Expression of streptavidin in tomato resulted in abnormal plant development that could be restored by biotin application.

Abstract

Biotin is an essential cofactor for a variety of carboxylase and decarboxylase reactions and is involved in diverse metabolic pathways of all organisms. In the present study we tested the hypothesis that controlling biotin availability by the expression of Streptomyces avidinii streptavidin, would impede plant development. Transient expression of streptavidin fused to plant signal peptide, bacterial signal peptide or both, in tomato (Lycopersicon esculentum cv. VF36) plants resulted in various levels of tissue impairment, exhibited as lesion development on 1-week-old tomato seedlings. The least toxic construct was introduced to tomato (stable transformation) under the constitutive CaMV 35S promoter, and lesions appeared on stems, flower morphologies were modified and numbers and sizes of fruits were altered. Furthermore, tissue-specific expression of the streptavidin, by means of the beta-phaseolin or TobRB7 promoters, resulted in localised effects, i.e., impaired seed formation or seedless fruits, respectively, with no alteration in the morphology of the other plant organs. External application of biotin on streptavidin-expressing tomato plants prevented the degeneration symptoms and facilitated normal plant development. It can be concluded that expression of streptavidin in the plant cell can lead to local and temporal deficiencies in biotin availability, impairing developmental processes while biotin application restores plant growth cycle.

Full text

Page 1

J. Plant Physiol. 161. 611–620 (2004)
http://www.elsevier.de/jplhp
Expression of streptavidin in tomato resulted in abnormal plant
development that could be restored by biotin application
Idit Ginzberg1*,Avihai Perl2, Mira Genser1, SmadarWininger1, Chaya Nemas1,Yoram Kapulnik1
1Department of Field and Garden Crops,ARO, theVolcani Center, PO Box 6, Bet Dagan 50250, Israel
2Department of FruitTree Breeding and Molecular Genetics,ARO, theVolcani Center, PO Box 6, Bet Dagan 50250, Israel
Received January 15, 2003 · Accepted May 2, 2003
Summary
Biotin is an essential cofactor for a variety of carboxylase and decarboxylase reactions and is
involved in diverse metabolic pathways of all organisms. In the present study we tested the hypothe-
sis that controlling biotin availability by the expression of Streptomyces avidinii streptavidin, would
impede plant development. Transient expression of streptavidin fused to plant signal peptide, bacte-
rial signal peptide or both, in tomato (Lycopersicon esculentum cv. VF36) plants resulted in various
levels of tissue impairment, exhibited as lesion development on 1-week-old tomato seedlings. The
least toxic construct was introduced to tomato (stable transformation) under the constitutive CaMV
35S promoter, and lesions appeared on stems, flower morphologies were modified and numbers and
sizes of fruits were altered. Furthermore, tissue-specific expression of the streptavidin, by means of
the β-phaseolin or TobRB7 promoters, resulted in localised effects, i.e., impaired seed formation or
seedless fruits, respectively, with no alteration in the morphology of the other plant organs. External
application of biotin on streptavidin-expressing tomato plants prevented the degeneration symptoms
and facilitated normal plant development. It can be concluded that expression of streptavidin in the
plant cell can lead to local and temporal deficiencies in biotin availability, impairing developmental
processes while biotin application restores plant growth cycle.
Key words: biotin – streptavidin – transgenic tomato
Abbreviations: mst = the entire bacterial streptavidin gene. – mprost = the bacterial streptavidin
gene without its signal peptide sequence. – mcyto = core streptavidin. – sps, prost, cst = the corre-
sponding mst, mprost and mcyto streptavidin sequences fused to wheat α/β-gliadin signal peptide. –
CaMV = cauliflower mosaic virus. – PCR = polymerase chain reaction. – rDNA = ribosomal DNA. –
rRNA = ribosomal RNA
* E-mail corresponding author: iditgin@volcani.agri.gov.il
Contribution from the Agricultural Research Organisation, the Volcani
Center, Bet Dagan, Israel, No.102/2002.
0176-1617/04/161/05-611$ 30.00/0

Page 2

612Idit Ginzberg et al.
Introduction
To control plant development by temporarily suppressing a
specific growth process, while retaining the ability to restore
normal growth, is a desirable treatment in many plant physio-
logical studies. Only a few methods, to date, to impede plant
cell development have been demonstrated, e.g., expression
of diphtheria toxin in plants (Day et al. 1995) and introduction
of the Barnase gene system (Goldberg et al. 1995). Cell abla-
tion by laser technology has been demonstrated to damage
specific cells in developing meristems (Berger 1998). One
common feature of all the above-mentioned strategies is that
the impairment of the targeted plant organ could not be
reversed at the same stage of growth. The ability to arrest the
development of a specific plant organ (resulting in seedless
fruit, delayed flowering, etc.) and then to restore normal
growth at will is of considerable economic importance in the
cultivation of many ornamental and field crop plants. In the
approach detailed in the present paper, a specific ligand that
has the capacity to bind and hold on to a specific molecule,
essential for plant organogenesis, has been used to impair
the development of transgenic plants. Reduction in the avail-
ability of the captured molecule would interrupt essential cel-
lular processes, and so induce tissue ablation. The advan-
tage of such a strategy lies in the fact that reversal of this ab-
lation effect could be achieved by replenishment of the es-
sential molecule in the affected cells. In the present study we
tested this possibility by using streptavidin as a ligand for bio-
tin, to control plant development in transgenic tomato plants.
Biotin is an essential cofactor for a variety of carboxylases
and decarboxylases that are found in diverse metabolic path-
ways of all organisms (Knowles 1989). For example, it is in-
volved in membrane biogenesis (Yanai et al. 1995), degrada-
tion pathways of several amino acids (Baldet et al. 1992), and
formation of oxaloacetate from pyruvate (Wurtele and Nikolau
1990). Differential biotinylation of biotin-containing proteins
and compartmentalization of these biotinylated enzymes
have been hypothesized to play a role in gene expression
and cellular regulation (Nikolau et al. 1984, 1993, Ohlrogge et
al. 1979, Wang et al. 1995). Correspondingly it was shown that
deficiencies in this essential vitamin impaired plant develop-
ment (Patton et al. 1998, Shellhammer and Meinke 1990).
Streptavidin is a non-glycosylated, neutral bacterial protein
from Streptomyces avidinii (Argarana et al. 1986) that, sim-
ilarly to the egg-white avidin (Gope et al. 1987) binds biotin ir-
reversibly (Ka 1015mol/L–1) (Green 1975). In the bacterial cell
streptavidin is first synthesised as a pro-streptavidin with a
molecular mass of 18. Post-secretory proteolytic digestion
processes yield 14-kDa core-streptavidin that forms tetramers
(Bayer et al. 1989). Very little information is available in re-
gards to streptavidin expression in plants. As no biotin is pre-
sent in the vacuoles, over expression of the core-streptavidin
fused to a vacuolar targeting sequence in tobacco plants re-
sulted in a high accumulation level of the protein, without any
measured changes in biotin metabolism (Murray et al. 2002).
However, similar transformation with construct lacking the tar-
geting sequence failed as no explants were regenerated
(Murray et al. 2002).
It was unknown whether streptavidin expression could be
controlled in order to manipulate biotin metabolism without af-
fecting the plant viability. Low levels of transgenic streptavidin
might induce minor changes in plant development, which in
turn could be used as a tool in studying biotin-dependent
physiological processes. We report here on toxic effects of
overexpression of several streptavidin constructs in tomato
plants. A construct that carried the full bacterial gene fused to
a plant signal peptide for secretion was found to induce de-
generation of plant tissues that could be restored by applica-
tion of biotin.
Materials and Methods
Construction of chimeric streptavidin
Several artificial streptavidin genes were prepared by the PCR amplifi-
cation method, with the BamHI genomic fragment of Streptomyces
avidinii as a template, and specific primers. Direct primers used for
this study were:
P1:5′-actgcagttATGCGCAAGATCGTCG
P2: 5′-gtaaacaatggctCGCAAGATCGTCGTTGCAG
P3: 5′-gactgcagttGACCCCTCCAAGGACTCGAAGGCCCAG
P4: 5′-gtaaacaatggctGACCCCTCCAAGGACTCGAAGGCCCAG
P5: 5′-actgcaGGCATCACCGGCACCTGGTACAAC
P6: 5′-gtaaacaatggctGGCATCACCGGCACCTGGTACAAC
Upper case letters correspond to streptavidin sequences. The di-
rect primers contained additional sequences according to further
construction and expression requirements. Primers P2, 4, 6 were 5′
flanked with plant-translation start-site sequences (lower case letters)
according to Lutcke et al. (1987). Primers P1, 3, 5 were 5′ flanked with
PstI restriction site (lower case, underlined) for further fusion to plant-
transient peptide. Reverse primers used were:
P7:5′-GACTACTGCTGAACGGCG and
P8: 5′-CTACGGCTTCACCTTGGTGAAG.
The following combinations of primers were used to amplify the
various streptavidin artificial genes: each of the direct primers P1, P2,
P3 and P4, together with the reverse primer P7, was used for prepara-
tion of the sps, mst, prost and mprost constructs, respectively. Each of
the direct primers P5 and P6, together with the reverse primer P8 was
used for preparation of cst and mcyto constructs, respectively. The
PCR reaction was performed with 50ng of template and 10pmole of
each direct and reverse primer. Following denaturation for 3 min at
95˚C, 36 amplification cycles of 30s at 94˚C, 45s at 59˚C and 45s at
72˚C, with a 5-min extended elongation step were applied in a PTC-
100 Programmable Thermal Controller (MJ Research Inc., USA). The
PCR products were gel purified, cloned in pGEMTA(Promega) and se-
quenced by the dideoxy chain termination method with an automated
DNA sequencer, dye terminators, and the standard T3, T7 and/or sp6
primers. Computer analyses of nucleotide and amino acid sequences
were carried out with software from the GCG/EGCG package of the
University of Wisconsin, running under a UNIX system.
The PCR products sps, prost and cst, cloned in pGEMTAvector
(Promega) were recloned as PstI fragments in-frame to signal peptide
for secretion, originated from wheat gene for α/β-gliadin storage pro-

Page 3

613Expression of streptavidin in transgenic tomato
tein (pW8233) (accession #×02539). These signal peptide-fused chi-
mera segments as well as the PCR fragments mst, mprost and mcyto,
were ligated between CaMV 35S promoter and NOS terminator to
create plant expression cassettes (Table 1). These cassettes were
cloned into pBin+ vector (Van Engelen et al. 1995) and were used for
either transient expression or stable transformation mediated by Agro-
bacterium tumefaciens strain EHA 105 (Hood et al. 1993). A binary
construct pME (kindly provided by M. Flieshman of the Volcani Cen-
ter) that does not contain any streptavidin sequences was used as a
negative control for the transient expression assay.
Transient expression assay and stable transformation
Seeds of tomato plants (Lycopersicon esculentum cv. VF36) were sur-
face sterilized (with 1% sodium hypochloride) and allowed to germi-
nate on solidified hormone-free Murashige and Skoog (MS) medium in
controlled-growth-room conditions (12/12-h white-light/dark photope-
riod at 25˚C). Seedlings (10 days old) containing two cotyledons and
the apical meristem were taken for transient expression assay. Agro-
bacteria containing the streptavidin and control constructs (see Table
1) were grown for 20 h in 50 mL of liquid YEB medium (pH 7.0) at
28˚C, with shaking at 200rpm, and were transferred to fresh YEB me-
dium (pH 5.2) for an incubation period of 4h under the same condi-
tions. Bacterial cells were precipitated by centrifugation (3000 g,
10min, room temperature), resuspended in 50mL of liquid MS salts
medium supplemented with 2 % sucrose, and added to the tomato
seedlings. Infection was performed by infiltration (2 min under vac-
uum) followed by further incubation (under vacuum) at room temper-
ature for 15min. Co-cultivation was performed for 3–4 days on 1 MM
Whatmann paper discs placed on solidified MS salts medium supple-
mented with 2% sucrose, 1% glucose, 0.25% gelrite and 100µmol/L
acetosyringone (pH 5.2). The seedlings were washed with liquid MS
salts supplemented with 2 % sucrose to remove the Agrobacterium
and were transferred to solidified MS medium supplemented with 3%
sucrose, 0.25 % gelrite, 0.5 mg/L zeatin and 300 mg/L claforan (pH
5.8). Co-cultivation of seedlings and further incubation were per-
formed in a controlled-conditions room with a 12/12-h white-light/dark
photoperiod at 25 ˚C. Tissue degeneration was first observed 48 h
after seedling transplantation. To determine the toxicity levels induced
by the various constructs, each of several Agrobacterium cultures,
each carrying a different streptavidin construct was applied by vac-
uum infiltration to three Petri dishes, each containing nine seedlings.
The seedlings were inspected for degeneration symptoms and were
sorted into four groups according to the severity of the damage. The
experiment was repeated twice and similar results were obtained.
Constructs that resulted in both low toxicity indexes and small num-
bers of affected cotyledons were selected for stable transformation.
Stable transformation of tomato was performed with cotyledons of
sterile-grown seedlings at the Genome Center, Department of Plant
Sciences, the Weizmann Inst., according to a modification of McCor-
mick’s protocol (McCormick 1991). Stable transformation of tobacco
plants (Nicotiana tabacum var. Samsun NN) was performed using the
leaf disc method (Horsch et al. 1985). In both procedures plantlets
were rooted on medium containing 50mg/L kanamycin for selection of
transformants, and 2mg/L biotin to ensure the development of all the
potential plants expressing streptavidin. Plantlets were further ana-
lyzed by PCR with both the nptII primers (direct 5′-CACGCAGGTTCT
CCGGCCGC-3′ reverse 5′-TGCGCTGCGAATCGGGAGCG-3′) and
the streptavidin primers P1 and P7. PCR-positive plantlets were
hardened for 2 weeks and then transferred to a greenhouse (natural
daylight, 25–30˚C) for further plant growth and development.
Segregation ratios of the transgenic trait in F1 plants were deter-
mined by germination of surface-sterilized seeds of the transgenic
plants on solidified MS medium supplemented with 300mg/L kanamy-
cin and 2 mg/L biotin. The kanamycin-resistant/-sensitive seedlings
ratio was determined 1month later.
Biotin determination andWestern blot analysis
Leaf tissues (0.1-g samples) were ground in phosphate-buffered sa-
line (PBS) pH 7.0, containing protease-inhibitor-cocktail tablet (Com-
plete; Boehringer Mannheim, Germany). Biotin levels were deter-
mined by ELISA with alkaline-phosphatase-conjugated Extravidin
(Sigma) according to Shiuan et al. (1997). Results representing sam-
pling of five different plants for each transgenic line were standar-
dized relative to the protein content (in micrograms) of each sample,
as determined by Bradford assay (Bradford 1976). The statistical sig-
nificance was analyzed with Student’s t-test (P ≤0.05).
The same leaf extracts were also used for determination of strepta-
vidin accumulation in the transgenic plants, by Western blot analysis.
Protein samples (10µg) were loaded onto a 15% SDS-polyacrylamide
gel and transferred to a PVDF filter (Boehringer Mannheim) with BIO-
RAD equipment, according to the manufacturer’s instructions. West-
ern assays were performed with anti-streptavidin antibodies (Sigma)
and secondary peroxidase-conjugated goat anti-rabbit IgG (Jackson
ImmunoResearch Lab Inc., PA, USA), with the ECL Kit (Amersham,
UK) used for detection according to the manufacturer’s protocol. Like-
wise, the amount of streptavidin in the plant extracts was evaluated,
with commercial streptavidin used as a standard. For determination of
protein molecular weights, a similar analysis was applied except that
10–20% Tris-tricine Ready Gel System (BIO-RAD), and Polypeptide-
SDS-PAGE Molecular Weight Standards (BIO-RAD) were used ac-
cording to the manufacturers’ protocols.
DNA and RNA gel-blot analysis
For Northern blotting, total RNA was isolated from young leaves of
transgenic tomato and control VF36 plants with the Tri-Reagent Kit
(Molecular Research Center Inc., USA). Samples of 10µg total RNA
were electrophorated on a 1.1% formamide-agarose gel (Sambrook et
al. 1989), blotted onto a nylon membrane (Hybond N, Amersham,
UK), and hybridized with a32P-streptavidin probe or with a 1-kb to-
mato
method (Feinberg and Vogelstein 1987). Hybridization was carried out
at 65˚C in 0.263mol/L Na2HPO4, 1% (w/v) BSA, 7% (w/v) SDS and
1mmol/L EDTA. The membrane was washed twice with 2× SSC, 0.1%
(w/v) SDS at room temperature for 10min, and twice with 0.2× SSC,
0.1% (w/v) SDS at 60 ˚C for 10 min. The blot was exposed either to
Biomax X-ray film (Kodak) with an intensifying screen, at –70˚C, or to
a Phosphor-Imager screen. The Phosphor-Imager Program (Fujix BAS
1500, Fuji, Japan) was used for radioactivity quantification.
Genomic DNA was isolated from young leaves of transgenic to-
mato and a wild-type strain according to Chee et al. (1991). For South-
ern blot analysis, 10µg DNA was digested with EcoRI so that one di-
gestion would be performed at the end of the transgenic fragment
and the others in the plant genome. The digested DNA samples were
electrophorated on a 0.8 % agarose gel in Tris-acetate (TAE) buffer
and blotted onto a Hybond-N+membrane (Amersham, UK) with
32P-rDNA fragment (28S) labelled by the random-priming

Page 4

614Idit Ginzberg et al.
0.4mol/L NaOH. The membrane was hybridized with the32P-760-bp
PCR fragment of nptII fragment, labelled by the random-priming
method. Hybridization, washing and blot exposure conditions were as
described for Northern analysis.
Results and Discussion
Evaluation of streptavidin toxicity levels
In plants the expression of the biotin-binding proteins, avidin
or streptavidin, succeeded or failed depending on the se-
quences added to the transgenic gene (Hood et al. 1997, Mur-
ray et al. 2002). Regeneration of transformants could not be
obtained unless a targeted signal was fused to these biotin-
binding proteins. Moreover, considerable accumulation of avi-
din was detected: the expression level of avidin localised to
the intercellular spaces of maize was found to be about 2% of
aqueous soluble extracted proteins from dry seeds (Hood et
al. 1997); and avidin localised to the vacuoles of tobacco ac-
cumulated up to 1.5% of total leaf protein (Murray et al. 2002).
Nevertheless, normal plant development was observed, and
no significant differences were observed in biotin levels com-
pared to non-transgenic controls (Murray et al. 2002).
In an effort to evaluate the implications of altered biotin me-
tabolism for plant development, several artificial genes of
streptavidin were constructed on the basis of its proteolysis
processes in the bacterial cell (Bayer et al. 1989) (Table 1). To
assess the correlation between toxicity level and cellular lo-
calisation of streptavidin, each artificial streptavidin gene was
also fused to a plant-signal peptide for secretion (from wheat
gene for α/β-gliadin storage protein; accession #X02539) to
exclude its activity from the cytoplasmic biotin pool. The tox-
icity levels induced by the various constructs were monitored
Table 1. Streptavidin constructs used for transient expression assay
and the toxicity level obtained on tomato seedlings after 48–96h.
Streptavidin constructs Cotyledons
affected
(%)
Toxicity
index
(0–3)
pME (control)
sps
mst
prost
mprost
cst
mcyto
00
2
3
1
1
3
3
40
50
39
38
63
52
CaMV35S promoter,
streptavidin signal peptide,
Core-streptavidin,
Toxicity level was rated in three categories depending on the intensity
and the severity of the symptoms on the affected cotyledons: 3 – high,
2 – moderate, 1 – low, or 0 – normal-looking tissue.
Plant signal peptide,
Streptavidin processed termini,
Nos terminator.
Bacterial
in a transient expression assay by scoring the appearance of
necrotic lesions-like symptoms on tomato seedling cotyle-
dons 48–96h after infection. All streptavidin-containing con-
structs (derived by the constitutive promoter CaMV 35S) in-
duced the development of lesion-like symptoms, but to differ-
ing extents (Table 1). The core-streptavidin constructs cst and
mcyto, and the pro-streptavidin construct mst caused the
most severe damage to the tomato seedlings: broad lesion-
like symptoms appeared on both stems and cotyledons of all
seedlings sampled (100 %); in addition, the cotyledons be-
came chlorotic even when lesions were not observed (desig-
nated as toxicity index 3). The expression of the streptavidin
constructs prost and mprost induced the least damage: fewer
seedlings were affected (75–79%) and the lesion-like symp-
toms were smaller and appeared mainly on cotyledons,
whose tissue, nevertheless, remained green in colour (toxicity
index 1). The damage induced by the sps construct seemed
to fall between the two levels described above: the affected
seedlings (80%) exhibited light chlorosis around the necrotic
spots (toxicity index 2). These results imply that the ‹core›
streptavidin constructs (mcyto and cst) are more toxic than
the unprocessed forms prost, mprost and sps (Table 1), sug-
gesting either that the streptavidin protein may not be cor-
rectly processed and folded in the plant cell, which would re-
duce its toxicity effect, or that the unprocessed forms of the
protein may be immediately degraded and, therefore, induce
the least damage to the cells. At this point it is hard to predict
the role of the bacterial signal peptide in the mst construct,
but the observation that its toxicity was greater than that of
the mprost construct (Table 1) suggests that it may have ac-
cumulated to a higher level. Presumably, the bacterial signal
peptide stabilized the streptavidin precursor, thus reducing its
degradation potential, whereas the addition of plant signal
peptide, as in the sps construct (Table 1), annulled that effect.
It was considered that using the mcyto, cst or mst con-
struct might lead to severe plant degeneration, whereas the
expression of the prost or mprost constructs might be too
weak to reduce the biotin level significantly. Therefore, the
sps construct was chosen as a means to obtain stable trans-
formation of tomato plants expressing streptavidin.
Streptavidin induces tissue degeneration in transgenic
tomato plants
In order to monitor the plant tissue responses to streptavidin
gene expression, the sps gene was fused to three different
plant promoters: the constitutive CaMV 35S promoter for ex-
pression in whole plant organs; the tobacco root-specific
TobRB7 (Yamamoto et al. 1991); and the French bean β-pha-
seolin promoter that directs endosperm- and embryo-specific
expression (Karchi et al. 1994). Nonetheless, whichever
streptavidin chimera was used for tomato transformation, the
plantlets were regenerated on media containing biotin, to
enable all the potential transgenic plants to develop, even
those expressing streptavidin at toxic levels.

Page 5

615Expression of streptavidin in transgenic tomato
Figure1. Transgenic tomato plants expressing streptavidin under the
CaMV 35S promoter. (a) Overexpression of sps-streptavidin gene in-
duced lesion-like symptoms on the stems of the transgenic plants
(marked with arrows) and (b) alteration in the morphology of their
leaves and flowers.
Figure2. Tissue specific expression of streptavidin. (a) Seedless fruits
of four tomato transgenic plants (T23, 40, 41, 44) expressing the sps-
streptavidin gene under the TobRB7 promoter compared with a fruit
of non-transgenic control (C, middle). (b) Non-viable, deformed to-
mato seeds of transgenic plant harboring the sps-streptavidin gene
under the β-phaseolin promoter (lower panel) compared with viable,
germinating seeds of non-transgenic control plant (upper panel).
Figure3. Restoration of degenerate transgenic tomato plant. Young stems of plant #13 were rooted into three pots; (a) constant spraying with
2-mg/L biotin resulted in normal growth, (b) no application of biotin resulted in the degenerate phenotype and (c) local application of biotin res-
tored the degenerate plant.