Planta (2000) 210: 232±240
Transgene expression driven by heterologous
ribulose-1,5-bisphosphate carboxylase/oxygenase small-subunit
gene promoters in the vegetative tissues of apple (Malus pumila Mill.)
John R. Gittins1, *, Till K. Pellny1, Elizabeth R. Hiles1, Cristina Rosa2, Stefano Biricolti2, David J. James1
1
Plant Breeding and Biotechnology, Horticulture Research International, East Malling, West Malling, Kent, ME19 6BJ, UK
Dipartimento di Orto¯orofrutticoltura, UniversitaÁ degli Studi di Firenze, 50144 Firenze, Italy
2
Received: 15 June 1999 / Accepted: 12 August 1999
Abstract. It is desirable that the expression of transgenes
in genetically modi®ed crops is restricted to the tissues
requiring the encoded activity. To this end, we have
studied the ability of the heterologous ribulose-1,5bisphosphate carboxylase/oxygenase (Rubisco) smallsubunit (SSU) gene promoters, RBCS3CP (0.8 kbp)
from tomato (hycopersion esculentum Mill.) and SRS1P
(1.5 kbp) from soybean (Glycine max [h.] Mers.), to drive
expression of the b-glucuronidase (gusA) marker gene in
apple (Malus pumila Mill.). Transgenic lines of cultivar
Greensleeves were produced by Agrobacterium-mediated
transformation and the level of gusA expression in the
vegetative tissues of young plants was compared with
that produced using the cauli¯ower mosaic virus
(CaMV) 35S promoter. These quantitative GUS data
were assessed for their relationship to the copy number of
transgene loci. The precise location of GUS activity in
leaves was identi®ed histochemically. The heterologous
SSU promoters were active primarily in the green
vegetative tissues of apple, although activity in the roots
was noticeably higher with the RBCS3C promoter than
with the SRS1 promoter. The mean GUS activity in leaf
tissue of the SSU promoter transgenics was approximately half that of plants containing the CaMV 35S
promoter. Histochemical analysis demonstrated that
GUS activity was localised to the mesophyll and palisade
cells of the leaf. The in¯uence of light on expression was
also determined. The activity of the SRS1 promoter was
strictly dependent on light, whereas that of the RBCS3C
promoter appeared not to be. Both SSU promoters
would be suitable for the expression of transgenes in
green photosynthetic tissues of apple.
*Present address: Department of Chemistry, Biochemistry and
Biophysics, GoÈteborg University, S-40530 GoÈteborg, Sweden
Abbreviations: CaMV = cauli¯ower mosaic virus; GUS =
b-glucuronidase; 4-MU = 4-methyl-umbelliferone; MUG =
4-methyl-umbelliferyl-b-D-glucuronide; SSU = small subunit;
X-Gluc = 5-bromo-4-chloro-3-indolyl-b-D-glucuronic acid
Correspondence to: D.J. James; E-mail: [email protected];
Fax: 44 (1732) 849 067
Key words: Gene promoter ± b-Glucuronidase ± Malus
(transgenic) ± Ribulose-1,5-bisphosphate carboxylase/
oxygenase(SSU) ± Transgene expression (gusA) ±
Transgenic apple
Introduction
Constitutive and non-speci®c promoters such as CaMV
(cauli¯ower mosaic virus) 35S and nos (nopaline synthase) have been extremely useful as experimental tools
to assess the e€ects of transgene expression in many
plant species. With such promoters, a gene is expressed
in the majority of tissues throughout plant growth and
development. This lack of temporal and spatial regulation may be suitable for proof-of-concept experiments
but has a number of potential drawbacks for use in
genetically improved crops.
Localisation of transgene expression may be essential
for the expression of certain antisense constructs,
biosynthetic genes or transcription factors in order to
obtain transgenic regenerants or avoid phenotypic
abnormalities (Kumar et al. 1996; Martin 1996). The
presence of multiple transgenes driven by the same
constitutive promoter in a single plant (e.g. for resistance-gene pyramiding) may result in homology-dependent gene silencing, particularly where the promoter is
also highly active (Vaucheret et al. 1998). There is also
the potential risk that constitutive expression of viral
capsid proteins in transgenic plants may increase the risk
of transcapsidation or viral recombination to generate
new strains of phytopathogen (Robinson 1996). Lastly,
localised and targeted gene expression may be desirable
to satisfy regulations regarding food safety.
To address these concerns we have undertaken a
study of the ecacy of various tissue-speci®c promoters
in transgenic apple plants. Here we report ®ndings
concerning the level and location of expression of the
marker gene gusA using promoters from Rubisco small
subunit (SSU) genes of tomato and soybean.
J.R. Gittins et al.: Transgene expression in apple driven by heterologous Rubisco SSU promoters
Ribulose-1,5-bisphosphate
carboxylase/oxygenase
catalyses the competing reactions of photosynthetic
carbon ®xation and photorespiration in higher plants
and green algae. It is a multimeric enzyme composed of
eight large and eight small subunits (Jensen and Bahr
1977). The large subunit is encoded by the chloroplast
genome, whereas the SSU polypeptides are encoded by
nuclear gene families. The coding regions of di€erent SSU gene family members are highly homologous, although the level and pattern of expression
can vary.
The tomato Rubisco SSU gene RBCS3C is one of the
®ve members of a gene family (RBCS1, 2, 3A, 3B, 3C )
whose pattern of expression and promoter sequences
have been extensively studied (Sugita and Gruissem
1987; Sugita et al. 1987; Manzara et al. 1991; Wanner
and Gruissem 1991; Carrasco et al. 1993; Manzara et al.
1993; Meier et al. 1995). Individual members of this gene
family show varied patterns of temporal and organspeci®c expression. However, all ®ve genes are actively
transcribed in light-grown cotyledons and leaves and
expression is undetectable in roots. Like other genes
involved in photosynthesis, the expression of the tomato
RBCS genes is regulated by light. Soon after the transfer
of plants to darkness, levels of RBCS2 and RBCS3A
mRNA in leaves are reduced, whereas transcripts of
RBCS1, RBCS3B and RBCS3C become undetectable.
Levels of transcript rapidly return to normal after
placing the plants under light.
Soybean contains at least six genes encoding the
Rubisco SSU, of which only the two most highly
expressed, SRS1 and SRS4, have been characterised in
detail (Berry-Lowe et al. 1982; Shirley et al. 1990).
These highly homologous genes are actively transcribed
in light-grown leaves and, as with the tomato RBCS
genes, expression of the soybean SRS genes is regulated
by light. Transcript levels are rapidly reduced upon
transferring plants to darkness and return to normal
when light is restored.
In this study, the tomato RBCS3C gene promoter
was chosen to drive transgene expression in apple
because this transcript is relatively abundant in leaf
tissue and is apparently not expressed in young fruit
tissues (Wanner and Gruissem 1991). The promoter for
the soybean SRS1 gene was chosen for comparison with
the RBCS3C promoter, and the activities of these two
SSU promoters were compared with that of the constitutive CaMV 35S promoter.
Despite originating from very di€erent plants, the
heterologous Rubisco SSU promoters were found to
give similar activities to those seen in their own species
and to one another when used to drive expression of a
marker gene in the woody perennial apple.
Materials and methods
Promoter DNAs and vectors. The tomato RBSCS3C promoter was isolated from a genomic HindIII fragment in construct
pRBCS3CHin (W. Gruissem, UC Berkeley, USA; Sugita et al.
1987). The soybean SRS1 promoter, originally isolated in the
laboratory of Professor R Meagher, University of Georgia,
233
USA (Berry-Lowe et al. 1982), was obtained as a genomic
EcoR I-EcoR V fragment in the construct pAB7 (V. BuchananWollaston, HRI-Wellesbourne, UK). All binary constructs were
based on vector pSCV1.6 (G. Edwards, Shell Forestry, East
Malling, UK, personal communication). pSCV1.6 contains the
repaired nptII gene driven by the CaMV 35S promoter next to the
right T-DNA border, an intron-containing gusA gene driven by 35S
(Vancanneyt et al. 1990) next to the left border and a T-DNA
transfer enhancer sequence (overdrive) outside the T-DNA adjacent to the right border.
Binary vector construction. The two Rubisco SSU promoters
were subcloned into pSCV1.6, replacing the CaMV 35S promoter
to produce transcriptional fusions with the gusA gene. Standard
procedures (Sambrook et al. 1989) were used except where
indicated.
pRBCS3CHin, puri®ed using a QIAprep Spin Miniprep kit
(QIAGEN), was digested with Sau3A I to release a 1.8-kbp
fragment containing the entire RBCS3C gene with approximately
0.8 kbp of 5¢ ¯anking sequence. This was puri®ed from a 0.8%
TAE-agarose gel (Geneclean II kit; BIO101 Inc., La Jolla, Calif.,
USA) and ligated to vector pGEM3Z (Promega) cleaved with
BamH I. Following ligation, Escherichia coli strain DH5a was
transformed by electroporation (E. coli Pulser; BioRad).
A recombinant clone with the EcoR I site of the vector located at
the 5¢ end of the RBCS3C fragment, named pRBCS3CSau, was
digested with endonuclease Eco57 I which cleaves at a single site
9 bp proximal to the RBCS3C coding region. The 3¢ overhang
generated by Eco57 I cleavage was removed by treatment with T4
DNA polymerase in the presence of dNTPs. The 0.8 kbp of 5¢
¯anking sequence was excised by digestion with EcoR I, gel-puri®ed
and ligated to vector pSCV1.6, cleaved with EcoR I and Sma I. The
recombinant binary vector was named pSCV1.6RBCS3CP.
pAB7 was digested with EcoRI and EcoRV which cleave at the
5¢ end of the SRS1 promoter sequence and 45 bp proximal to the
start of the SRS1 coding region, respectively. The excised 1.5-kbp
SRS1 promoter fragment was gel-puri®ed and ligated to vector
pSCV1.6 cleaved with EcoRI and Sma I to produce construct
pSCV1.6SRS1P.
The identity of recombinant binary plasmids was veri®ed by
restriction analysis and sequencing of the cloning junctions using a
PRISM dye-terminator cycle sequencing kit (Applied Biosystems).
Sequencing reactions were analysed at the University of Durham,
UK.
Plant material. Shoots of apple (Malus pumila Mill. cv. Green-
sleeves) were propagated in vitro as described by James and
Dandekar (1991). For experimentation, plants were induced to root
in vitro and transferred to potting compost and grown in a
propagator under a 16 h light/8 h dark photoperiod. The photon
¯ux density at photosynthetically active wavelengths (400±700 nm)
was measured to be 73±80 lmol m)2 s)1 using a quantum radiometer (Q102; Macam Livingston, UK).
The control plants included in the study were: untransformed
Greensleeves clone GS92 ()ve) and clone B3 (+ve), a single-copy
apple transgenic produced by transformation with the vector
pSCV1.6 in which gusA expression is driven by the CaMV 35S
promoter. Among CaMV 35S-gusA apple transgenics, B3 represents a typical clone with respect to the level and pattern of
expression (data not shown).
Apple transformation. The binary vectors pSCV1.6RBCS3CP
and pSCV1.6SRS1P were introduced into the disarmed Agrobacterium tumefaciens strain EHA101 (Hood et al. 1986) by electroporation. Apple cultivar Greensleeves, clone GS92, was
micropropagated in vitro and transformed using the leaf disc cocultivation method (Horsch et al. 1985) with modi®cations described by James and Dandekar (1991). Transgenic shoots were
selected by growth in the presence of kanamycin and identi®ed by
histochemical staining for b-glucuronidase (GUS) activity (Je€erson 1987).
234
J.R. Gittins et al.: Transgene expression in apple driven by heterologous Rubisco SSU promoters
Sampling of vegetative tissues for analysis. For sampling of
tissues to determine the levels of GUS activity, three duplicate
shoots of each clone were rooted in compost and grown at 20 °C
under a 16 h light/8 h dark photoperiod until they had reached a
height of approximately 15 cm. Samples were taken from the plants
according to the following scheme:
(i)
From all three plants, one disc was taken from each of the ®rst
pair of expanded leaves. The discs were cut with a cork borer
(size 2 = 6 mm diameter) from either side of the main vein.
(ii) From two of the plants, two samples of root were taken. One
was a section of brown/pink pigmented mature root, from
close to the base of the stem, in most cases formed in vitro. The
other was a section of actively growing ®ne white young root.
(iii) From one of the three duplicate plants, one petiole and one
section of green stem was taken. The petiole was cut from one
of the sampled leaves and the stem section was taken from the
®rst internode beneath the sampled leaves.
All samples were placed in 1.5-ml microfuge tubes, frozen in liquid
nitrogen and then stored at )80 °C.
Fluorometric GUS assay. Frozen tissues were brie¯y ground
in 1.5-ml microfuge tubes using a chilled plastic pellet pestle
(Berkhard Scienti®c Sales, Uxbridge, UK) before addition of 250 ll
of GUS extraction bu€er (Je€erson 1987). Methanol was added to
this bu€er to a concentration of 20% (v/v) to enhance GUS activity
(Wilkinson et al. 1994). For roots, 2-mercaptoethanol (10 mM)
was added to the bu€er to prevent browning. After further grinding
in bu€er, the tubes were centrifuged for 15 min at 13,000 g and the
supernatant removed to fresh tubes; 150 ll for GUS assay and 50 ll
for protein assay.
The activity of GUS in apple vegetative tissue samples was
quanti®ed by ¯uorometric assay (Je€erson 1987) using the
substrate 4-methyl-umbelliferyl-b-D-glucuronide (MUG). The reaction was carreid out for 60 min after preliminary experiments
con®rmed that the reaction was linear to this time point, as
reported by Je€erson (1987). The reaction was stopped by the
addition of 0.2 M Na2CO3. The amount of 4-methyl-umbelliferone
(4-MU) produced in each reaction mixture was determined by
¯uorometric measurement using a Hoefer LKB DyNA quant 200
¯uorometer (b excitation 365 nm/b emission 455 nm).
In parallel with the GUS assay, the protein content of each
sample was determined using the protein assay II kit (BioRad) with
BSA as the standard. The speci®c enzyme activity was calculated
for each sample in nmol 4-MU (mg protein))1 min)1.
Genomic DNA isolation and analysis. Genomic DNA was
isolated using a Nucleon Phytopure kit (Amersham Pharmacia
Biotech) from leaf tissue (1 g). Puri®ed DNA (6 lg) was digested
with restriction endonuclease Kpn I. Digested DNAs were electrophoretically separated on a 0.8% TAE-agarose gel and capillaryblotted onto a nylon membrane (Magna; Micron Separations Inc.,
Westborough, Mass., USA). The blot was probed separately using
nptII and gusA probe fragments labelled with b-[32P] dCTP
(Amersham) using a Ready-to-Go kit (Pharmacia). Conditions
for hybridisation, washing and stripping of blots were those
recommended for use with digoxygenin-labelled DNA probes
(Boehringer Mannheim). Hybridisation of the probe DNAs to the
blot was recorded on blue-sensitive X-ray ®lm (GRI, Dunmow,
Essex, UK).
Isolation and analysis of RNA. Three duplicates of each clone
under study were propagated in compost as described. All plants
were grown under the same lighting conditions and one was
sacri®ced each day for sampling. Leaf samples were collected at
midday (to eliminate possible e€ects of circadian rhythms) and
immediately frozen in liquid nitrogen and stored at )80 °C.
Samples of RNA were isolated from between 0.5±1.5 g of leaf
tissue using the procedure of Goldsbrough and Cullis (1981).
Samples of 20 lg of total RNA were electrophoretically separated
on a 1.2% formaldehyde agarose gel and then capillary blotted
onto a Magna nylon membrane. The blot was probed using a gusA
probe fragment labelled with b-[32P]dCTP. Conditions for hybridisation and washing of blots were those recommended for use with
digoxygenin-labelled DNA probes (Boehringer Mannheim). Hybridisation of the probe DNAs to the blot was recorded on bluesensitive X-ray ®lm.
Histochemical GUS assay. Leaves taken from plants propa-
gated in compost (approx. 15 cm tall) were placed in 100 mM
sodium phosphate (pH 7.0), 10% (v/v) glycerol and held under
mild vacuum (300 mbar) for approx. 1 h. Leaf pieces of 1 cm2 were
then rapidly embedded in ice on a freezing microtome (Leitz
Wetzlar) sample holder which had been pre-chilled with CO2.
Sections (20±30 lm) were cut using a disposable blade. These were
initially ¯oated on water and then transferred to fresh 5-bromo-4chloro-3-indolyl-b-D-glucuronic acid (X-Gluc) substrate [1 mM XGluc in 100 mM phosphate bu€er (pH 7.0); 10 mM EDTA; 1 mM
potassium ferricyanide; 1 mM potassium ferrocyanide; 0.1% (v/v)
Triton X100] and held at 37 °C for 2 h. Chlorophyll was cleared
from the stained sections by overnight incubation at 55 °C in 80%
(v/v) lactic acid. They were then mounted in 80% lactic acid and
covered with a coverslip. The sections were examined using a light
microscope (Leitz Dialux), and staining patterns recorded by
photography using Kodak EPY 64 T ®lm.
Results
Production of binary vectors and apple transformation. To ensure that the expression produced by the
tomato and soybean SSU promoters would be directly
comparable, each was subcloned into the binary vector
pSCV1.6 to replace the CaMV 35S promoter driving
gusA (Fig. 1A). Transcriptional fusions were favoured
over translational fusions because it was thought that
with the former, variations in translational eciency
would be less likely.
The strategy to subclone the SRS1 promoter introduced an out-of-frame ATG triplet 19 bp before the
actual gusA start codon. This non-authentic start codon
is likely to overlap with the predicted transcription start
site (Fig. 1B; Berry-Lowe et al. 1982). Therefore, it is
probable that it is not included in the mRNA, or is so
close to the 5¢ terminus that the decoding site of the
ribosome is downstream of this codon when the 40S
subunit is bound to the 5¢ end of the message (FuÈtterer
and Hohn 1996). This feature did not apparently a€ect
the level of expression driven by the SRS1 promoter in
this study.
Agrobacterium tumefaciens EHA101 containing binary
vectors pSCV1.6RBCS3CP and pSCV1.6SRS1P was
used in transformation of apple cultivar Greensleeves.
Twelve separate transgenic shoots containing the
RBCS3C and twelve containing the SRS1 promoter
were identi®ed, multiplied in vitro and then propagated
in compost for sampling. For comparison, untransformed plants of cultivar Greensleeves (GS92; )ve) and
plants of clone B3 (CaMV 35S-gusA; +ve) were
propagated in an identical manner.
Activities of the RBCS3C and the SRS1 promoters in
young vegetative tissues of apple. b-Glucuronidase
(GUS) activities in samples of leaves, petioles, green
J.R. Gittins et al.: Transgene expression in apple driven by heterologous Rubisco SSU promoters
235
Fig. 1A,B. Binary vectors used in
this study. A Structure of the
pSCV1.6 T-DNA. B, BamH I; E,
EcoR I; H, Hind III; K, Kpn I; OD,
overdrive T-DNA transfer enhancer;
S, Sma I; Sn, SnaB I. B gusA 5¢ leader
sequences in pSCV1.6,
pSCV1.6RBCS3CP and
pSCV1.6SRS1P. The positions of the
TATA box and ATG start codons
are shown in bold text, underlined.
The Sma I site of pSCV1.6 used in
insertion of the SSU promoters is
marked and the SSU promoter-gusA
transcriptional fusion junctions are
indicated beneath the sequences (D).
Arrows indicate the location of the
start of transcription for the CaMV
35S (Guilley et al. 1982), RBCS3C
(Sugita et al. 1987) and SRS1 (BerryLowe et al. 1982) promoters. This is
not certain for the pSCV1.6SRS1P
construct because the promoter/gusA
fusion junction is located at the
putative transcription start site
stems, mature roots and young roots were determined by
¯uorometric assay. Background GUS activity in untransformed tissues of clone GS92 was found to be
negligible (data not shown). The data for speci®c GUS
activity in leaf discs of the RBCS3C and SRS1 promoter
transgenics, respectively are shown in Fig. 2 with the
separate clones presented in ascending order of activity.
The speci®c activity seen in separate leaf discs taken
from duplicates of the same clone was fairly constant.
Absolute levels of GUS activity produced by either the
tomato or soybean SSU promoters in leaf tissue were
similar. The overall mean values for the activity of the
RBCS3C and SRS1 promoters in leaf discs were 11.3
and 9.6 nmol 4-MU (mg protein))1 min)1 respectively.
In comparison, the mean GUS activity in the leaves of
clone B3 (constitutive CaMV 35S promoter +ve
control) was 18.1 nmol 4-MU (mg protein))1 min)1.
Direct comparison of the GUS expression levels in
other vegetative tissues within and between individual
transgenic clones was not possible due to low sample
sizes, widely di€erent tissue characteristics (protein
content/cell density) and di€ering extraction bu€er
composition. To allow comparison of the level of GUS
activity in di€erent vegetative tissues, mean speci®c
activity values in leaf discs, petioles, stems, young roots
and mature roots for the RBCS3C and SRS1 promoter
transgenics were related to the activity in the equivalent
tissues of clone B3. In each tissue, the mean B3 value
was nominally called 100% activity and the mean SSU
promoter GUS activity values were expressed as a
percentage of this (Fig. 3). It can be seen that GUS
activity driven by the SSU promoters was largely
con®ned to green tissues. Relative activity was highest
in leaf discs, lower in petioles and lower still in stems.
Although GUS activity in the roots was very low, as
would be expected, it was markedly higher in the
RBCS3CP plants than in the SRS1P plants (Fig. 4).
In¯uence of transgene locus copy number on the level of
GUS activity in leaves. The transgene locus copy
number of individual RBCS3CP and SRS1P clones
was estimated from a Southern blot of Kpn I-digested
genomic DNA. Separate probing of this blot with gusA
and nptII gene fragments gave independent estimates of
the locus copy number. Because of the presence of a Kpn
I site in the SRS1 promoter only the nptII result gave
useful copy-number data for the SRS transgenics. Seven
of the 12 clones for each of the promoters RBCS3CP
and SRS1P contained single transgene loci, but it is
apparent that the highest expressors all contained
multiple transgene loci (clones 8 and 14 for RBCS3CP
and clones 15, 24, 37 for RBCS3CP. In the RBCS3CP
plants, two of the lowest expressors also possessed
multiple copies (clones 1, 21).
The tomato RBCS3CP promoter appeared to show
more than one copy for those clones having the highest
and the lowest levels of expression but this pattern was
not repeated for the other promoter from soyabean,
SRS1P.
Histochemical localisation of GUS activity. The activity
of GUS was localised by examination of sections of
leaves treated with the chromogenic GUS substrate XGluc (Fig. 5). The results presented are typical of the
staining pattern in experiments using two separate
transgenic clones for each promoter. In leaves of plants
236
J.R. Gittins et al.: Transgene expression in apple driven by heterologous Rubisco SSU promoters
Fig. 3. b-Glucuronidase activities of SSU promoters in vegetative
tissues related to the constitutive expression driven by the CaMV 35S
promoter in clone B3. The mean B3 GUS activity values were
obtained from leaf discs (22 samples/11 plants), petioles (9 samples/9
plants), stems (7 samples/7 plants), mature roots (10 samples/10
plants) and young roots (10 samples/10 plants). Assigning the level of
GUS activity in this CaMV 35S clone as the 100% value for each
tissue, the mean GUS activity for the 12 separate clones of the
RBCS3C and SRS1 promoter transgenics is presented as a percentage
of this value. The number of samples used to derive each mean value
is given below the bars
Fig. 2. b-Glucuronidase activities in leaves of RBCS3C- and SRS1promoter transgenics, with locus copy numbers for each clone. The
mean speci®c activity (‹SE) of GUS in leaf discs of 12 separate apple
clones containing RBCS3C-gusA and SRS1-gusA is presented., For
each clone, a total of six leaf discs (two from each of three duplicate
plants) was assayed. The locus copy number established by Southern
blotting is shown beneath the name of each clone. Where the copy
number is uncertain, two ®gures are given
containing either of the SSU promoters driving gusA
expression, histochemical staining was mainly localised to the mesophyll layer and the palisade cells
(Fig. 5A,B,D,E). The SSU promoters were very active
in the palisade layer but much less active in the spongy
mesophyll, the epidermis and cells of the vascular bundle
of the mid-vein. The lack of expression in the inner and
outer phloem of the vascular bundle is particularly
obvious (Fig. 5D,E). Most cells of the leaf, with the
exception of the cortex, were heavily stained in an
equivalent section of a leaf from a CaMV 35S-gusA
plant (clone B3; Fig. 5C,F).
In¯uence of light on the regulation of SSU-promoter
activity. The SSU-transcript is one of at least 100 plant
transcripts whose abundance is regulated by light
(Terzaghi and Cashmore 1995) and the fusion of SSU
promoters to reporter genes has often been used to study
light regulation in transgenic plants. When used in
Fig. 4. b-Glucuronidase activities in root tissues of RBCS3C- and
SRS1-promoter transgenics. The mean speci®c activity (‹SE) in
young roots (24 samples) and mature roots (24 samples) of apple
clones containing RBCS3C-gusA and SRS1-gusA is presented
heterologous species, SSU promoters typically show
light regulation, suggesting that the regulatory mechanisms are conserved during evolution. To determine
whether the tomato RBCS3C and the soybean SRS1
promoters show light regulation in apple, steady-state
levels of gusA transcript were examined in plants grown
in darkness for 24 h and then returned to a 16 h light/
8 h dark photoperiod. This light regime has previously
been used to examine the light-dependence of expression
of the tomato and soybean SSU genes in their own
species. Under such a regime, the level of the two SSU
transcripts rapidly decreases to become virtually undetectable after 24 h in darkness (Sugita and Gruissem
1987; Shirley et al. 1990).
Before being placed in the dark (Fig. 6, L1), gusA
mRNA was clearly present in the leaves of plants
containing the RBCS3CP promoter (clones 14 and 16)
and the SRS1 promoter (clones 17 and 24) driving
gusA. The steady-state levels of the gusA transcript
J.R. Gittins et al.: Transgene expression in apple driven by heterologous Rubisco SSU promoters
237
Fig. 5A±F. Localisation of GUS activity in leaves of RBCS3C-,
SRS1- and CaMV-35S promoter transgenics by staining with the
chromogenic substrate X-Gluc. Stained transverse leaf sections from
RBCS3C-gusA clone 16 (A, D) and SRS1-gusA clone 24 (B, E) plants
are compared with an identical section from a CaMV 35S-gusA
transgenic, clone B3 (C, F). Transverse sections through the leaf midvein. i, inner phloem; o, outer phloem; p, palisade mesophyll; s,
spongy mesophyll. The scale of each micrograph is shown by a size
bar representing 100 lm
correlated with the level of GUS activity found in the
leaves of the respective clones, i.e. the higher expressors
(SRS1P 24 and RBCS3CP 14) had higher levels of
transcript than lower expressors (SRS1P 17 and
RBCS3CP 16).
After 24 h in darkness (Fig. 6, D), gusA transcript
levels in the two SRS1P clones were greatly reduced,
con®rming a marked down-regulation of the SRS1
promoter in the absence of light. Unexpectedly, the
RBCS3CP clones showed similar levels of gusA transcript before and after 24 h of darkness.
Plants grown in darkness were then moved back into
the light for a further 24 h under a 16 h light/8 h dark
photoperiod and the steady-state levels of gusA transcript examined (Fig. 6, L2). In SRS1P clones, the gusA
transcript levels had recovered substantially but had not
yet reached the amounts seen in the leaves of plants
before the dark treatment. The gusA transcript levels in
238
J.R. Gittins et al.: Transgene expression in apple driven by heterologous Rubisco SSU promoters
SRS1P
17
L1
D
SRS1P
24
L2 L1 D
RBCS3CP
14
L2 L1
RBCS3CP
16
D L2 L1 D L2
Fig. 6. Light regulation of the RBCS3C and SRS1 promoters in
transgenic apple leaves. Steady-state gusA mRNA levels were
examined by Northern blotting (upper panel). Two independent
transgenic clones were analysed for each SSU promoter. Leaves were
harvested from duplicates of each clone grown under a 16 h light/8 h
dark photoperiod (L1), after 24 h in total darkness (D) and after 24 h
following placement back under the 16 h light/8 h dark photoperiod
(L2). Total RNA was isolated from the leaf material and 20 lg was
loaded into each lane. Ethidium bromide staining of ribosomal RNA
(lower panel) was used to demonstrate equivalent loading
the RBCS3CP clones were slightly increased by replacement in the light. It appears that the activity of the SRS1
promoter is strictly dependent on light, whereas that of
the RBCS3C promoter fragment used here is not.
Discussion
The aim of this study was to investigate the activity of
heterologous Rubisco SSU promoters in apple. We have
produced populations of transgenic plants in which the
expression of gusA is driven by promoters from soybean
(SRS1) and tomato (RBCS3C). The expression observed
in young apple plants was largely con®ned to leaves,
with some activity seen in the petiole and green stem
tissues. b-Glucuronidase activity was localised to the
photosynthetically active leaf mesophyll and palisade
cells, as is normally seen for Rubisco. The mean activity
of the two SSU promoters in apple leaves appeared to be
approximately half that of the constitutive CaMV 35S
promoter in the typical clone B3.
In¯uence of transgene locus copy number on the level of
GUS activity. When the copy number of transgene loci
in separate apple clones is related to the GUS activity in
leaves, the extremes of expression can sometimes be
related to multiple transgene loci. In the case of the
RBCS3CP clones, some of the lowest expressors also
contained more than one copy of the transgene.
However the strength of the relationship between levels
of expression and transgene copy number is dicult to
judge from the present data since some clones in the
mid-range of GUS values also had more than one copy.
This suggests that, depending on the location and nature
of the transgene integration, multiple inserts can either
raise the expression level by a gene-dosage e€ect or
lower it by some gene-silencing mechanism. This is in
agreement with the ®ndings in our laboratory (data not
shown) and of a previous study on the e€ect of transgene
copy number on expression level in tobacco (Hobbs
et al. 1993). Further study is required to determine
whether the putative transgene silencing observed in
some clones operates at the transcriptional or at the
post-transcriptional level (Meyer and Saedler 1996;
Vaucheret et al. 1998).
In¯uence of light on the regulation of SSU promoter
activity. The activity of both SSU promoters is largely
con®ned to green photosynthetic tissues in apple.
However, the soybean SRS1 promoter is strictly regulated by light whereas the tomato RBCS3C promoter is
not. The higher GUS activity in roots driven by the
RBCS3C promoter compared with the SRS1 promoter
(Fig. 4) may be another indication that the promoter is
defective in light-dependent regulation. Both promoters
used here have previously been used to drive authentic
light-regulated gene expression in di€erent heterologous
species (Quandt et al. 1992; Kyozuka et al. 1993). This
apparent discrepancy may have one of several explanations:
(i) The gusA mRNA transcribed by the RBCS3C
promoter may be extremely stable so that levels persist
in leaves grown in the dark for 24 h. Although the gusA
mRNA may be inherently more stable than the RBCS3C
transcript, which in tomato becomes undetectable after
3 h following a shift to darkness (Sugita and Gruissem
1987), the gene and its 3¢ untranslated region including
the polyadenylation signals are identical to those in the
SRS1 promoter construct. The features of plant mRNAs
which govern their stability are not fully characterised
(Abler and Green 1996) although the nature of the 3¢
end is thought to be important (Ingelbrecht et al. 1989).
Therefore, it is thought likely that the stability of the two
SSU promoter-gusA mRNAs will be much the same.
(ii) Apple may lack certain trans-acting factors
required for the correct regulation of expression, or
they may have diverged so that they no longer recognise
cis-elements within the tomato promoter. In general, it
has been shown that light regulation mechanisms are
highly conserved through plant evolution (Terzaghi and
Cashmore 1995). However, there are precedents for
tissue-speci®c yet light-independent expression driven by
photosynthetic gene promoters in heterologous species.
Promoters isolated from photosynthetic genes of conifers can drive tissue-speci®c but light-independent
reporter gene expression in species such as tobacco
(Kojima et al. 1994) and rice (Yamamoto et al. 1994).
Conversely, the tobacco rbcS promoter drives chloroplast-dependent but light-independent gene expression
in transgenic black spruce (Gray-Mitsumune et al.
1996). It is likely however, that this pattern of lightindependent promoter function marks a fundamental
di€erence between angiosperms and conifers. In the
latter group, promoters driving genes involved in photosynthesis have been shown to function, to varying
degrees, independently of light (Alosi et al. 1990;
Yamamoto et al. 1991; Mukai et al 1992).
J.R. Gittins et al.: Transgene expression in apple driven by heterologous Rubisco SSU promoters
Other studies have shown that the promoters of
photosynthetic genes from dicot (pea, tobacco, Arabidopsis) and monocot (wheat, maize, rice) species require
di€erent classes of cis elements for normal lightregulated activity in homologous and heterologous host
plants (SchaȀner and Sheen 1991; Luan and Bogorad
1992). However, the tomato RBCS3C promoter has
previously been shown to function in a tissue-speci®c
and light-dependent manner in transgenic rice (Kyozuka
et al. 1993), so it would appear to contain elements
necessary for authentic activity in widely di€erent plant
species.
(iii) A ®nal and more likely hypothesis to explain why
the RBCS3C promoter is not strictly regulated by light
in apple is that certain cis-elements required for
authentic expression were not included in the construct
pSCV1.6RBCS3CP. The 0.8-kbp fragment of 5¢ ¯anking
sequence from the RBCS3C gene was chosen because it
includes cis-acting sequence elements such as Box II,
GT-1 and LRE thought to be involved in light-regulated
expression in tomato (Manzara et al. 1991; Carrasco
et al. 1993). In the study of Kyozuka et al. (1993) a
larger 1.5-kbp fragment of the RBCS3C promoter was
used to drive light-dependent gusA expression in transgenic rice. Therefore, some essential cis-elements may
have been omitted in the 0.8-kbp RBCS3C promoter
used in this study. Interestingly, studies in tomato have
shown that the DNA-footprint pro®le of an RBCS3C
promoter region contained within the 0.8-kbp fragment
shows no apparent di€erences in seedlings grown in the
dark compared to those grown in the light (Carrasco
et al. 1993). It is suggested that the light-regulation of
transcription is therefore likely to be the result of
modi®cation of the trans-acting factors or changes in
their interactions (Carrasco et al. 1993). An alternative
explanation is that there is a requirement for additional
sequence elements outside the 0.8-kbp fragment that we
have used here. Interactions between trans factors
binding these ``missing'' cis-elements with other factors
bound nearer to the transcription start point may be
required for authentic light-regulated expression.
Despite the fact that the RBCS3C promoter fragment
used in this study is apparently not down-regulated by
the absence of light for 24 h, there is some stimulation of
expression when plants are returned to the light after a
period in darkness. Also, expression is largely speci®c to
green tissues, and the histochemical analysis of GUS
activity demonstrates that the activity of this promoter is
restricted to chloroplast-containing photosynthetic cells.
These ®ndings indicate that there is some involvement of
light in the regulation of the 0.8-kbp RBCS3C promoter
and also that promoter regions directing expression in
photosynthetic cells may to some degree be separate
from those regulating light-responsive expression. It is
probable that there are multiple cis-acting promoter
elements involved in conferring a speci®c pattern of gene
expression, and the truncated RBCS3C promoter we
have used here appears to lack a number of the elements
which ®ne-tune expression. It is likely to be necessary to
use a larger fragment of the RBCS3C 5¢ ¯anking region
to include the complete promoter.
239
We have demonstrated that factors controlling tissuespeci®c gene expression, driven by two heterologous
SSU promoters are present and active in the woody
perennial, apple. Both promoters would be appropriate
for con®ning the expression of bene®cial transgenes to
green photosynthetic tissues. In the mature tree these
tissues are the leaves and possibly the fruit skin. An
e€ective leaf-speci®c promoter would be useful to direct
the expression of pest and disease resistance factors. For
example, leaf damage caused by insect pests may be
reduced using targeted expression of insecticidal toxins.
Also, apple scab caused by Venturia inequalis may be
countered by leaf-speci®c expression of antifungal proteins. Both of these biotechnological control measures
would help to reduce the use of harmful chemical sprays
which are required at present.
Although these data essentially address issues concerning the localisation of transgene expression, they
also demonstrate that heterologous promoters can
function in response to environmental signals, such as
light when transferred to a non-host species. Uniquely,
too, they provide the opportunity to examine the
stability of promoters that normally would cease to
exist after one year in their natural herbaceous hosts. In
perennial crops their ecacy over several years can be
monitored. Ultimately this can tell us much about the
turnover of the photosynthetic machinery and the
genetic control of function.
We acknowledge the support of Ministry of Agriculture, Fisheries
and Food (grant HH1104STF) and the Apple and Pear Research
Council. We thank Dr. Glyn Edwards of Shell Forestry, East
Malling for the binary vector pSCV1.6. We thank Dr. Wilhelm
Gruissem of UC Berkeley, Dr. Rich Meagher at the University of
Georgia and Dr. Vicky Buchanan Wollaston at HRI-Wellesbourne
for the Rubisco SSU promoters used in this study. Lastly, we
acknowledge those at East Malling who provided assistance
including Cathy Gedman, Penny Greeves, Gail Kingswell, Andy
Passey and Fiona Wilson.
References
Abler ML, Green PJ (1996) Control of mRNA stability in higher
plants. Plant Mol Biol 32: 63±78
Alosi MC, Neale DB, Kinlaw CS (1990) Expression of cab genes in
Douglas-®r is not strongly regulated by light. Plant Physiol 93:
829±832
Berry-Lowe SL, McKnight TD, Shah DM, Meagher RB (1982)
The nucleotide sequence, expression, and evolution of one
member of a multigene family encoding the small subunit of
ribulose-1,5-bisphosphate carboxylase in soybean. J Mol Appl
Genet 1: 483±498
Carrasco P, Manzara T, Gruissem W (1993) Developmental and
organ-speci®c changes in DNA-protein interactions in the
tomato rbcS3B and rbcS3C promoter regions. Plant Mol Biol
21: 1±15
FuÈtterer J, Hohn T (1996) Translation in plants: rules and
exceptions. Plant Mol Biol 32: 159±189
Goldsbrough PB, Cullis CA (1981) Characterisation of the genes
for ribosomal RNA in ¯ax. Nucleic Acids Res 9: 1301±1309
Gray-Mitsumune M, Yoo BY, Charest PJ (1996) Chloroplastdependent and light-independent expression of the tobacco rbcS
promoter ± GUS chimeric gene in black spruce. Can J For Res
26: 909±917
240
J.R. Gittins et al.: Transgene expression in apple driven by heterologous Rubisco SSU promoters
Guilley H, Dudley RK, Jonard G, Balazs E, Richards KE (1982)
Transcription of Cauli¯ower mosaic virus DNA: detection of
promoter sequences, and characterization of transcripts. Cell
30: 763±773
Hobbs SLA, Warkentin TD, Delong CMO (1993) Transgene copy
number can be positively or negatively associated with transgene expression. Plant Mol Biol 21: 17±26
Hood EE, Helmer GL, Fraley RT, Chilton MD (1986) The
hypervirulence of Agrobacterium tumefaciens A281 is encoded
in a region of pTiBo542 outside of T-DNA. J Bacteriol 168:
1291±1301
Horsch RB, Fry JE, Ho€man NL, Eicholtz D, Rogers SG, Fraley
RT (1985) A simple and generalised method of transferring
genes into plants. Science 227: 1229±1231
Ingelbrecht ILW, Herman LMF, Dekeyser RA, Van Montagu
MC, Depicker AG (1989) Di€erent 3¢ end regions strongly
in¯uence the level of gene expression in plant cells. Plant Cell
1: 671±680
James DJ, Dandekar AM (1991) Regeneration and transformation
of apple (Malus pumila Mill.). In Lindsey K (ed) Plant tissue
culture manual. Kluwer, Dordrecht, B8 pp 1±18
Je€erson RA (1987) Assaying chimeric genes in plants: the GUS
gene fusion system. Plant Mol Biol Rep 5: 387±405
Jensen RG, Bahr JT (1977) Ribulose-1,5-bisphosphate carboxylase-oxygenase. Annu Rev Plant Physiol 28: 379±400
Kojima K, Sasaki S, Yamamoto N (1994) Light-independent and
tissue-speci®c expression of a reporter gene mediated by the
pine cab-6 promoter in transgenic tobacco. Plant J 6: 591±596
Kumar A, Taylor MA, Mad Arif SA, Davies HV (1996) Potato
plants expressing antisense and sense S-adenosylmethionine
decarboxylase (SAMDC) transgenes show altered levels of
polyamines and ethylene: antisense plants display abnormal
phenotypes. Plant J 9: 147±158
Kyozuka J, McElroy D, Hayakawa T, Xie Y, Wu R, Shimamoto K
(1993) Light-regulated and cell-speci®c expression of tomato
rbcS-gusA and rice rbcS-gusA fusion genes in transgenic rice.
Plant Physiol 102: 991±1000
Luan S, Bogorad L (1992) A rice cab gene promoter contains
separate cis-acting elements that regulate expression in dicot
and monocot plants. Plant Cell 4: 971±981
Manzara T, Carrasco P, Gruissem W (1991) Developmental and
organ-speci®c changes in promoter DNA-protein interactions
in the tomato rbcS gene family. Plant Cell 3: 1305±1316
Manzara T, Carrasco P, Gruissem W (1993) Developmental and
organ-speci®c changes in DNA-protein interactions in the
tomato rbcS1, rbcS2 and rbcS3A promoter regions. Plant
Mol Biol 21: 69±88
Martin C (1996) Transcription factors and the manipulation of
plant traits. Curr Opin Biotechnol 7: 130±138
Meier I, Callan KL, Fleming AJ, Gruissem W (1995) Organspeci®c di€erential regulation of a promoter subfamily for the
ribulose-1,5-bisphosphate carboxylase/oxygenase small subunit
genes in tomato. Plant Physiol 107: 1105±1118
Meyer P, Saedler H. (1996) Homology-dependent gene silencing in
plants. Annu Rev Plant Physiol Plant Mol Biol 47: 23±48
Mukai Y, Tazaki K, Fujii T, Yamamoto N (1992) Light-independent expression of three photosynthetic genes, cab, rbcS and
rbcL, in coniferous plants. Plant Cell Physiol 33: 859±866
Quandt HJ, Broer I, Puhler A (1992) Tissue-speci®c activity and
light-dependent regulation of a soybean rbcS promoter in
transgenic tobacco plants monitored with the ®re¯y luciferase
gene. Plant Sci 82: 59±70
Robinson DJ (1996) Environmental risk assessment of releases of
transgenic plants containing virus-derived inserts. Transgen Res
5: 359±362
Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning:
a laboratory manual (2nd edn). Cold Spring Harbor Laboratory Press, New York
SchaȀner AR, Sheen J (1991) Maize rbcS promoter activity
depends on sequence elements not found in dicot rbcS
promoters. Plant Cell 3: 997±1012
Shirley BW, Ham DP, Seneco€ JF, Berry-Lowe SL, Zur¯uh LL,
Shah DM, Meagher RB (1990) Comparison of the expression of
two highly homologous members of the soybean ribulose-1,5bisphosphate carboxylase small subunit gene family. Plant Mol
Biol 14: 909±925
Sugita M, Gruissem W (1987) Developmental, organ-speci®c, and
light-dependent expression of the tomato ribulose-1,5-bisphosphate carboxylase small subunit gene family. Proc Natl Acad
Sci USA 84: 7104±7108
Sugita M, Manzara T, Pichersky E, Cashmore A, Gruissem W
(1987) Genomic organization, sequence analysis and expression
of all ®ve genes encoding the small subunit of ribulose-1,5bisphosphate carboxylase/oxygenase from tomato. Mol Gen
Genet 209: 247±256
Terzaghi WB, Cashmore AR (1995) Light regulated transcription.
Annu Rev Plant Physiol Plant Mol Biol 46: 445±474
Vancanneyt G, Schmidt R, O'Connor-Sanchez A, Willmitzer L,
Rochasosa M (1990) Construction of an intron-containing
marker gene ± splicing of the intron in transgenic plants and its
use in monitoring early events in Agrobacterium-mediated plant
transformation. Mol Gen Genet 220: 245±250
Vaucheret H, BeÂclin C, Elmayan T, Feuerbach F, Godon C, Morel
J-B, Mourrain P, Palauqui J-C, Vernhettes S (1998) Transgeneinduced gene silencing in plants. Plant J 16: 651±659
Wanner LA, Gruissem W (1991) Expression dynamics of the
tomato rbcS gene family during development. Plant Cell 3:
1289±1303
Wilkinson JE, Twell D, Lindsey K (1994) Methanol does not
speci®cally inhibit endogenous b-glucuronidase (GUS) activity.
Plant Science 97: 61±67
Yamamoto N, Mukai Y, Matsuoka M, Kano-Murakami Y,
Tanaka Y, Ohashi Y, Ozeki Y, Odani K (1991) Lightindependent expression of cab and rbcS genes in dark-grown
pine-seedlings. Plant Physiol 95: 379±383
Yamamoto N, Tada Y, Fujimura T (1994) The promoter of a
pine photosynthetic gene allows expression of a b-glucuronidase reporter gene in transgenic rice plants in a lightindependent but tissue-speci®c manner. Plant Cell Physiol 35:
773±778