High-level expression of apple PGIP1 is not sufficient to protect transgenic potato against Verticillium dahliae

Physiological and Molecular Plant Pathology 65 (2004) 145–155
www.elsevier.com/locate/pmpp
High-level expression of apple PGIP1 is not suf?cient to protect
transgenic potato against Verticillium dahliae
I. Gazendama,b, D. Oelofsea, D.K. Bergerb,*
aBiotechnology Division, ARC-Roodeplaat Vegetable and Ornamental Plant Institute, Private Bag X293, Pretoria 0001, South Africa
bDepartment of Botany, Forestry and Agricultural Biotechnology Institute (FABI), University of Pretoria, Pretoria 0002, South Africa
Accepted 7 January 2005
Abstract
Polygalacturonase-inhibiting proteins (PGIPs) are plant proteins believed to play a role in the defence against pathogenic fungi. Puri?ed
apple PGIP1 inhibited polygalacturonases (PGs) secreted by Verticillium dahliae grown on potato root cell walls and pectin. We therefore
hypothesised that apple PGIP1 could be used to confer resistance against Verticillium-wilt, a major disease of potato caused by the fungus V.
dahliae. Transgenic lines containing the apple pgip1 gene under control of the enhanced CaMV 35S (e35S) promoter were generated. Stable
integration of the apple pgip1 transgene into the potato genome was shown by the polymerase chain reaction (PCR) and Southern blot. Cross
hybridisation with potato pgip(s) was not observed. High-level expression of the apple PGIP1 in several independent transgenic potato events
was veri?ed by silver staining of SDS-PAGE separated proteins and Western blot. All but one of the PGIP1 extracts prepared from the
transgenic potato lines were successful in inhibiting V. dahliae PGs. Active PGIP1 was expressed in the leaves as well as the roots of the
transgenic plants. The apoplastic localisation of PGIP activity in the pgip-transgenic potato plants was demonstrated by a vacuum
in?ltration–extraction experiment. A glasshouse trial indicated that six transgenic lines (A10, B10, B13, A3, A14 and B16) had signi?cantly
reduced disease symptoms compared to the untransformed control and other lines when grown in the inoculated soil, but ?ve of them also
showed signi?cantly slower senescence symptoms when grown in the control soil. It is proposed that an extended juvenile phase in the
transgenic lines resulted in the apparent increased disease resistance, and this could not be attributed to inhibition of V. dahliae PGs, despite
high-level expression of the apple PGIP1.
q 2005 Elsevier Ltd. All rights reserved.
Keywords: Verticillium dahliae; Solanum tuberosum; Polygalacturonase; PG; Polygalacturonase-inhibiting protein; PGIP; Apple PGIP1; Transgenic potato;
Glasshouse trial
1. Introduction
symptoms [15]. The disease is characterised by chlorosis
and wilting of lower leaves, followed by browning and
Verticillium-wilt is a major disease that limits potato
drying-out, after which the symptoms spread to the rest of
(Solanum tuberosum L.) yield under arid growing con-
the stem or even the whole plant. The yield of potatoes is
ditions [9]. It is caused by the destructive soil-borne fungal
reduced because the growing period is shortened, resulting
pathogen, Verticillium dahliae. V. dahliae causes vascular
in a reduction in size of the daughter tubers [21]. Current
wilt diseases on more than 160 plant species, which includes
control approaches include sanitation of ?elds and equip-
cotton, tomatoes and potatoes. The fungus penetrates the
ment, crop rotation, soil fumigation and chemical control
host through its roots, spreads systemically through the
[34]. Since current control strategies are not effective in
xylem, and subsequently leads to the appearance of wilt
managing the disease, and chemical treatments may be
deleterious to the environment, alternative strategies are
required. Breeding for resistant potato lines faces many
* Corresponding author. Address: Room 6-26, Agricultural Sciences
challenges, therefore genetic modi?cation to confer resist-
Building, University of Pretoria, Pretoria 0002, South Africa. Tel.: C27 12
ance against this disease is proposed.
420 4634; fax: C27 12 420 3947.
E-mail address: dave.berger@up.ac.za (D.K. Berger).
To be successful in infection, plant and insect pathogens
must breach their host’s outer integument [31]. The plant
0885-5765/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
cell wall has important functions in maintaining the cell
doi:10.1016/j.pmpp.2005.01.002

---

146
I. Gazendam et al. / Physiological and Molecular Plant Pathology 65 (2004) 145–155
and tissue integrity. It also plays a complex role in resistance
The elicitor active molecules are accumulated at the site
to invading pathogens. It acts as a physical barrier to
of infection. The PGIP can therefore act against fungal
infecting pathogens [16], and is composed of complex
invasion of the plant by causing fungal PGs to increase their
polysaccharides capable of regulating gene expression and
elicitation of plant defence responses [6].
defence responses. Also localised in the plant cell wall are
PGIP has been shown previously to be successful in
enzymes and proteins involved in host defence mechanisms.
conferring fungal resistance to a heterologous plant. Pear
Pectin is a complex saccharide that contains large amounts
PGIP was able to confer resistance to tomato in vivo against
of galactosyluronic acid residues [39], and can be broken
the fungal pathogen Botrytis cinerea, whereas endogenous
down by a range of enzymes, such as endo- and exo-
tomato PGIP was not effective [25].
polygalacturonases, pectate lyases, pectin lyases and pectin
cDNA library studies showed that the mRNAs of cell
methylesterases.
wall degrading enzymes (CWDEs) of V. dahliae are
Fungal endopolygalacturonases (endo-PGs) play an
induced by growing the fungus in the presence of cellulose
important role during the early stages of plant pathogenesis,
[24]. It is proposed that V. dahliae produces a variety of
and are often the ?rst detectable enzymes secreted by plant
CWDEs during colonisation of the vascular system to obtain
fungal pathogens [13]. They must work ?rst before other
nutrients by limited degradation of the plant cell wall. An
enzymes, such as glycosidases, cellulases, hemicellulases
extracellular PG from V. dahliae isolated from cotton, and
and pectinases (pectin hydrolase, lyase and esterase), can
its inhibition by a PGIP present in the same host, has been
degrade their substrates, which are other plant cell wall
studied previously [15]. In this study, puri?ed apple PGIP1
components [16]. During pathogenesis, PGs spread into the
inhibited PGs secreted by V. dahliae grown on pectin
host tissue in advance of the invading fungal mycelium, and
medium. Since it was established from the literature that
hydrolyse the pectic components in primary plant cell walls
pectinases are important for V. dahliae during infection of
and middle lamellas [38]. This causes the cells to separate
its plant hosts [3,31], we therefore hypothesised that apple
and the host tissue macerates, facilitating pathogen pen-
PGIP1 could be used to confer resistance against V. dahliae.
etration and colonisation of the plant tissues. Extensive
The aim of this study was the molecular characterisation of
degradation of the plant cell walls ultimately leads to the
transgenic potato lines containing the apple pgip1 gene, and
death of the host cell. The products of the degradation process
to assess whether it confers enhanced fungal resistance to V.
are used by the fungus as a nutrient source for growth [16].
dahliae in the glasshouse.
The species within the Verticillium genus represent
diverse econutritional groups, including saprophytes and
pathogens of plants, insects, humans, nematodes and
2. Materials and methods
mushrooms [3,31]. The plant pathogens (V. dahliae,
Verticillium albo-atrum and Verticillium nigrescens) form
2.1. Potato transformation
a clade [3]. They all produce high levels of pectinase
enzymes able to degrade plant cell walls, while the insect
2.1.1. Construction of transformation plasmid
(Verticillium indicum, Verticillium lecanii) and mushroom
The apple pgip1 gene was isolated using inverse PCR
(Verticillium fungicola) pathogens cannot degrade pectin.
methods as described by Arendse et al. [1]. The sequence
These, on the other hand, produce high levels of a subtilisin-
obtained was identical to that published by Yao et al. ([36],
like protease, capable of degrading insect cuticles. Together
accession number U77041). An e35S-pgip1 construct for
with the nematode pathogens, the insect and mushroom
the constitutive expression of the apple pgip1 in transgenic
pathogens are distinguishable from the plant pathogens by
plants had been generated (pAppRTL2-NcoI-applePGIP;
their ability to produce chitinases. Strains of Verticillium
Oelofse, Arendse, Dubery and Berger, manuscript in
therefore show enzymic adaptation to the polymers present
preparation). The apple pgip1 gene was inserted into the
in the integument of their particular host, being either plant
binary vector pCAMBIA2300 for plant transformation,
or insect depending on their ecological niche [31].
using standard molecular biology methods, as part of a
Polygalacturonase-inhibiting proteins (PGIPs) are basic
cassette containing the enhanced CaMV 35S promoter
proteins present in the cell wall of most dicotyledonous
(e35S), the Tobacco Etch Virus (TEV) leader element,
plants. PGIP is a speci?c, reversible, saturable, high-af?nity
followed by the CaMV 35S terminator. Transformation
‘receptor’ for fungal, but not plant, endo-PGs [6–8]. It
constructs containing the apple pgip1 cassette in either
reduces the activity of endo-PGs, to different extents and is
orientation were generated, and called pCAMBIA2300-
highly speci?c. PGIP is structurally related to several
applepgip1A and pCAMBIA2300-applepgip1B.
resistance gene products, since it belongs to the super-
family of leucine-rich repeat (LRR) proteins [20]. In vitro
2.1.2. Agrobacterium-mediated potato transformation
evidence led to the hypothesis that the complex formed
Potato (Solanum tuberosum L.) cultivar BP1 were
between the endo-PG and PGIP leads to longer half-lives of
transformed with the pCAMBIA2300-applepgip1A and
oligogalacturonides that have elicitor activity, due to
pCAMBIA2300-applepgip1B constructs using standard
the reduced catalytic rate of the endo-PGs [8,11].
Agrobacterium tumefaciens-mediated transformation [23].

---

I. Gazendam et al. / Physiological and Molecular Plant Pathology 65 (2004) 145–155
147
Transgenic plants were selected on kanamycin and main-
CA, USA) using BSA as protein standard [5]. Two
tained in tissue culture. From the transformation with both
micrograms of undialysed in vitro root PGIP extract from
constructs, 29 kanamycin-resistant lines were produced, and
lines A6, A9, A10, A11, B10, B11, B16 and untransformed
the lines were labelled A or B depending on the construct
BP1 potato were separated on a 4% stacking/10% separating
transformed.
SDS-PAGE gel at 100 V for 30 min and 180 V for 1 h at 4 8C
[17]. Rainbow marker RPN 756 High range (Amersham-
2.2. Molecular analysis of the transgenic plants
Pharmacia Biotec) was used as protein molecular weight
marker. Pure apple PGIP1 (250 ng) was loaded as a positive
2.2.1. PCR
control. After electrophoresis, the gel was stained with silver
Genomic DNA was isolated on a small scale from
nitrate [4].
glasshouse-grown potato leaves using the method from
Murray and Thompson [22]. The primers used for detection
2.2.4. Western blot
of the apple pgip1 gene were: AP-PGIP-L2: 50-GCAGC-
The same PGIP extracts (0.5 mg) were separated by SDS-
CATGGAACTCAAGTTCTC-30 and AP-PGIP-R: 5 0-
PAGE as described before. The gel was electroblotted to
CCCGGATCCATCTGCAGTTGTGGCCATTAC-30. For
PVDF-Plus transfer membrane (MSI, Westborough, Ma,
ampli?cation of the nptII gene (conferring kanamycin
USA) at 30 V, 4 8C overnight. ECL biotinylated protein
resistance):
NPTII-L:
50-GAGGCTATTCGGCTAT-
molecular marker (ECL RPN2107, Amersham-Pharmacia
GACTG-30 and NPTII-R: 50-ATCGGGAGCGGCGA-
Biotec) was used instead of the Rainbow marker. The
TACCGTA-30. One unit of Taq DNA Polymerase
membrane was blocked and incubated with the anti-bean
(Promega, Madison, WI, USA.) was used in a reaction
PGIP antibody (prepared against puri?ed bean PGIP
containing the supplied buffer and 50–120 ng genomic
prepared from bean pods by D. Oelofse (unpublished)) at
DNA as template. An annealing temperature of 58 8C was
room temperature overnight on a platform rocker (Bibby
used for the pgip1 primers and 62 8C for the nptII primers.
Sterilin Limited, Staffordshire, UK). The membrane was
PCR products were separated and visualised by agarose gel
incubated with the secondary anti-rabbit IgG-horseradish
electrophoresis.
peroxidase conjugate (HRP) and streptavidin-HRP conju-
gate and detected using the ECL Western Blotting Analysis
2.2.2. Southern hybridisation
system (Amersham-Pharmacia Biotec). The membrane was
Genomic DNA was isolated on a large scale from
exposed to X-ray ?lm (Hyper?lm ECL High performance
glasshouse-grown potato leaf material using the method of
chemiluminescence ?lm, Amersham-Pharmacia Biotec) for
Dellaporta et al. [12]. The pAppRTL2-NcoI-applePGIP
1 min and the ?lm developed.
vector containing apple pgip1 was digested with PstI and
used to spike untransformed BP1 potato genomic DNA for
2.3. Enzyme activity assays
determination of the transgene copy number in the
transgenic lines during Southern blot. A labelled DNA
2.3.1. Potato root cell wall preparation
probe was generated using the PCR DIG Probe synthesis kit
Approximately 200 g (fresh weight) BP1 potato roots
(Roche Diagnostics GmbH, Germany) and the AP-PGIP
were harvested from plants grown in vermiculite in a
primers. Ten micrograms of genomic DNA, extracted from
glasshouse. The roots were washed and freed from
lines A5, A6, A9, B10, B11 and B13 and digested with
vermiculite particles and dried before freezing in liquid
NcoI, NsiI or BamHI, were separated on an agarose gel and
nitrogen and storing at K70 8C. The frozen roots were
blotted to a nylon membrane (Osmonics Magnacharge
ground to a powder in liquid nitrogen and a cell wall extract
nylon transfer membrane, Amersham-Pharmacia Biotec,
was prepared following the protocol of English et al. [13].
Little Chalfont, UK) using standard protocols [30]. The
The ground roots were blended in a Waring blender and
membrane was probed with the DIG-labelled apple pgip1
washed with cold 0.1 M potassium phosphate buffer (pH
probe at a hybridisation temperature of 42 8C. The
7.0). Thereafter it was repeatedly extracted with chloro-
denatured probe was added to DIG Easy Hyb solution
form: methanol (1:1) and acetone, and the residue air-dried.
containing 125 mg/ml denatured salmon testes DNA. The
A yield of dry cell walls that was 10% of the original fresh
membrane was blocked and DIG detected using the DIG
weight was obtained.
Wash and Block Buffer set and the DIG Luminescent
detection kit for nucleic acids (Roche Diagnostics GmbH).
2.3.2. V. dahliae PG preparation on pectin and potato root
The membrane was exposed to X-ray ?lm (Hyper?lm ECL
cell walls
High performance chemiluminescence ?lm, Amersham-
V. dahliae was isolated from infected potato of the
Pharmacia Biotec) for 1 h and the ?lm developed.
cultivar Lady Rosetta collected in 1998 from the Worcester
area (South Africa). The culture was stored as isolate 61 in
2.2.3. SDS-PAGE and silver staining of proteins
the culture collection of C. Millard at ARC-Roodeplaat. The
Protein concentrations of the crude PGIP extracts were
fungus was grown in a citrate: phosphate buffer (pH 6.0)
determined with the Bio-Rad protein assay kit (Hercules,
containing the following ?nal concentrations: 100 mg mlK1

---

148
I. Gazendam et al. / Physiological and Molecular Plant Pathology 65 (2004) 145–155
ampicillin; 2 mM MgSO4; 0.6 mM MnSO4; 25 mM KNO3;
weeks to support their vertical growth. Ten weeks after
3.5 mM ZnSO4; 0.6 mM CuSO4 and 3.6 mM FeSO4. The
planting, leaves were harvested, vacuum-in?ltrated and
carbon source was either 1% pectin (w/v) or 1% potato root
extracted with 20 mM NaAc buffer (pH 5.0) and 0.3 M
cell walls. PG extracts were prepared as described in Berger
NaCl according to Salvi et al. [27] and Terry and Bonner
et al. [2], except the culture was grown at 27 8C, the culture
[33]. Centrifugation to collect apoplastic ?uid took place at
?ltrate was treated with 85% ammonium sulphate to collect
1000g for 10–20 min in a Beckman TJ-6R tabletop
the PG proteins and the pellet was resuspended in one-
centrifuge ?tted with a TH-4 rotor with buckets. For
twentieth the original volume 20 mM NaAc buffer, (pH
comparison, total PGIP extracts were prepared from potato
4.7).
leaves as described above.
2.3.3. Native polyacrylamide gel electrophoresis (PAGE) of
2.3.6. Agarose diffusion assay (ADA)
V. dahliae PGs
Petri dishes containing the assay medium [1% type II
PG extracts were dialysed against 20 mM ammonium
agarose, 0.01% Polygalacturonic acid (PGA) and 0.5%
acetate, pH 4.7, at 4 8C in a 12 kDa molecular weight cut-off
ammonium oxalate in citrate-phosphate buffer, pH 4.6]
dialysis membrane (Sigma GmbH, Germany). Two micro-
were prepared according to Taylor and Secor [32] with
gram quantities were freeze-dried in 1.5 ml tubes and stored
modi?cations. Equal volumes of V. dahliae PG (diluted 1 in
at K70 8C. These aliquots were dissolved in protein loading
5, speci?c activity of 9 nmol reducing ends/min/mg protein)
buffer and electrophoresed on a native PAGE gel, consisting
and 20 mM NaAc buffer (pH 4.7) or various PGIP extracts
of a 15% separating and a 4% stacking gel. It was
were incubated at 25 8C for 20 min, after which 25 ml was
electrophoresed at 5 8C for 2 h 50 min at 100 V [17].
loaded into a well of an ADA plate and incubated at 27 8C
Rainbow marker RPN 756 High range (Amersham-
overnight. Puri?ed apple PGIP1 was used as a positive
Pharmacia Biotec) was used as protein molecular weight
control. All reactions were done in triplicate. The fungal PG
marker. After electrophoresis, one-half of the gel was
activity was visualised by staining with 0.05% ruthenium
stained with silver nitrate [4] and the other half processed
red (Sigma GmbH) for 1 h at 37 8C. The plates were
for PG activity. The gel was placed in 20 mM NaAc (pH
destained overnight at 4 8C with dH2O before the zone
4.8) for 10 min, and then overlaid with an overlay made up
diameters were measured. A zone radius was calculated
of 0.1% polygalacturonic acid (PGA) and 0.8% agarose in
from the zone diameter by subtracting the well size
100 mM NaAc buffer (pH 4.8), cast onto a Gelbond ?lm.
(5.5 mm) and dividing by two. The percentage inhibition
Similar PAGE gels were overlaid with agarose overlays
relative to NaAc buffer was calculated on the radius
incorporating 0.88 mg/ml puri?ed apple PGIP1 or 0.88 mg/
measurements. The nanogram PGIP extract used in each
ml boiled puri?ed apple PGIP1. After overnight incubation,
assay was plotted against its resulting percentage inhibition.
the overlay was stained for PG activity with 0.05%
The amount of PGIP extract (in ng) to inhibit V. dahliae PG
ruthenium red (Sigma GmbH) for 1 h 20 min at 37 8C.
activity by 50% was read from the resulting graph.
Statistical analysis was carried out using the software
2.3.4. Preparation of PGIP extracts and puri?ed
GenStat 5 (1993 release, GenStat 5 committee of the
apple PGIP1
Statistics Department, IACR-Rothamsted, UK, Clarendon
PGIP extracts were prepared from roots and leaves
Press: Oxford).
collected from either in vitro plantlets or glasshouse-grown
plants. Samples of ground plant material were extracted
2.3.7. Reducing sugar assay for PG activity
with two volumes of 1 M NaCl, 20 mM NaAc buffer (pH
Release of reducing sugars by fungal PG activity was
4.7) for 2 h at 4 8C. The supernatants were dialysed against
measured by the PAHBAH (p-hydroxybenzoic acid hydra-
20 mM NaAc buffer (pH 4.7) at 4 8C. A 12 kDa molecular
zide) procedure [18,39] as described in Berger et al. [2]. PG
weight cut-off dialysis membrane (Sigma GmbH) was used.
enzyme activity was expressed as mmoles reducing ends
Non-dialysed extracts were used for the agarose diffusion
released per min at 30 8C with 0.25% polygalacturonic acid
assay. Apple PGIP1 was puri?ed to apparent homogeneity
as substrate. The data enabled the selection of the 30 min
(see Fig. 3(a), lane 10) from the transgenic tobacco line LA
time point and a 1C4 dilution of V. dahliae PG (37 mg/ml
Burley: pgip1 #8, which had been transformed with
protein, speci?c activity of 9 nmol reducing ends/min/mg
pCAMBIA2300-applepgip1B (Oelofse, Arendse, Dubery
protein), which was in the linear range of activity, for PGIP
and Berger, manuscript in preparation).
inhibition studies. The PGIP inhibition studies were carried
out as described in Berger et al. [2]. Statistical analysis was
2.3.5. Apoplastic PGIP extraction
done using GenStat.
Seven-week-old in vitro potato plantlets of untrans-
formed BP1 and the apple pgip1 transgenic lines A6 and A9
2.4. Glasshouse trial
were hardened off and grown in a glasshouse for apoplastic
PGIP extractions. The pots were placed in a glasshouse at
In vitro transgenic plantlets were hardened off and grown
25 8C, watered three times daily and tied up to stakes after 5
in a glasshouse to produce mini-tubers. The mini-tubers were

---

I. Gazendam et al. / Physiological and Molecular Plant Pathology 65 (2004) 145–155
149
planted into pots containing sterile sandy soil and vermicu-
Table 1
lite (3:1, v/v). Nine replicates of the twenty transgenic lines
Apple PGIP1 inhibits PGs of Verticillium dahliae
and eighteen replicates of untransformed BP1 potato (two
V. dah-
Source of PGIP
Treatment
Mean PG
SDId
groups of nine each, called BP1A and BP1B) were planted in
liae PGa
of PGIPb
activity
a randomised block design [28]. They were planted in pots
(%)c
containing V. dahliae infected soil (62 microsclerotia gK1
C
NaAc buffer
–
100.0
b
soil, termed ‘inoculated’ soil) and uninoculated control soil.
C
Transgenic tobaccoe
–
11.1G4.7
a
The pots were placed in a glasshouse at 25 8C, watered three
C
Untransformed tobaccof
–
96.5G3.1
b
C
Transgenic tobaccoe
Boiled
101.3G1.3
b
times daily and tied up to stakes to support their vertical
C
Puri?ed PGIP1g
–
11.2G0.9
a
growth. Nine to 16 weeks after planting, visual assessments
C
Puri?ed PGIP1g
Boiled
102.1G5.1
b
of disease symptoms were performed twice weekly using a 5-
The PG activity was determined by the reducing sugar assay and is shown
point scale of Robinson et al. [26] and Isaac and Harrison
as the mean of three separate reactions. The PG:PGIP mixtures were
[14]. The stems were divided into three equal regions and
incubated for 20 min at 25 8C prior to addition of the substrate PGA, and
class values assigned to each plant according to the following
incubation for a further 30 min at 30 8C. The amount of reducing sugars
scale: 1, no symptoms of yellowing/ wilting; 2, single yellow
released was assessed using the PAHBAH method.
a
leaf or symptoms up to the bottom third of the plant; 3,
V. dahliae PG (speci?c activity of 9 nmol reducing ends/min/mg
protein) was mixed with the PGIP from the indicated sources.
symptoms up to the middle third; 4, symptoms up to the top
b Where indicated the extracts had been boiled for 10 min and cooled
third or the whole plant symptomatic; and 5, the whole plant
prior to mixing with the PG.
wilted, dried out and completely dead. At the end of the
c The PG activity is presented as a percentage of the activity obtained in
growth stage, stem sections were collected and plated onto
the presence of NaAC buffer (20 mM, pH 4.7).
d
PDA plates amended with 100 mg/ml streptomycin sulphate.
SDI, Signi?cant difference indicator. PGIP sources with different lower
case letters had signi?cantly different PG activity % from one another at the
The plates were incubated in a growth room at 25 8C at 12 h
1% con?dence level using Fisher’s protected least signi?cant difference test
light and 12 h darkness. The plates were microscopically
(F-test).
examined 3–5 days later to identify V. dahliae fungal
e Transgenic tobaccoZLA Burley: pgip1 #8 (0.21 mg crude PGIP extract).
cultures growing on the stems. Disease indices were
f Untransformed tobaccoZLA Burley (0.35 mg crude PGIP extract).
g
calculated for each replicate using a modi?cation of the
Puri?ed PGIP1ZApple PGIP1 protein puri?ed to homogeneity from
LA Burley: pgip1 #8 (0.188 mg pure PGIP1 protein).
index of Corsini et al. [10] (Eq. 1). It incorporated the visual
disease severity on week 16 after planting of the tubers.
Puri?ed PGIP1 and transgenic tobacco PGIP extract
Colonisation was a 1 if V. dahliae colonies could be
caused a signi?cant reduction in V. dahliae PG activity
identi?ed microscopically. The median time (in weeks) for
(from 100% to 11%) (Table 1). When these samples were
symptoms to appear was determined by arranging the weeks
boiled, activity returned to normal (101% and 102%,
when symptoms started to appear for each individual plant in
respectively). This indicated that the inhibitor is a protein
an increasing order, and choosing the middle value.
since it is heat denaturable.
Statistical analysis was done using GenStat.
ðWilt severity 1–5 scaleÞ !ðindividual showing wilt 0=1Þ Cðindividual stem colonised 0=1Þ
10
Disease index :
!
(1)
Median time for symptoms to appear
1
Further experiments were done to verify that apple
PGIP1 also inhibited PGs produced by V. dahliae in vivo.
When grown on potato root cell walls, the PGs produced
3. Results
were comparable to in vitro pectin media, as shown in a
native gel (Fig. 1, lanes 4 and 5). The activity of these PGs
was inhibited by apple PGIP1 (Fig. 1, lanes 6 and 7),
3.1. Apple PGIP1 is an active inhibitor of V. dahliae PGs
consistent with the results shown in Table 1. Inhibition was
removed by boiling of PGIP1, indicating it was due to active
Puri?ed apple PGIP1 was tested in an inhibition assay
protein (Fig. 1, lanes 8 and 9).
against V. dahliae PGs isolated from a culture grown on
pectin medium. The apple pgip1 gene was introduced into
tobacco cv. LA Burley to generate the transgenic line LA
3.2. Production of transgenic potatoes expressing the apple
Burley: pgip1 #8. PGIP extracts and puri?ed apple PGIP1
PGIP1 protein at high levels
were prepared from this plant, and tested together with
untransformed control extracts against V. dahliae PGs.
Because of the successful inhibition of V. dahliae PGs by
Table 1 summarises the results of the PG activity obtained
apple PGIP1 in vitro, it was proposed to transform the pgip1
in the presence of various PGIP preparations. The results are
gene into potato in an effort to protect it against
expressed as a percentage relative to the PG activity in the
Verticillium-wilt. Transgenic potato lines containing the
presence of NaAc buffer (20 mM, pH 4.7).
apple pgip1 gene were generated as described in Section 2.

---

150
I. Gazendam et al. / Physiological and Molecular Plant Pathology 65 (2004) 145–155
Fig. 1. Native polyacrylamide gel electrophoresis (PAGE) and polygalacturonic acid overlay assays of PGs secreted by Verticillium dahliae grown on potato
root cell walls and pectin medium. PAGE was conducted in a 15% separating polyacrylamide gel with a 4% stacking gel [17]. (A) Proteins were stained with
silver staining. Lane 1: Rainbow protein MW marker; Lane 2: 2 mg V. dahliae PG extract (potato root cell wall); Lane 3: 2 mg pectin V. dahliae PG extract.
Similar PAGE gels were overlaid with agarose overlays containing 0.1% polygalacturonic acid and incubated for 17 h before staining with ruthenium red. (B)
Standard overlay; (C) overlay incorporating 0.88 mg/ml puri?ed apple PGIP1; (D) overlay incorporating 0.88 mg/ml boiled puri?ed apple PGIP1. Lanes 4, 6
and 8: 10 mg V. dahliae PG extract (potato root cell wall); Lanes 5, 7 and 9: 1 mg pectin V. dahliae PG extract.
PCR with the apple pgip1 primers yielded the expected
Furthermore, both detection methods are consistent since
1024 bp product for all twenty lines selected for the
they showed the protein present in six transgenic lines and
glasshouse trial (results not shown). PCR with NPTII
absent in the untransformed BP1 and one transformed line
primers also showed the presence of the kanamycin
(B16) (Fig. 3(a) and (b)). The broadness of the bands in the
resistance gene in all the lines (results not shown).
Western blot compared to the SDS-PAGE (including the
Southern blot (Fig. 2) of six randomly selected lines (A5,
pure apple PGIP1; Fig. 3(b), lane 10 vs. Fig. 3(a), lane 10)
A6, A9, B10, B11, B13) indicated the integration of a single
can be attributed to diffusion during the blotting process.
copy (lines A5, A6, A9 and B11) and double copies of
the pgip1 transgene (lines B10 and B13) into the potato
3.3. Transgenic potato lines contain active PGIP
genome. NsiI digestion was used to determine the number of
in their roots
insertion events in both A and B lines. BamHI and NcoI were
used to determine the copy number in the A and B lines,
PGIP extracts were prepared from in vitro potato roots,
respectively. BamHI digestion was incomplete leading to
which are at the infection point of V. dahliae. The agarose
hybridising fragments that were larger than expected (Fig. 2,
lanes 6, 8 and 10, 1943 bp fragment was predicted), while NcoI
excised the expected 2356 bp hybridising fragment (Fig. 2,
lanes 12, 14 and 16). The probe did not hybridise to the
untransformed BP1 gDNA (Fig. 2, lane 17).
SDS-PAGE (Fig. 3(a)) of PGIP extracts from eight
selected lines, four of them representative of the lines shown
in the Southern blot, showed similar amounts of a highly
expressed protein in extracts A6, A9, A10, A11, B10 and
B11 (Fig. 3(a), lanes 3–8) but not in BP1 and B16 (Fig. 3(a),
lanes 2 and 9, respectively). The highly expressed protein’s
size was between 30 and 45 kDa. Silver staining of 250 ng
puri?ed apple PGIP1 (Fig. 3(a), lane 10) showed a protein in
the same size range.
A Western blot (Fig. 3(b)) of the same samples was
incubated with anti-bean PGIP antibody. Non-speci?c
Fig. 2. Southern blot of apple pgip1 transgenic potato lines A5, A6, A9,
B10, B11 and B13 probed with pgip1 gene. Lane 1: lDNA/HindIII marker
background signals were detected in all the potato lines
(DIG-labelled); lanes 2–4: 1, 10 and 20 copies of apple pgip1 spiked into
(Fig. 3(b), lanes 2–9), but a protein of the same size as pure
10 mg NsiI digested BP1 gDNA, respectively; lanes 5–16: 10 mg of the
apple PGIP1 (Fig. 3(b), lane 10) was detected in six
following transgenic potato lines gDNA digested with the indicated
transgenic lines while absent from the BP1 and B16 lines
restriction enzymes; lanes 5 and 6: A5 digested with NsiI and BamHI,
(Fig. 3(b), lanes 2 and 9, respectively). We conclude that
respectively; lanes 7 and 8: A6 digested with NsiI and BamHI, respectively;
lanes 9 and 10: A9 digested with NsiI and BamHI, respectively; lanes 11
this protein is the highly expressed apple PGIP1, due to the
and 12: B10 digested with NsiI and NcoI, respectively; lanes 13 and 14: B11
correlation in size to the pure apple PGIP1 on both the SDS-
digested with NsiI and NcoI, respectively; lanes 15 and 16: B13 digested
PAGE gel and the Western blot (Fig. 3(a) and (b)).
with NsiI and NcoI, respectively; lane 17: 10 mg NsiI digested BP1 gDNA.

---

I. Gazendam et al. / Physiological and Molecular Plant Pathology 65 (2004) 145–155
151
Table 2
Inhibition of V. dahliae PGa by PGIP extracts from apple pgip1 transgenic
potato roots
Source of PGIP extract
PGIP extract (ng)b to cause
SDIc
50% PG inhibition
BP1 untransformed
O1200
a
A6
37G2
b
A9
53G8
b
A10
47G1
b
A11
51G10
b
B10
58G22
b
B11
38G3
b
B16
O1200
a
Puri?ed apple PGIP1d
35G3
b
a The speci?c activity of V. dahliae PG was 9 nmol reducing ends/min/mg
protein.
b The amount of PGIP extract (in ng) to inhibit V. dahliae PG activity by
50% was determined in an agarose diffusion assay. It is shown as the mean
of three separate reactions.
c SDI, Signi?cant difference indicator. PGIP sources with different lower
case letters had signi?cantly different PGIP activities from one another at
the 1% con?dence level using Fisher’s protected least signi?cant difference
test (F-test).
d Apple PGIP1 protein puri?ed to homogeneity from transgenic tobacco
LA Burley: pgip1 #8.
Up to 1200 ng of extract from one transgenic line, B16, and
untransformed BP1, still did not give 50% inhibition.
Fig. 3. SDS-PAGE and Western blot of PGIP extracts prepared from roots
Extracts of leaves were assayed for PGIP, and all the
of apple pgip1 transgenic potato. (a) SDS-PAGE with silver staining of
lines showing root activity (A6, A9, A10, A11, B10, B11;
proteins. SDS-PAGE was conducted in a 10% separating SDS-polyacryl-
Table 2) also showed leaf activity (data not shown). Line
amide gel with a 4% stacking gel [17]. Lane 1: rainbow protein MW
B16 and untransformed BP1 showed statistically signi?cant
marker; lanes 2–9: 2 mg of the following potato lines PGIP extracts:
untransformed BP1, A6, A9, A10, A11, B10, B11 and B16, respectively;
less PGIP activity in leaves (data not shown). An additional
lane 10: 250 ng pure apple PGIP1. (b) Western blot with chemiluminescent
13 transgenic lines also showed PGIP activity (data not
detection of hybridised anti-PGIP antibody. SDS-PAGE was conducted in a
shown). Activity was abolished when the extracts were
10% separating SDS-polyacrylamide gel with a 4% stacking gel [17]. The
boiled, as expected (data not shown).
gel was electroblotted overnight to PVDF-Plus transfer membrane at 30 V.
Lane 1: ECL biotinylated protein MW marker; lanes 2–9: 0.5 mg of the
following potato lines PGIP extracts: untransformed BP1, A6, A9, A10,
3.4. Apple PGIP1 is localised to the apoplast
A11, B10, B11 and B16, respectively; lane 10: 100 ng pure apple PGIP1.
in transgenic potato
The apoplastic localisation of PGIP activity in the pgip1-
diffusion assay was performed with V. dahliae PG and a
transgenic potato plants was demonstrated by a vacuum
dilution series of PGIP extracts to quantify the root
in?ltration–extraction experiment. Lines A6 and A9 were
inhibition. The data was used to construct Table 2.
chosen as representatives since they were shown to be stably
Statistical analysis was carried out using the software
transformed with apple pgip1, they expressed the protein and
GenStat. The PGIP extracts from transgenic potato roots
exhibited PGIP activity (Figs. 2 and 3, Table 2). Apoplastic
contained an active PGIP since it inhibited the zones formed
extracts from both line A6 and line A9 showed PGIP activity in
by the V. dahliae PGs. The PG inhibition activity of the
contrast to apoplastic extracts from untransformed BP1 potato
PGIP extracts was shown to decrease with dilution. Table 2
(Table 3). The total extracts from A6 and A9 also showed
contains, for each chosen line, the nanograms of root PGIP
PGIP activity, as expected, although it appeared that higher
extract needed to inhibit the V. dahliae PGs by 50%.
activity was obtained from the apoplast (Table 3).
From the results it was concluded that there were high
levels of apple PGIP1 expression in the potato roots, since
3.5. Apple PGIP1 does not confer enhanced resistance
similar quantities of extract from six lines (37–58 ng)
in the glasshouse
produced 50% inhibition of V. dahliae PG, which is
comparable to the amount of puri?ed apple PGIP1 that
Analysis of Verticillium-wilt resistance or susceptibility
was required (35 ng). The expression levels of the six high-
of the transgenic potato lines was based on visual
expressing lines did not differ signi?cantly from each other.
assessments of the foliage symptoms typical for

---

152
I. Gazendam et al. / Physiological and Molecular Plant Pathology 65 (2004) 145–155
Table 3
difference test (F-test) was applied separately to the disease
Apple PGIP1 is localised to the apoplast in apple pgip1 transgenic potato
indices of plants grown in each soil type (inoculated or
Source of PGIP1
Line
PGIP extract (ng)a
SDIc
control). Data were analysed using the statistical program
extract
to cause 50% PGb
GenStat. Data were tested for statistically signi?cant
inhibition
differences between the disease indices of untransformed
Apoplastic extract
BP1 untransformed
O1500
a
BP1 and the transgenic potato lines. Table 4 summarises the
Apoplastic extract
A6
54G0
b
DI and statistical analysis data, with the DI of the inoculated
Apoplastic extract
A9
56G0
b
group arranged in an increasing order. The overall F-test was
Total extract
BP1 untransformed
O1500
a
Total extract
A6
145G12
c
signi?cant at the 1% level of signi?cance for both the inoculated
Total extract
A9
138G6
c
and control groups. The null hypothesis of no difference
Puri?ed apple
63G6
b
between lines was therefore rejected. The least signi?cant
PGIP1d
difference (lsd) of index values at the 1% level of signi?cance
a The amount of PGIP extract (in ng) to inhibit V. dahliae PG activity by
was determined to be 1.0676 for the lines planted in inoculated
50% was determined in an agarose diffusion assay. It is shown as the mean
soil. Lines planted in control soil had an lsd of 0.9386.
of three separate reactions.
b
Infection of the potato plants during the glasshouse trial
The speci?c activity of V. dahliae PG was 9 nmol reducing ends/min/mg
protein.
with