-aMF

pMB origin
ampR

Sfi I

B GGCT ORF
ATTA

TTGGACAAGAGAGAGGCTGA TAATCAAGAGGATGT
AACCTGTTCTCTCTCCGACTCCGA GTTCTCCTACA
L D K R E A E A

Figure 13.1 Map of Biogrammatics expression vector, pJAZ-aMF.

Cloning a gene whose product is to be secreted can be a little trickier.

Extracellular, or secreted expression using a genes native secretion signal
sequence will follow the same procedures as above for intracellular expres-
sion; however, several options exist when using a foreign signal sequence,
such as the S. cerevisiae alpha mating factor secretion signal sequence (aMF),
encoded in the expression vector. In this case, the subcloning procedure
must place the gene of interest in frame with the aMF codons. This aMF
secretion signal is very commonly utilized for secretion because it has
proved to be very good at secreting recombinant proteins of many types.
Although the recombinant protein may be successfully secreted using the
aMF signal, proper processing for the aMF at the NH2-terminus of
the desired protein may not occur and modifications of the aMF signal,
or the use of an alternative secretion signal sequence, may ultimately be
necessary to obtain a properly processed protein. In this regard, the aMF
signal sequence, may include two Glu–Ala repeats at the junction between
the signal peptide and the NH2-terminus of the mature protein of interest.
The Biogrammatics vector, pJAZ-aMF, is designed for making a construction
176 James M. Cregg et al.

with these Glu–Ala repeats (Fig. 13.1). Another set of Alternative Biogram-
matics vectors such as pJAZ-aMF-KR do not contain the Glu–Ala repeats in
the cloning site. To utilize the aMF signal in a pPICZ alpha vector one must
add sequences to the 50 of the gene such that when ligated to the portion of
aMF present in the vector it results in the reconstruction of a functional aMF
signal sequence (either with or without the sequences encoding the Glu–Ala
repeats in aMF). Typically, XhoI is used to cut the pPICZ vector which cuts it
inside the aMF signal encoding sequences. To restore these sequences, an
oligonucleotide with the sequence:
XhoI
TCT CTC GAG AAA AGA GAG GCT GAA GCT-ABC-DEF. . ..
Ser Leu Glu Lys Arg Glu Ala Glu Ala
is synthesized where ‘‘ABC DEF. . .’’ denotes the nucleotide sequences
encoding the first amino acids of the mature recombinant protein. This
oligonucleotide will hybridize appropriately to the 50 end of the gene
encoding the mature protein and result in the incorporation of the missing
portion of the aMF sequence. Similarly, the ‘‘seamless’’ cloning scheme
used to clone genes into the Biogrammatics vectors utilize type IIS restric-
tion sites to join the ABC DEF nucleotides of a gene of interest to the last
Alanine in the aMF by creating a four base ‘‘sticky’’ end comprising the last
nucleotide in a Glu codon and an entire Ala codon (Fig. 13.1).

6. Transformation by Electroporation
At least four different procedures to introduce foreign plasmid DNA
into P. pastoris have been developed using: spheroplast-generation, LiCl,
polyethelene glycol1000 and electroporation. The electroporation proce-
dure is most commonly used and therefore a modified version of that
described by (Becker and Guarante, 1991) will be outlined in detail. For
the other procedures, readers are referred to either of the volumes of
Methods in Molecular Biology: Pichia Protocols (Cregg, 2008; Higgins
and Cregg, 1998).

7. DNA Preparation
For all transformation methods, linear plasmid DNA is most commonly
transformed into P. pastoris for integration into the yeast genome. The DNA
sequence at the ends of the linear plasmid DNA stimulate integration by a
single crossover recombination event into the locus shared by vector and host
genome. Therefore, linearization of an expression plasmid is performed in
Expression in the Yeast Pichia pastoris 177

Pichia DNA in the plasmid, such as in the promoter (i.e., a PmeI site in the
AOXI promoter). The final vector, prepared in E. coli, is cut with a restriction
enzyme that linearizes the vector, and then the DNA is purified and concen-
trated to at least 100 ng/ul in water prior to transformation. At this point, the
vector is ready for transformation into P. pastoris.
Procedure for preparation of electrocompetent cells (Lin-Cereghino
et al., 2005).
Prepare the following (all solutions should be autoclaved except for the
DTT and HEPES solutions, which should be filter sterilized):
1. 500 ml liquid YPD media in a 2.8 l Fernbach culture shaking flask.
2. H2O (1 l).
3. 1 M sorbitol (100 ml).
4. Appropriate selective agar plates.
5. 1 M DTT (2.5 ml).
6. BEDS solution (9 ml): 10 mM Bicine–NaOH, pH 8.3, 3% ethylene
glycol, 5% DMSO (dimethyl sulfoxide), 1 M sorbitol and 0.1 M DTT.
7. 1 M HEPES buffer, pH 8.0 (50 ml).
8. Sterile 250 ml centrifuge tubes.
9. Sterile electroporation cuvettes.
10. Electroporation instrument: BTX Electro Cell Manipulator 600 (BTX,
San Diego, CA); Bio-Rad Gene Pulser (Bio-Rad, Hercules, CA);
Electroporator II (Invitrogen, San Diego, CA). Parameters for electro-
poration with different instruments vary with the instrument. Be sure
to check instructions for each type of instrument. (Becker and
Guarante, 1991; Grey and Brendel, 1995; Pichia Expression Kit
Instruction Manual; Stowers et al., 1995).
Protocol:
1. Inoculate 10 ml YPD media with a single fresh P. pastoris colony of the
strain to be transformed from an agar plate and grow overnight shaking at
30
C.
2. Use the overnight culture to inoculate a 500 ml YPD culture in a 2.8 l
Fernbach culture flask to a starting OD600 of 0.01 and grow to an OD600 of
1.0 ( 12 h).
3. Harvest the cells by centrifugation at 2000g at 4
C, discard the superna-
tant and resuspend the cells in 100 ml of fresh YPD medium plus HEPES
(pH 8.0, 200 mM ) in a sterile 250 ml centrifuge tube.
4. Add 2.5 ml of 1 M DTT and gently mix.
5. Incubate at 30
C for 15 min with slow rotating.
6. Add 150 ml cold water to the culture and wash by centrifugation at 4
C
with an additional 250 ml of cold water. At this stage and from here on,
keep the cells ice cold and do not vortex the cells to resuspend them
(slow pippeting is best).
178 James M. Cregg et al.

7. Wash cells a final time in 20 ml of cold 1 M sorbitol, then resuspend in 0.5 ml

of cold 1 M sorbitol (final volume including cells will be 1.0–1.5 mls).
8. Use these cells directly without freezing to achieve the most transformants.
9. To freeze competent cells, distribute in 40 ul aliquots to sterile 1.5 ml
minicentrifuge tubes and place the tubes in a  70
C freezer.
Electroporation procedure:
1. Add up to 1 ug of linearized plasmid DNA sample in no more than 5 ul
of water to a tube containing 40 ul of frozen or fresh competent cells and
then transfer the entire mixture to a 2 mm gap electroporation cuvette
held on ice.
2. Pulse cells according to the parameters suggested for yeast by the manu-
facturer of the specific electroporation instrument (Table 13.1).
3. Immediately add 0.5 ml of cold 1 M sorbitol and 0.5 ml of cold YPD,
then transfer the entire cuvette contents to a sterile 1.5–2.0 ml mini-
centrifuge tube.
4. Incubate for 3.5–4 h at 30
C with slow shaking (100 rpm).
5. Spread aliquots onto selective agar plates and incubate for 2–4 days.
6. To avoid mixed colonies, pick and restreak transformants on selective
medium at least once before proceeding with further analysis.
Rapid preparation of electrocompetent P. pastoris:
1. Grow a 5 ml culture of P. pastoris in YPD overnight, shaking at 30
C.
2. Dilute the overnight culture to an OD600 of 0.15–0.20 in 50 ml YPD
medium in a flask large enough to provide good aeration.
3. Grow to an OD600 of 0.8–1.0 at 30
C with shaking (4–5 h).
4. Centrifuge cells at 500g for 5 min at room temperature and discard
supernatant.
5. Resuspend cells in 9 ml of ice-cold BEDS solution supplemented with
DTT.
6. Incubate the cell suspension for 5 min, shaking at 30
C.
7. Centrifuge cells at 500g for 5 min at room temperature and resuspend in
1 ml of BEDS (without DTT).
8. Perform electroporation as described above, immediately or freeze cells
in small aliquots at  80
C.

8. Examination of Strains for Recombinant

Protein Production
Yeast expression systems have been successful at generating large
quantities of recombinant proteins. For example, production levels of
between 1 and 10 g/l of culture supernatant have been secreted from
Table 13.1 Parameters for electroporation using selected instruments

Cuvette Sample Charging Field Pulse

gap volume voltage Capacitance Resistance strength length
Instrument (mm) (ul) (V) (uF) (O) (kV/cm) ( ms) References
ECM600 (BTX) 2 40 1500 Out 129 7500 5 Becker and
Guarante
(1991)
Electroporator II 2 80 1500 50 200 7500 10 Pichia Manual
(Invitrogen)
Gene-Pulser 2 40 1500 25 200 7500 5 Grey and Brendel
(Bio-Rad) (1995)
Cell-Porator 1.5 20 480 10 Low 2670 NS Lorow-Murray
(BRL) and Jesse (1991)
and Stowers
et al. (1995)
180 James M. Cregg et al.

P. pastoris strains. However, one should not expect to see a band on a

coomassie-stained gel in initial expression experiments. Thus, it is impera-
tive that by the time a recombinant strain is ready to begin expression
studies, one or more sensitive assays for the detection of the targeted protein
is in place. Assays can be based on enzyme activity, an epitope tag fused to
ones product or an antibody against the desired gene product; however, the
point is: serious efforts to develop sensitive detection methods should begin
at the same time or even prior to expression studies. Good assays facilitate
strain development as illustrated in the following examples.
The most convenient way to detect foreign protein expression in yeast is
via a plate activity assay. Plate assays allow one to crudely quantify and
compare productivity of 100s to 1000s of transformants directly on diag-
nostic plates at a single glance using replica-plating or other techniques.
An example of a plate activity assay is shown in Fig. 13.2 for secretion of
the enzyme phytase. This enzyme degrades phytate which results in the
clearing of a zone around the expressing colony on agar plates. The size of
the zone roughly correlates to the amount of phytase being secreted. As a
second example, the expression of bacterial b-lactamase as an intracellular
protein is shown in P. pastoris (Fig. 13.3). The colonies, which are typically
yellow in color, turn pink to purple with the expression of b-lactamase.
The intensity of the purple color roughly indicates the amount of enzyme
expressed and, in this case, the number of copies of the expression cassette in

Figure 13.2 Plate assay for detection of phytase constitutively expressed and secreted
by selected P. pastoris strains. Top spot: negative control; remaining spots show five
transformants secreting various amounts of the enzyme.
Expression in the Yeast Pichia pastoris 181

Figure 13.3 Expression of intracellular b-lactamase in P. pastoris. The two spots at the
top of the plate are non-b-lactamase expressing negative control strains.

each strain. Finally, with good quality polyclonal or monoclonal antibodies

against a tag region or the recombinant protein itself, a plate antibody assay
or ‘‘Yeastern Blot’’ is possible, as described below.
Once colonies of transformed cells have been selected, single cell pur-
ified and collected onto a master plate, samples of multiple strains are
examined for expression of the foreign protein. Before analysis of expres-
sion, one can screen for the presence of the recombinant gene in transfor-
mants by PCR. Simply prepare genomic DNA in a cell-free extract by glass
bead disruption as described below and utilize it as template in a PCR
reaction with primers that are complementary to a portion of the recombi-
nant gene. Different methods of detection for the expression of proteins in
P. pastoris can be applied depending on what kind of promoter is used for
expression (inducible or constitutive), and what kind of vector is used
(intracellular or secretory):
A. If the gene of interest is expressed constitutively, inoculate a sample
colony into YPD medium, grow the culture for 2–3 days with good
aeration and analyze the proteins in samples taken periodically during
this time by any available detection method.
B. For methanol-induced expression, a colony should be grown for
17–24 h in YPD and then transferred to a fresh methanol-containing
medium for induction. The induction can be performed in 15-ml tubes
182 James M. Cregg et al.

with 2 ml medium containing 0.5% methanol with good aeration at a

starting OD600 of about 10. For intracellular proteins an induction time
of 6–12 h in methanol medium is sufficient.
Intracellularly expressed samples require the preparation of cell extracts
for analysis. Culture samples can be harvested after 10–12 h and prepared for
cell breakage by glass bead disruption. Harvest approximately 10–50 OD600
units of cells and resuspend in 150 ul of a breakage buffer. Add an equal
volume of sterile acid washed 0.45 mm diameter glass beads. Vortex the
mixture on the highest setting for 1 min, then place the sample on ice for
1 min; repeat this process at least five times. Alternatively, load samples into
a vortexer with a head made to hold multiple microfuge tubes. Place the
vortexer in a 4
C cold box or cold room and vortex on high for approxi-
mately 10 min. Examine cultures for cell breakage under a microscope,
80–90% of the cells should be disrupted. After disruption, draw off the cell
debris and buffer into a fresh microcentrifuge tube. Rinse the beads with an
additional 100 ul portion of buffer by a brief vortexing and transfer the wash
to the tube with the rest of the cell debris. Centrifuge the samples on high
speed for 5 min at 4
C. Draw off the top liquid phase containing your
protein and transfer to a fresh microcentrifuge tube. This is your crude
protein sample ready for SDS–PAGE, western blotting or enzyme assay.
Secreted proteins build up in the medium much more slowly and
require at least 2 days to reach high levels. Allow the cultures to incubate
with shaking for 2–5 days. Add fresh methanol to a final concentration of
0.5% every 12 h and collect 50–100 ul supernatant samples during this
induction period. The supernatant samples are ready for SDS–PAGE,
western blotting or enzyme assay and can be stored at  20
C.

9. Assay Development—The Yeastern Blot

There is much in common in different antibody assays. One useful and
yeast-specific antibody assay is the yeast colony western blot, sometimes
referred to as the ‘‘Yeastern’’ blot for secreted proteins (Fig. 13.4). For
standard western blot assay procedures, you are referred to (Sambrook et al.,
1989).
The Yeastern blot is a convenient means of qualitatively screening large
numbers of yeast colonies directly on plates for expression of a recombinant
protein. However, readers should be aware that the correlation between the
size of the ‘‘halo’’ surrounding a colony and the amount of recombinant
product is not always linear and all results with this method should be
confirmed with a standard western blot or other assay. The procedure for
Yeastern blotting is as follows:
Expression in the Yeast Pichia pastoris 183

Figure 13.4 ‘‘Yeastern Blot’’ of P. pastoris colonies secreting both heavy and light
chains of IgG antibodies. Negative controls of P. pastoris strain that does not secrete
antibody shown on second row second streak from the left and at first two positions
from the left on bottom row.

1. Transfer freshly grown yeast colonies from the surface of an agar Petri
plate onto a sterile piece of Whatman No. 1 filter paper by a standard
replica-plating method. The filter paper should be cut to a size that
exactly fits within the plate.
2. Place the filter with yeast cells onto the surface of a fresh plate containing
an appropriate induction medium and incubate the plate for 1–2 days.
3. Prepare a piece of nitrocellulose membrane the same size and dimensions
as the filter paper by soaking the membrane for 5 min or more in 15 ml of
transfer buffer (25 mM Tris base, pH 8.5, 0.2 M glycine, 20% methanol).
Soak two additional pieces of cut filter paper in transfer buffer.
4. Prepare a sandwich of the papers as described below:
a. Place one piece of the filter paper on the anode platform of a western
blotter. (It is essential to remove all bubbles between membrane
layers. This can be done by rolling a pipette over their surface.).
b. Place the soaked nitrocellulose filter on top of the filter paper.
c. Place the filter paper with replica-plated yeast cells on top of the
nitrocellulose paper.
d. Place the second piece of soaked filter paper on top of the filter with
cells.
e. Finally, place the cathode plate on top of the sandwich.
5. Transfer proteins to the nitrocellulose membrane with a constant
current (1–4 mA/cm2) for 1 h.
6. Remove the nitrocellulose membrane and wash it for 5 min in 15 ml
TBS buffer (50 mM Tris–HCl, pH 7.6, 150 mM NaCl). Replace the
TBS buffer with 15 ml of TBST (TBST is prepared by adding Tween-
20 to 0.05% to TBS) containing 1–5% bovine serum albumin (BSA).
184 James M. Cregg et al.

Place filter in blocking buffer and rock gently at room temperature

for 1 h.
7. Transfer filter to 15 ml of TBST buffer and wash by rocking at room
temperature for 5 min.
8. Prepare primary antibody dilution in 15 ml of blocking buffer accord-
ing to the vendor’s recommendations (typically, 0.5 ug/ml).
9. Incubate membrane overnight at 2–8
C with rocking action (can be as
short as 1–3 h at room temperature depending on antibody).
10. Wash membrane in at least four changes of TBST buffer (15 ml each,
5 min/wash) then briefly rinse in fresh TBST.
11. Prepare an enzyme-conjugated secondary antibody dilution in block-
ing buffer according to the vendor’s recommendations (typically 3 ul in
15 ml of blocking buffer).
12. Incubate membrane in secondary antibody solution for 1 h, rocking at
room temperature.
13. Repeat Step 10.
14. Treat membrane in the dark room with visualization reagents accord-
ing to the vendor’s recommendations (e.g., Pierce ECL Western Blot-
ting Substrate), remove excess liquid by blotting with filter paper and
place the filter in a plastic protector and expose to X-ray film for
anywhere from 30 s to several minutes. The intensity and size of the
signal around each colony approximately reflects the secretion level of
the recombinant product by the different colonies/strains.

10. Posttranslational Modification of the

Recombinant Protein (Proteinases and
Glycosylation)
Posttranslational problems that affect the quality of recombinant pro-
teins expressed in P. pastoris include proteolysis and glycosylation.
A proteinase deficient strain of P. pastoris can be tested if initial results by
SDS–PAGE analysis (coomassie staining or western blots) suggest proteo-
lytic degradation of the protein. Signs of proteolysis include low recombi-
nant protein levels and active or immunoreactive products that are smaller
than the full-length product. Degraded protein can also run as a ‘‘smear’’
after PAGE, running from approximately the correct size of the product to
smaller sizes. The Pichia strain SMD1168 ( pep4 his4) is deleted for much of
the PEP4 gene (Gleeson et al., 1998). The PEP4 gene product is responsible
for activating many of the proteases in the vacuole of P. pastoris which enter
the vacuole as inactive zymogens and are activated there by the PEP4
product. Although secreted recombinant proteins do not go to the vacuole,
they can contact proteases in the culture medium from the lysis of a small
Expression in the Yeast Pichia pastoris 185

portion of the cells. Due to the extremely high culture densities used with
P. pastoris, the concentration of vacuolar proteases in the medium can be
considerable. To utilize a PEP4 strain, one must transform one of these
strains (i.e., SMD1168 ( pep4 his4) or SMD1168H ( pep4)) with the expres-
sion vector or create deletions in the PEP4 gene in a strain already expres-
sing a protein of interest. To determine if PEP4 deficient strains benefit
the expression of the protein, the recombinant protein should be examined
from wild type and the PEP4 deficient strains after an induction performed
in parallel. Note: PEP4 deficient strains of P. pastoris are not as robust as
wild-type strains. In particular, PEP4 deficient strains die more quickly
when stored on plates, do not transform as efficiently as wild-type strains,
grow more slowly in culture, and are more difficult to induce on methanol.
Furthermore, PEP4 deficient strains are difficult to mate with other non-
PEP4 P. pastoris strains.
If a recombinant protein expressed in P. pastoris is larger than expected
and somewhat heterogenous in size by SDS–PAGE analysis, it may be
glycosylated. One should first examine the amino acid sequence of the
protein for potential glycosylation sites: the signal for addition of N-linked
oligosaccharides is ASN-X-Ser/Thr, and O-linked sugars can be added to
the OH-group of any Ser or Thr residue. Glycosylation can be confirmed by
deglycoslyating suspect protein with PNGase F and examining the product
by SDS–PAGE. A good protocol for deglycoslation of proteins can be found
at: http://www.neb.com/nebecomm/products/productP0704.asp. If this
treatment reduces the apparent molecular size of your protein resulting in a
more homogenous product, the protein is almost certainly glycosylated.

11. Selection for Multiple Copies

of an Expression Cassette
Perhaps the most productive means of increasing the per cell amount
of a recombinant protein using the P. pastoris system is by increasing the
number of copies of the expression cassette in a strain (Brierley, 1998; Thill
et al., 1990). Two general approaches have been developed to create multi-
copy expression strains in P. pastoris. The first approach involves construct-
ing a vector with multiple head-to-tail copies of an expression cassette
(Brierley, 1998). The key to generating this construction is a vector that
has an expression cassette flanked by restriction sites that have complemen-
tary termini (e.g., BamHI–BglII, SalI–XhoI combinations). The process of
repeated cleavage and reinsertion results in the generation of a series of
vectors that contain increasing numbers of expression cassettes. A particular
advantage to this approach, especially in the production of human
186 James M. Cregg et al.

pharmaceuticals, is that the precise number of expression cassettes is known

and can be recovered for direct verification by DNA sequencing.
The second approach utilizes expression vectors that contain a drug
resistance gene as the selectable marker, and selection for strains resistant
to higher levels of the drug (Scorer et al., 1994). Drug resistant genes used in
P. pastoris include the bacterial KanR, ZeoR, and BsdR genes, as well as, the
P. pastoris FLD1 gene (Shen et al., 1998). Each of these genes supports
significant enrichment for strains with increased copy number of the
expression vector with higher levels of drug resistance. For example, selec-
tion of zeocin resistant P. pastoris transformants is performed on plates with
1–2 mg/ml zeocin instead of the standard 100–200 ug/ml zeocin. How-
ever, no matter which drug is used, the vector copy number will still vary
greatly. After selection most transformants still contain only a single vector
copy, even if they are resistant to high drug levels. Thus, 50–100 indepen-
dent transformants selected at the high drug concentration should be ana-
lyzed for copy number and expression level to identify better expressing
strains. By this approach, strains carrying up to 30 copies of an expression
cassette have been isolated (Scorer et al., 1994). Importantly, once isolated,
multicopy strains are stable with standard microbial handling procedures
and do not require continued drug selection on plates or in liquid medium
(i.e., maintain a stock of a given strain selected on medium with the drug
stored frozen at 80
C, then use a working stock kept on plate of
noninducing medium for a limited number of experiments or a single
production run).
One drawback of this selection procedure is the difficulty in obtaining
enough clones to screen for multiple expression cassettes. First, the number
of transformants resistant to high levels of drug is very low, often 0.1–1% of
the number on low levels of drug (100 ug/ml for zeocin). Therefore, unless
ones transformation efficiency is at its peak, there may not be any transfor-
mants resistant to the highest drug levels. Furthermore, 50–100 transfor-
mants are needed to screen for a multicopy strain since only about 1–5% of
the transformants resistant to high levels of the drug are due to added copies
of the resistance gene and most are resistant due to other unknown factors.
In part due to the difficulty in obtaining multicopy strains by direct
selection a new method for obtaining strains with high copy numbers of
vector and elevated recombinant protein expression levels was developed
(Sunga et al., 2008). Briefly, P. pastoris transformants selected on a low level
of drug and containing only one or a few copies of the vector are subse-
quently subjected to higher drug levels to obtain strains with higher num-
bers of copies. Simply streak transformants on agar plates containing higher
and higher levels of zeocin. For example, if the original transformant was
selected on 100 ug/ml of zeocin, streak the strain on plates containing
500 ug/ml of the drug. Collect colonies that are resistant to the higher
level of drug as individual strains and confirm that one or more have
Expression in the Yeast Pichia pastoris 187

elevated recombinant protein expression levels and are resistant due to a

higher number of vector copies by PCR. Once the strain is confirmed as a
better expressing/multicopy strain, the process can be repeated at an even
higher concentration of zeocin (2 mg/ml). Again, high resistance strains are
collected and examined for expression and the number of vector copies.
This iterative process has been termed Posttransformational vector amplifi-
cation (PTVA) and results in strains containing multiple head-to-tail copies
of the entire vector integrated at a single locus in the genome. An analysis of
PTVA-selected clones indicates 40% showed a three- to fivefold increase
in vector copy number. So-called ‘‘jackpot’’ clones, with greater than
10 copies of the expression vector, represented 5–6% of selected clones
and in some cases had a proportional increase in recombinant protein
production.
Although the molecular details of the process(s) by which these amplifi-
cation events occur are not well understood, key observations about the
process have been made. First, the amplification process appears to occur
naturally in a small percentage of cells in virtually any vector containing
strain. Second, the PTVA process leads to a considerable increase in copy
number of the entire vector and not just portions of the vector such as the
resistance gene. This is clearly important if a uniform recombinant product
is desired. Third, Southern blot data demonstrated that all the copies are
inserted into the P. pastoris genome in the same location as the original copy
and in a head-to-tail configuration (Sunga et al., 2008).
Finally, the PTVA method works with other drug resistant selectable
markers in P. pastoris and not just with zeocin vectors. Thus, this amplification
process seems to be a general response to high drug levels in this yeast
species. Given that most yeast species have similar homologous recombination
systems, this technique should work in other yeast species as well.

REFERENCES
Ausubel, E. M., Brent, R., Kingston, E., Moore, D. D., Seidman, J. G., Smith, J. A., and
Struhl, K. (eds.) (2001). In ‘‘Current Protocols in Molecular Biology’’ John Wiley and
Sons, Inc., New York.
Becker, D. M., and Guarante, L. (1991). High-efficiency transformation of yeast by electro-
poration. Methods Enzymol. 194, 182–187.
Boer, E., Gellisson, G., and Kunze, G. (2005). Arxula adeninivorans. In ‘‘Production of
Recombinant Proteins’’ (G. Gellisson, ed.), pp. 89–110. Wiley-VCH Verlag GmbH and
Co. KGaA, Weinheim, Germany. Chapter 5.
Brierley, R. A. (1998). Secretion of recombinant human insulin-like growth factor I (IGF-I).
In ‘‘Methods in Molecular Biology Vol. 103 Pichia Protocols’’ (D. R. Higgins and
J. M. Cregg, eds.), pp. 149–178. Humana Press, Totowa, NJ.
Cregg, J. M. (2008). DNA-mediated transformation. In ‘‘Methods in Molecular Biology:
Pichia Protocols’’ ( J. M. Cregg ed.), 2nd ed. pp. 27–42. Humana Press, Totowa, NJ.
Chapter 3.
188 James M. Cregg et al.

Gellissen, G. (ed.) (2002). In ‘‘Hansenula polymorpha: Biology and Applications’’ Wiley-VCH

Verlag GmbH and Co. KGaA, Weinheim, Germany.
Gellissen, G. (ed.) (2005). In ‘‘Production of Recombinant Proteins’’ Wiley-VCH Verlag
GmbH and Co. KGaA, Weinheim, Germany.
Gleeson, M. A. G., White, C. E., Meininger, D. P., and Komives, E. A. (1998). Generation
of protease-deficient strains and their use in heterologous protein expression.
In ‘‘Methods in Molecular Biology, Vol. 103, Pichia Protocols’’ (D. Higgins and
J. M. Cregg, eds.), pp. 81–94. Humana Press, Totowa, NJ.
Grey, M., and Brendel, M. F. (1995). Ten-minute eletrotransformation of Saccharomyces
cerevisiae. In ‘‘Methods in Molecular Biology, Vol. 47: Electroporation Protocols for
Microorganisms’’ ( J. A. Nickoloff, ed.), pp. 269–272. Humana Press, Totowa, NJ.
Heerikhuisen, M., van den Hondel, C., and Punt, P. (2005). Aspergillus sojae.
In ‘‘Production of Recombinant Proteins’’ (G. Gellissen, ed.), pp. 191–214. Wiley-
VCH Verlag GmbH and Co. KGaA, Weinheim, Germany. Chapter 9.
Higgins, D. R., and Cregg, J. M. (eds.) (1998). Methods in Molecular Biology, Vol. 103,
Pichia Protocols, Humana Press, Totowa, NJ.
Kuck, U., and Poggeler, S. (2005). Sordaria macrospora. In ‘‘Production of Recombinant
Proteins’’ (G. Gellissen, ed.), pp. 215–232. Wiley-VCH Verlag GmbH and Co. KGaA,
Weinheim, Germany. Chapter 10.
Lin-Cereghino, J., and Lin Cereghino, G. P. (2008). Vectors and strains for expression.
In ‘‘Methods in Molecular Biology: Pichia Protocols’’ ( J. M. Cregg, ed.), pp. 111–126.
Humana Press, Totowa, NJ. Chapter 2.
Lin-Cereghino, J., Wong, W. W., Xiong, S., Giang, W., Luong, L. T., Vu, J.,
Johnson, S. D., and Lin-Cereghino, G. P. (2005). Condensed protocol for competent
cell preparation and transformation of the methylotrophic yeast Pichia pastoris. Biotechni-
ques 38, 44–48.
Lorow-Murray, D., and Jesse, J. (1991). High efficiency transformation of Saccharomyces
cerevisiae by electroporation. Focus 13, 65–68.
Madzak, C., Nicaud, J.-M., and Gaillardin, C. (2005). Yarrowia lipolytica. In ‘‘Production
of Recombinant Proteins’’ (G. Gellissen, ed.), pp. 163–190. Wiley-VCH Verlag GmbH
and Co. KGaA, Weinheim, Germany. Chapter 8.
Pichia Expression Kit Instruction Manual p. 63. Version E, Invitrogen, San Diego, CA.
Romanos, M. A., Scorer, C. A., and Clare, J. J. (1992). Foreign gene expression in yeast:
A review. Yeast 8, 423–488.
Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory
Manual. 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Scorer, C. A., Clare, J. J., McCombie, W. R., Romanos, M. A., and Sreekrishna, K. (1994).
Rapid selection using G4118 of high copy number transformants in Pichia pastoris for
high-level foreign gene expression. Bio/Technology 12, 181–184.
Shen, S., Sulter, G., Jeffries, T. W., and Cregg, J. M. (1998). A strong regulated promoter for
controlled expression of foreign genes in the yeast Pichia pastoris. Gene 216, 93–102.
Stowers, L., Gautsch, J., Dana, R., and Hoekstra, M. F. (1995). Yeast transformation and
preparation of frozen spheroplasts for electroporation. In ‘‘Methods in Molecular
Biology, Vol 47: Electroporation Protocols for Microorganisms’’ ( J. A. Nickoloff, ed.),
pp. 261–267. Humana Press, Totowa, NJ.
Sunga, A. J., Tolstorukov, I., and Cregg, J. M. (2008). Post transformational vector amplifi-
cation in the yeast Pichia pastoris. FEMS Yeast Res. 8, 870–876.
Thill, G. P., Davis, G. R., Stillman, C., Holtz, G., Brierley, R., Engel, M., Buckholtz, R.,
Kinney, J., Provow, S., Vedvick, T., and Siegel, R. S. (1990). Positive and negative
effects of multi-copy integrated expression vectors on protein expression in Pichia pastoris.
In ‘‘Proceedings of the 6th International Symposium on Genetics of Microorganisms’’
Expression in the Yeast Pichia pastoris 189

(H. Heslot, J. Davies, J. Florent, L. Bobichon, G. Durand, and L. Penasse, eds.) Vol. II,
pp. 477–490. Societe Francaise de Microbiologie, Paris.
Tolstorukov, I., and Cregg, J. M. (2008). Classical Genetics. In ‘‘Methods in Molecular
Biology: Pichia Protocols’’ ( J. M. Cregg, ed.), 2nd ed. pp. 189–202. Humana Press,
Totowa, NJ. Chapter 14.
van Ooyen, A. J., Dekker, P., Huang, M., Olsthoorn, M. M., Jacobs, D. I., Colussi, P. A.,
and Taron, C. H. (2006). Heterologous protein production in the yeast Kluyveromyces
lactis. FEMS Yeast Res. 6, 381–392.