Journal of Experimental Botany, Vol. 54, No. 385, pp. 1163±1174, April 2003
DOI: 10.1093/jxb/erg128
RESEARCH PAPER
Aluminium-responsive genes in sugarcane: identi®cation
and analysis of expression under oxidative stress
Derek A. Watt1
Biotechnology Department, South African Sugar Association Experiment Station, Private Bag X02,
Mount Edgecombe 4300, South Africa
Received 19 July 2002; Accepted 10 January 2003
Abstract
Suppression subtractive hybridization (SSH) technology was used to gain preliminary insights into gene
expression induced by the phytotoxic aluminium
species, Al3+, in sugarcane roots. Roots of hydroponically-grown Saccharum spp. hybrid cv. N19 were
exposed to 221 mM Al3+ at pH 4.1 for 24 h, a regime
shown to inhibit root elongation by 43%, relative to
unchallenged roots. Database comparisons revealed
that, of a subset of 50 cDNAs ostensibly up-regulated
by the metal in the root tips, 14 possessed putative
identities indicative of involvement in signalling
events and the regulation of gene expression, while
the majority (28) were of unknown function. All of the
50 cDNAs sequenced displayed signi®cant similarity
to uncharacterized plant expressed sequence tags
(ESTs), approximately half (23) of which had been
derived from other graminaceous crop species that
had been subject to a variety of stresses. Analysis of
the expression of 288 putative Al3+-inducible genic
fragments indicated higher levels of expression
under oxidative (1 mM diamide for 4 h) rather than
Al3+ stress. By deploying SSH, this study has provided an indication of the nature of genes expressed
in sugarcane roots under Al3+ stress. It is anticipated
that the information obtained will guide further
exploration of the potential for manipulation of the Al
tolerance characteristics of the crop.
Key words: Aluminium, oxidative
suppression subtractive hybridization.
stress,
sugarcane,
Introduction
The incidence and negative consequences of Al phytotoxicity on plant growth and crop production are of
1
worldwide relevance and, consequently, have been the
focus of numerous international research efforts (Taylor,
1991; Snowden and Gardner, 1993; Delhaize and Ryan,
1995; Kochian, 1995; Matsumoto, 2000). The effects of
the metal on agricultural productivity are also of concern
within the South African sugar industry where severe soil
acidi®cation has resulted from intensive sugarcane monocropping (Schroeder et al., 1994). Although current
strategies used within the industry to alleviate the negative
effects of soil acidity and by association Al phytotoxicity
have met with some success (Schumann et al., 1999), they
do not offer a sustainable solution for a number of
agronomic, economic and environmental reasons. Hence,
the exploitation and potential manipulation of Al tolerance
characteristics of sugarcane, through the application of
molecular technology, are increasingly viewed as desirable
adjuncts to existing agronomic practices. However,
fundamental to the deployment of these technologies is
knowledge of the mechanisms through which sugarcane
perceives and responds to such potentially harmful stimuli
in the rhizosphere.
Attempts to unravel the molecular mechanisms underlying the response of sugarcane to Al have been con®ned to
the general data-mining approach adopted within the
Brazilian sugarcane Expressed Sequence Tag (EST)
project, SUCEST (Drummond et al., 2001). However,
studies on other graminaceous crops (Snowden and
Gardner, 1993; Cruz-Ortega et al., 1997; Hamel et al.,
1998) and dicotyledonous species (Ezaki et al., 1995;
Richards et al., 1998) have identi®ed a number of genes
expressed speci®cally as a consequence of a de®ned Al
stress. Those investigations revealed that the metal induces
the expression of diverse genes, including several involved
in general plant stress-responsive (Snowden and Gardner,
1993; Ezaki et al., 2000), pathogenesis (Cruz-Ortega et al.,
1997; Hamel et al., 1998) and anti-oxidant (Richards et al.,
1998) pathways. In instances where such Al-induced ESTs
Fax: +27 31 5395406. E-mail: [email protected]
Journal of Experimental Botany, Vol. 54, No. 385, ã Society for Experimental Biology 2003; all rights reserved
1164 Watt
have been used as transgenes, reduced susceptibility to
both Al- and oxidative-stresses has been demonstrated
(Ezaki et al., 1999; Ezaki et al., 2000). Hence, given the
diversity and apparently indiscriminate nature of Alresponsive genes isolated to-date (Richards et al., 1998),
derivation of conclusions regarding the contributions of
these individual determinants to overall Al tolerance is
dif®cult.
Detailed investigations into plant responses to agronomically important stresses, other than Al, have revealed
that degree of tolerance may reside within variations in
signalling pathways and gene regulatory mechanisms
(Scheel and Wasternack, 2002). Hence, subsequent to the
successful isolation of numerous genes with up-regulated
expression in response to such stresses, considerable effort
has been expended on the discovery of proteins that
regulate gene transcription, which may ultimately confer
tolerance (Kirch et al., 2002; Xiong and Zhu, 2002). To
date, however, investigations into signalling and regulatory pathways elicited by Al stress have been limited,
possibly due to the challenges associated with the identi®cation of genes involved in the perception of phytotoxic
levels of Al in the rhizosphere and the consequent
transmission of the stress response. In this regard, the
advent of suppression subtractive hybridization (SSH)
technology (Diatchenko et al., 1996) has provided plant
physiologists with a powerful means to construct subtractive cDNA libraries enriched for rare transcripts, such
as those involved in signalling and the regulation of gene
expression.
Towards the ultimate goal of elucidating the molecular
responses of sugarcane to Al, SSH was used in the current
study to capture and enrich rare transcripts expressed in
root tips as a consequence of exposure to a demonstrably
phytotoxic level of the metal. Insights into the nature of
pathways operational under Al stress were gathered
through the assignment of putative identities to a subset
of these genic fragments through electronic database
homology searches. In addition, the induction speci®city
of the isolated genic fragments was assessed by analysis of
their expression under conditions of Al- and oxidative
stress. These investigations aimed at establishing the
nature of gene expression in sugarcane roots under Al
phytotoxic conditions and the information obtained will
permit more focused study of the Al tolerance characteristics of the crop.
Materials and methods
Plant material and growth conditions
Entire transverse nodal culm sections, bearing a single intact axillary
bud, of Saccharum spp. hybrid cv. N19 (N19) were planted to a
depth of 1 cm in acid-washed graded silica. To induce bud break and
subsequent plantlet growth, water was supplied twice daily and
supplemented weekly with nutrient medium (Hydroponic Nutrient
Mix, Hygrotech Seeds [PTY] Ltd, Silverton, RSA) under glasshouse
conditions (2866 °C). After shoot emergence, plantlets were
transferred to silica medium contained within 1.0 l volume pots
and supplied with water and Long Ashton nutrients (Hewitt, 1966)
by capillary action from a 20 l volume sump. Once the ®rst
population of metabolically active roots was established (approximately 5 weeks after bud break) the remaining portions of the
original culm were excised and the plantlets introduced into a
hydroponics system. Ten litre volume plastic buckets served as the
basis of this system, with aeration and agitation of the medium by
means of a diaphragm pump at approximately 0.5 l air delivered
vessel±1 min±1. The growth medium was the Long Ashton formulation (Hewitt, 1966), modi®ed to contain 2 mM NH4Cl, 0.09 mM
Fe-EDTA and 0.0033 mM CuSO4, at pH 5.5. To accommodate the
plants, four holes of 2.5 cm diameter were cut in the lid of each
vessel, through each of which a single 5-week-old plant was inserted
and supported by a 5 cm wide Neopreneâ collar. A 32 cm length of
polycarbonate tubing (internal diameter 1.1 cm, external diameter
1.2 cm) with multiple perforations was inserted through the centre of
each lid, which served to deliver air from the diaphragm pump. The
four plants within each of 12 vessels were cultured for 4 weeks and
supplied weekly with fresh nutrient medium. Twenty-four hours
prior to stress imposition, plants were supplied with fresh nutrient
medium.
Al3+ and oxidative stress: application and measurement of
effects
Challenge of roots with Al was conducted under conditions identical
to those used in the hydroponic culture of plants, except that the
nutrient medium was replaced with various concentrations of Al in
1 mM CaCl2. In the formulation of the medium for Al challenge, a
0.1 M AlCl3 stock was prepared by adding an appropriate amount of
the chemical to polished water, acidi®ed to a pH value of 3.0 with
concentrated HCl (Hamel et al., 1998). Various volumes of this
stock were added to 1 mM CaCl2 (pH 4.5) to give Al of
concentrations 0, 0.05, 0.10, 0.25, 0.50, and 1.00 mM. The ®nal
pH value of these solutions was adjusted to 4.15 with concentrated
HCl. The activity of the Al3+ ion at each concentration was
determined by means of the ion speciation programme,
MINTEQA2/PRODEFA2 (Allison et al., 1990). Plants exposed to
1 mM CaCl2 under identical conditions but in the absence of AlCl3
served as controls. Before exposure to the Al3+-containing media,
roots were rinsed three times with deionized water and blotted dry to
remove traces of nutrient medium, after which the distal 10 mm
region of each root was demarcated with indelible ink. Any increase
in length of the tip of each root was measured after 24, 48, and 72 h
and the average increase in root length for each treatment expressed
as a relative root growth inhibition index (%RGI), calculated
according to the equation cited by Hamel et al. (1998). The
signi®cance of the effects of the Al3+ treatments on retardation of
root elongation was assessed by means of an unpaired student t-test
(SigmaPlotâ version 4.0, Jandel Scienti®c). After measurement of
root growth, the organs from three of the plants subjected to each
treatment were separated and dried at 180 °C for 24 h. The Al
content of roots, stems and leaves was determined by catechol violet
dye colourimetry, following acid digestion (Wilson, 1984).
Diamide [(CH3)2NCON=NCON(CH3)2] (Sigma), a thiol-oxidizing compound (Kosower et al., 1969), was selected as agent for the
imposition of oxidative stress on roots (Ezaki et al., 2000). Roots
were exposed to 1 mM diamide for 4, 8, and 12 h, under conditions
used for the Al challenge. After the elapse of each exposure period,
roots were rinsed three times with deionized water, blotted dry and
the distal portion (approximately 10 mm) of each root excised, and
immediately frozen to ±196 °C in liquid nitrogen. Total protein was
extracted (Ibrahim and Cavia, 1975) from a portion of the harvested
tips and quanti®ed colourimetrically (Bradford, 1976). The effect of
diamide on the roots was assessed through the determination of total
Al-responsive genes in sugarcane 1165
reduced glutathione levels in a further portion of the root tips (Baker
et al., 1990). The remaining root tips (approximately 2 g) were
stored at ±80 °C until required.
cDNA synthesis and subtraction
The protocol described by Carson and Botha (2000) was followed in
the extraction of total RNA from approximately 2 g of frozen (±80
°C) root tips harvested from control plants and those challenged by
Al3+ at a concentration and for a duration shown to have maximum
inhibitory effect on root elongation. The total RNA preparations
were further puri®ed by selective binding and elution from silicagel-based membranes (RNeasy Plant Mini Kit, Qiagen) prior to poly
A+ RNA (mRNA) isolation by means of Dynabeadsâ Oligo (dT)25
(Dynalâ), with ®nal elution into a 10 ml volume. To increase the
representation of 5¢ ends within the ®nal double-stranded (ds) cDNA
population, reverse transcription of mRNA and second strand
synthesis reactions were facilitated by means of a SMARTÔ PCR
cDNA Synthesis Kit (Clontech). A suppression subtractive hybridization (SSH) approach (Diatchenko et al., 1996) (PCR-SelectÔ
cDNA Subtraction Kit, Clontech) was adopted to isolate fragments
of genes up-regulated in root tips in response to Al3+ challenge.
During subtraction, ds cDNA populations derived from control and
Al3+-challenged roots tips served as driver and tester cDNA,
respectively. After cloning (pGEMâ-T Easy Vector System,
Promega), 288 E. coli (strain JM 109) colonies containing
recombinant plasmid vector DNA were selected randomly for
further analysis.
Array printing, querying and analysis
The cDNA inserts, which were to serve as probes in the reverse
northern hybridization analyses, were ampli®ed directly from each
of the 288 bacterial clones by means of the PCR and vector-speci®c
primers (pGEMâ-T Easy Vector System, Promega). To determine
approximate insert size and verify the speci®city and ef®ciency of
PCR ampli®cation, aliquots (2 ml) of the ampli®ed PCR products
were fractionated by means of standard agarose gel electrophoresis.
Following this veri®cation, amplicons were denatured by the
addition of NaOH to a ®nal concentration of 0.2 N, heating to 65
°C for 30 min (Cairney et al., 1999) and quenching on ice. Aliquots
(0.2 ml) of the denatured PCR products, representing approximately
20 ng of probe DNA, were transferred in triplicate to 150±100 mm
portions of positively-charged nylon membrane (HybondÔ-N+,
Amersham Pharmacia Biotech) with a 96 pin replicator (V&P
Scienti®c, San Diego, Ca). The ampli®ed cDNAs were deposited in a
333 format with a single blank row and column intervening
amongst each array unit. Once printed, the membranes were dried
under a stream of ®ltered air for 2 h and the denatured probe cDNA
cross-linked to the membrane with short-wavelength ultra-violet
radiation (120 000 mJ cm±2 for 2 min). The membranes were stored
desiccated at room temperature until required.
The mRNA populations that served as template for the production
of driver and tester cDNA for SSH reactions, as well as those
isolated from diamide challenged and control root tips, were used to
synthesize target total cDNA populations for array querying. The
target cDNA was synthesized and labelled with [a-33P] dATP (AEC
Amersham) by means of the Advantage PCR system (Clontech),
using single-strand (ss) cDNA generated by SMART technology
(SMARTÔ PCR cDNA Synthesis Kit, Clontech), according to the
principle described by Cairney et al. (1999). The labelled target was
puri®ed from unincorporated dNTPs with NucTrapâ Probe
Puri®cation Columns (Stratagene), and heated to 100 °C for 5
min, with subsequent quenching to 0 °C, immediately prior to array
querying.
The cDNA probes on the arrays and the 33P-labelled target cDNA
populations were allowed to hybridize overnight under conditions
identical to those described by Carson et al. (2002). Used for array
querying were target cDNA populations derived from Al3+- and
diamide-treated roots, as well those from prepared from the roots of
plants in two experimental controls. Each querying event was
repeated twice. After array washing (Carson et al., 2002), the
hybridization patterns were captured through exposure of the
membranes for 18 h to high-resolution phosphor screens with
subsequent laser scanning (CycloneÔ Storage Phosphor System,
Packard Bioscience). Array images were analysed using
QuantArrayâ Microarray Analysis Software (Version 3.0, Packard
Bioscience), which permitted the quanti®cation of comparative
intensity of hybridization amongst probes and target cDNAs derived
from root tips in the aluminium and diamide treatments, relative to
their respective controls. The software also generated quality
measures for each querying event, including spot diameter and
intensity, background and signal to noise ratios. Using a threshold of
20% above local backgrounds, poorly hybridized spots were
eliminated from the data set, as were cases in which the relative
hybridization intensities for triplicate spots varied from each other
by more that 30% (Kurth et al., 2002).
Northern and cDNA Southern hybridization analysis
Size fractionation of total RNA (12 mg) under denaturing conditions
and subsequent transfer and immobilization onto nylon support
membranes were conducted essentially according to the method
developed by Ingelbrecht et al. (1998). The only deviation involved
the inclusion of 0.45 M formaldehyde in the electrophoretic tank
buffer to avoid the formation of formaldehyde gradients during
fractionation (Tsang et al., 1993). Labelling of the probe cDNA with
[a-32P] dCTP (AEC Amersham), hybridization and subsequent
visualization procedures were as described by Carson et al. (2002).
In certain instances, cDNA Southern hybridization was used as an
alternative to northern analysis. In these cases, the method of Jaakola
et al. (2001) was used, although for the purpose of this study, the
target cDNA population was synthesized by means of the Advantage
PCR system (Clontech), using ss cDNA generated by a SMARTÔ
PCR cDNA Synthesis Kit (Clontech).
DNA sequencing and sequence data analysis
Selected inserts within puri®ed (QIAprepâ Spin Miniprep Kit,
Qiagen) recombinant plasmid DNA were sequenced by dye
terminator cycle chemistry (BigDye Terminator Cycle Sequencing
Kit, Applied Biosystems) and automated capillary electrophoresis
(ABI Prism 310 Genetic Analyser, Applied Biosystems). The
universal reverse primer was used to generate single-pass partial
sequences. After removal of vector and ambiguous sequences
(Sequence Navigator, Applied Biosystems), comparative sequence
analysis was conducted with the BLASTx and BLASTn algorithms
(Altschul et al., 1997) against the National Centre for
Biotechnological Information (NCBI) non-redundant protein and
nucleotide Expressed Sequence Tag (dbEST) databases, respectively. Matches were considered signi®cant when the E values were
below 10±5 and the PAM120 similarity scores were above 80
(Newman et al., 1994).
Results
Assessment of sensitivity to Al3+ stress.
To determine the level and duration of exposure for
maximal phytotoxic effect of the octahedral hexahydrate
species of Al (Al(H2O)63+), commonly abbreviated as Al3+
(Kochian, 1995), sugarcane roots were exposed to concentrations of AlCl3 ranging from 0 to 1 mM at pH 4.15 for
1166 Watt
1, 2 and 3 d. As sugarcane is reportedly more tolerant of Al
than other graminaceous crops (Hetherington et al., 1986),
the upper limits of concentration and exposure period used
in this investigation were extended beyond those reported
for other studies (Hamel et al., 1998; Ezaki et al., 2000).
The activity of Al3+ in the media was calculated by means
of the ion speciation program, MINTEQA2/PRODEFA2,
which indicated that the phytotoxic Al3+ species was
present at levels only 11% lower than the molar concentrations of the metal (results not shown). At the lowest
concentration tested (45 mM), Al3+ stimulated relative root
elongation by between 6% and 14%, an effect that was
consistent over the entire duration of the assay (Fig. 1A).
However, at 88 mM Al3+, the stimulatory effect was
reversed and root elongation was inhibited by 10-15%,
when compared to the unexposed plants. This symptom of
phytotoxicity was exacerbated by higher concentrations of
the metal, reaching an apparent maximum of 43% relative
root growth inhibition at 221 mM Al3+, a trend that was
maintained over the three exposure periods assessed. Over
the entire higher concentration range (221±897 mM Al3+),
relative root elongation was reduced on average by
between 36±46%.
The Al content of roots, stems and leaves was
determined in plants that had been challenged with the
selected concentration range of Al3+ for 24 h. This shorter
challenge period was chosen because the effect of the
metal on root elongation was shown not to increase
signi®cantly upon further exposure (Fig. 1A). Under these
conditions, Al accumulated to levels signi®cantly higher in
the roots than in the aerial parts of the plant, regardless of
the concentration of the metal within the challenge
medium (Fig. 1B). After exposure to the metal, roots
contained between 230 and 322 mmol Al g±1 dry weight,
while the aerial portions of the plants contained approximately 8-fold less (14±51 mmol g±1 dry weight). However,
despite measures to remove free Al3+ from the apoplasm, it
is possible that metal adsorbed to the cell walls may have
accounted for a proportion of the levels measured in the
roots.
Identi®cation of Al3+- responsive genes
Genes responsive to Al3+ challenge were captured through
the construction of a subtractive cDNA library from
mRNA isolated from unexposed root tips and those
exposed to 221 mM Al3+ for 24 h. An SSH approach was
adopted due the capacity of the technology to enrich
speci®cally for rare transcripts, thereby substantially
reducing the number of genic fragments required to obtain
a representation of the changes in gene expression
occurring in response to external and internal stimuli.
Furthermore, as the intent of this study was to catalogue
genes up-regulated by exposure to the metal, only forward
subtractions were performed, in which cDNA derived from
Fig. 1. Relative root elongation and Al3+ accumulation in sugarcane
as a consequence of Al3+ challenge. Roots of hydroponically-grown
plants were exposed to various concentrations of Al3+ in 1 mM CaCl2
at a pH value of 4.15 for 24 h, after which the effects of the challenge
on (A) root elongation and (B) levels of the metal in roots, stems and
leaves were assessed. Root growth inhibition was calculated as a
percentage of the elongation of the terminal 1 cm portion of roots
exposed to Al3+ relative to that of unchallenged root tips (mean 6SE,
n=16). Values not signi®cantly different (0.01 signi®cance level) share
the same alphabetical character while those signi®cantly different do
not.
control and challenged root tips served as driver and tester
populations, respectively.
Assessment of subtraction ef®ciency by means of
reverse northern hybridization analysis of the 288 fragments isolated revealed that 182 were up-regulated by Al3+
(results not shown). Results of such array analyses further
served as the basis for the selection of 50 cDNAs with the
most obvious Al3+-inducible expression patterns for
further characterization. This subset of ostensibly Alresponsive cDNAs was subjected to sequence determinations with subsequent homology searches against the
NCBI non-redundant protein and Expressed Sequence Tag
(dbEST) databases (Table 1). Over half (28 out of 50) of
the cDNAs sequenced were of unknown function, either
demonstrating no similarity to genes lodged in the data
base or being homologous to genes encoding unknown or
hypothetical proteins. However, analysis of the putative
identities assigned to the remaining 22 sequences provided
additional evidence of the potential of the SSH technology
to enrich for rare transcripts, in that the majority (13 out of
22) appeared to be direct or indirect participants in the
Al-responsive genes in sugarcane 1167
regulation of gene expression (clones 1E9, 1F9, 1G7,
1G12, 2C7, and 3C2) and signalling (clones 1C1, 1E10,
1H12, 2B11, 2G4, 2G11, and 3G8), rather than basic
metabolic events (clones 1D5, 1D12, 1F7, and 2C9)
(Table 1). Interestingly, four sequences displayed similarity to genes involved in vesicular traf®cking and membrane transport (clones 2A2, 2G9, 3A1, and 3G10).
Although caution should be exercized in their interpretation, such functional categorizations may serve to illustrate general trends in gene expression under particular
circumstances. In this study, it is also of note that only a
limited degree of redundancy (two out of 50) (clones 1F7
and 2C9) was apparent amongst the genic fragments
isolated, suggesting that effective normalization between
abundant and rare transcripts had occurred during subtraction. All of the cDNAs sequenced displayed signi®cant
similarity to sequences contained within the dbEST
(Table 1). A number (18 out of 50) of these ESTs were
derived from Sorghum spp. that had been subject to biotic
and a variety of abiotic stresses. Further identity matches
(®ve out of 50) were obtained to cDNAs induced by
disease or cold in other graminaceous crop species,
namely, Oryza sativa, Triticum aestivum and Zea mays.
To verify expression inducibility, three of the isolated
ESTs (clones 1F9, 1E10 and 1H12) were subjected to
further hybridization analysis. Due to the propensity of
SSH to capture rare transcripts, either conventional
northern or cDNA Southern hybridization was deployed,
the latter being used when increased detection sensitivity
was demanded. For the three cDNAs assessed, an Al3+responsive expression pattern was con®rmed (Fig. 2).
These data, together with those obtained regarding the
ef®ciency of SSH, indicate that the technology effectively
captured a representation of Al3+-responsive genes.
Although this investigation ful®lled the objective of
providing an insight into the nature of genes responsive
to the metal in sugarcane roots, characterization of an
increased number of genic fragments would be required for
a detailed dissection of the molecular mechanisms operating under the stress.
Determination of responses common to Al3+ and
oxidative stress
Determination of possible pluralities in induction of
expression of the isolated genic fragments required the
imposition of the additional stress under conditions similar
to those used during Al challenge. To this end, Al3+ in the
challenge medium was replaced with 1 mM diamide, a
compound known to impose severe oxidative stress on
biological material. The effect of this superoxide generator
on sugarcane roots was assessed through the monitoring of
levels of reduced glutathione (GSH) after 4, 8 and 24 h of
exposure (Table 2). The GSH content of the tips of
oxidatively stressed roots increased from an initial concentration of 4.1 to 13.8 mmol g±1 total protein after 4 h of
exposure, with the latter level being approximately 46-fold
that measured in the control (0.3 mmol g±1 total protein) for
the same period. After this sharp rise, the levels of GSH
subsequently declined to 6.5 and 6.1 mmol g±1 total protein
after 8 and 24 h of exposure to diamide, respectively. The
concentration of GSH in the root tips of control (6.0 mmol
g±1 total protein) and stressed (6.1 mmol g±1 total protein)
root tips reached similar levels after 24 h, although these
remained elevated relative to the initial level of 4.1 mmol
per g±1 total protein (Table 2). The observed increase in
GSH content of roots upon exposure to diamide is a pattern
not without precedent, in that increased rates of GSH
synthesis have been reported in plants subjected to either
oxidative (Xiang and Oliver, 1998) or low, non-freezing
temperatures (Kocsy et al., 2000) and are believed to ful®l
a detoxi®cation function. In such cases, this response has
been attributed to the induction of adenosine 5¢-phosphate
reductase and g-glutamylcysteine synthetase, which are the
key enzymes of cysteine and GSH synthesis (Kocsy et al.,
2001). After the initial rise in GSH content of the diamidetreated roots, a decline was observed until, after 24 h,
levels in the challenged and control roots were similar (c.
6.0 mmol g±1 total protein) (Table 2). Furthermore, a
substantial ¯uctuation in GSH levels in the control roots
was observed over the 24 h diamide challenge period,
possibly indicating a diurnal response (Table 2). Circadian
rhythms in mRNA levels have been reported for genes
encoding several enzymes involved in cysteine synthesis,
cysteine being a precursor for glutathione synthesis
(Leustek, 2002). Hence, the observed response of root
GSH levels to diamide challenge, during which maximal
levels were reached after 4 h of exposure, and the similar
levels after 24 h in both challenged and control roots
precluded the choice of 24 h exposure period, as was used
for Al3+ challenge.
As the perturbation of root metabolism by 1 mM
diamide, as re¯ected by variations in GSH content, was
most apparent after 4 h of exposure (Table 2), it was under
this experimental regime that the expression of the putative
Al3+-responsive genic fragments was examined. To
accomplish this, arrays of the 288 genic fragments,
enriched for Al3+-induced sequences through SSH, were
queried with 33P-labelled total cDNA populations derived
from root tips of sugarcane plants that had been subject to
conditions imposing quanti®ed levels Al3+ (Fig. 1A) and
oxidative (Table 2) stress. The hybridization patterns
resulting from these query events were analysed with
software (QuantArrayâ Microarray Analysis Software)
that facilitated comparison of relative hybridization intensity to each genic fragment amongst the two target and
control total cDNA populations.
When arrays were queried with total cDNA populations
originating from unstressed roots, a hybridization signal
was detected for only 78 of the 288 (27%) cDNA probes on
the membrane (Fig. 3A), con®rming initial estimates of the
1168 Watt
Table 1. Putative identities and characteristics of selected sequences expressed in the root tips of Saccharum spp. hybrid cv. N19
as a consequence of challenge by 250 mM Al3+ at pH 4.1 for 24 h
Clone
cDNA
BLASTx
reference size (bp)
Corresponding
or related
protein
sequence
1A3
146
1A11
173
1B2
366
1B6
212
1B12
355
1C1
461
1D3
329
1D5
367
1D12
168
1E2
354
1E3
172
1F3
358
1E9
549
1E10
384
1F7
370
1F9
373
1G2
346
1G7
377
1G12
556
1H5
477
1H6
381
1H12
207
2A2
479
2A4
389
2A12
424
2B3
288
2B4
389
No signi®cant
similarity
No signi®cant
similarity
No signi®cant
similarity
No signi®cant
similarity
No signi®cant
similarity
Digitalis lanata
Acyl-CoA
binding protein
No signi®cant
similarity
Z. mays adenine
nucleotide
translocator
Capsicum annuum
TMV-induced
protein
No signi®cant
similarity
Z. mays cDNA
clone MEST41-B08
No signi®cant
similarity
Arabidopsis thaliana
probable RNA
binding protein
S. bicolor serine/
threonine kinase
Nicotiana glutinosa
60S ribosomal
protein
Z. mays probable
histone deacetylase
A. thaliana
hypothetical
protein F9D16.100
Z. mays probable
histone deacetylase
Homo sapiens
SWI/SNF gene
No signi®cant
similarity
No signi®cant
similarity
O. sativa
RAS-related
protein RGP1
Z. mays kinesin
heavy chain (KIN15)
No signi®cant
similarity
No signi®cant
similarity
No signi®cant
similarity
No signi®cant
similarity
BLASTn (dbEST)
E-value
Sequence
Accession
identity (%) number
Characteristics
of cDNA
E-value
Sequence Accession
identity number
(%)
±
±
±
9310±6
91
BG836703
±
±
±
Zea mays,
pathogen-induced
Sorghum bicolor,
water-stressed
S. bicolor,
pathogen-induced
S. bicolor,
water-stressed
Triticum aestivum,
cold-induced
S. bicolor,
pathogen-induced
1310±9
97
BE638185
88
BE594873
1310±11 86
BE592729
±11
88
BF200586
1310±163 92
BE367649
Z. mays,
stressed root
S. bicolor,
pathogen-induced
5310±51 104
AI855319
1310±139 96
BE600734
3310±20 87
BE594885
7310±7
AW070097
±
±
±
±
±
±
±
±
±
3310±36 86
AJ249833
±
±
±
1310±28 72
X15711
2310±18 60
AF242731
±
±
±
2310±22 98
±
±
4310±31 92
S. bicolor,
pathogen-induced
Oryza sativa,
pathogen-induced
BG842882 Z. mays, inbred
tassel
±
S. bicolor,
dark-grown seedling
T49019
S. bicolor,
light-grown seedling
±29
5310
1310
60
9310±60 93
AI939876
±64
BE358013
1310
127
1310±168 96
AW287250
1310±40 93
BE599698
8310±16 88
S. bicolor,
pathogen-induced
NGU23784 S. bicolor,
water-stressed
1310±122 95
AW677661
4310±40 90
P56521
1310±129 92
BG320015
1310±30 51
±30
AP002482
92
T05595
2310±24 81
P56521
6310
2310
±
±
±19
90
AF109733
±
±
±
±
4310±16 83
P25766
1310±169 601
AF272759
±
±
±
±
±
±
±
±
±
±
±
±
Z. mays,
cold-stressed
S. bicolor,
water-stressed
±142
1310
89
AW745310
S. bicolor,
pathogen-induced
O. sativa, panicle
at ¯owering stage
O. sativa, root
1310±116 94
1310±12 83
AU032267
Z. mays, inbred
tassel
S. bicolor,
pathogen-induced
±60
93
AI939876
2310±77 94
BE595980
1310±167 595
BU098549
3310±45 188
BQ282541
4310±78 297
BM329690
1310±22 113
BI992021
3310±76 291
BM348090
Z. mays, tassel
primordium
Triticum aestivum,
cold-stress induced
S. bicolor,
pathogen-induced
Z. mays, Unigene II
Z. mays, cDNA
clone MEST286-D05 3¢
±52
6310
9310
89
BE599391
C72606
Al-responsive genes in sugarcane 1169
Table 1. Continued
Clone
cDNA
BLASTx
reference size (bp)
Corresponding
or related
protein
sequence
BLASTn (dbEST)
E-value
Sequence
Accession
identity (%) number
2B11
486
Z. mays calnexin
3310±10 61
2C7
316
2310±35 147
2C9
313
2E2
216
2E3
281
2E7
304
2F3
293
2F10
420
2G4
500
2G6
225
O. sativa putative
G-box binding
protein
O. sativa putative
60S ribosomal
protein L37a
No signi®cant
similarity
No signi®cant
similarity
No signi®cant
similarity
No signi®cant
similarity
No signi®cant
similarity
O. sativa
24-methylene
lophenol C24(1)
methyltransferase
No signi®cant
similarity
A. thaliana putative
ABC transporter
A. thaliana putative
amine oxidase
2G9
305
2G11
531
2H5
340
No signi®cant
similarity
Z. mays actin
related protein
A. thaliana
RING-H2 ®nger
protein RHF1a
No signi®cant
similarity
O. sativa
unknown protein
zwh13.1
A. thaliana similar
to unknown protein
emb|CAB89322
A. thaliana
GTP-binding
protein (RAB1Y)
O. sativa genomic
DNA, chromosome 4
Characteristics
of cDNA
E-value
Sequence Accession
identity number
(%)
S. propinquum,
¯oral-induced
meristem 1 (FM1)
AAL76334 S. bicolor,
pathogen-induced 1 (PI1)
0.0
781
T03251
BF585670
BE363888
2310±7
55
AP003335
S. bicolor,
water-stressed 1 (WS1)
1310±68 266
BE363812
±
±
±
1310±15 90
AW054220
±
±
±
2310±8
66
BE599251
±
±
±
Z. mays,
root cDNA
S. bicolor,
pathogen-induced 1 (PI1)
Z. mays, mixed tissue
8310±11 74
BM381305
±
±
±
2310±61 289
AW922283
±
±
±
4310±8
D22024
5310±5
48
AAC34989 S. bicolor,
pathogen-induced (PI1)
1310±115 420
BE597306
±
±
±
1310±64 252
AW745885
±92
345
BG739317
2310±63 250
AL606652
4310±57 115
BI1351176
±127
BG319679
1310
±14
79
4310±27 297
±
7310
±
±19
3A1
421
3C2
370
3D9
116
3E10
421
3F5
350
3G8
487
3G9
514
3G10
501
A. thaliana g-soluble
5310±23 90
NSF attachment protein
3H3
502
3H8
348
A. thaliana
hypothetical protein
No signi®cant
similarity
77
S. bicolor,
dark-grown 1 (DG1)
O. sativa, callus
S. bicolor
water-stressed 1 (WS1)
AY086511 S. bicolor
embryo 1 (EM1)
AC006224 O. sativa
genomic DNA,
chromosome 4
±
S. bicolor
immature panicle 1 (IP1)
AJ223200 Z. mays, cold-stressed
2310
1310
66
460
3310±12 71
NP_193158 S. bicolor,
immature panicle 1 (IP1)
±
S. bicolor
1310±23 115
water-stressed 1 (WS1)
CAB55397 Z. mays
4310±85 321
mRNA sequence CL36095_1
±
2310±17 89
1310±22 105
9310±32 132
2310±17 89
5310±38 125
±
±
4310±47 115
±
BI351176
AW745885
AY109821
BAB10214 S. propinquum
¯oral-induced
meristem 1 (FM1)
AAF79660 S. bicolor
immature pannicle 1 (IP1)
8310±15 88
BF481893
1310±168 599
BI211750
AL117264
1310±107 394
AY105682
1310±162 577
AY109821
0.0
BG102656
Z. mays
PCO085208 mRNA
sequence
AF177990 Z. mays
CL36095_1 mRNA
sequence
NP_176310 S. propinquum
Rhizome2 (RHIZ2)
±
Z. mays Unigene II
718
2310±18 100
BI993147
1170 Watt
Fig. 2. Al3+-responsive expression of selected cDNAs. Con®rmation of expression patterns was facilitated by means of northern (A) and cDNA
Southern hybridization (B, C) analyses. The lowest panel (D), representing rRNA, serves to illustrate equal loading of 12 mg total RNA for
northern analysis (A). Size fractionated total cDNA populations (1 mg) (SMARTÔ PCR cDNA Synthesis Kit, Clontech) were used as alternatives
to total RNA in cDNA Southern analyses (B, C).
Table 2. Effect of diamide on reduced glutathione (GSH)
levels in sugarcane roots
Levels of GSH were determined in root tips exposed to 1 mM of
diamide for 4, 8 and 24 h (n=3, 6SE).
Duration of exposure (h)
0
4
8
24
GSH concentration
(mmol g±1 total protein)
0 mM diamide
1 mM diamide
4.160.5
0.360.1
2.260.9
6.060.5
4.160.5
13.861.2
6.560.8
6.160.8
subtraction ef®ciency of SSH. Of the remaining 210 probes
to which species within the control total cDNA populations
did not hybridize, 92 (32%) and 6 (2%) emitted signal
exclusively upon querying with total cDNA populations
derived from diamide and Al3+ challenged root tips,
respectively. By contrast, hybridization of species common to both the Al3+- and oxidative-stress derived targets
occurred to 29% (83 out of 288) of the probes (Fig. 3A).
Considerable variation in relative hybridization intensity
was detected for these apparently dual Al3+- and oxidativestress inducible genic fragments (Fig. 3B). When the data
were expressed as the proportion of the total hybridization
signal contributed by the Al-challenge derived target
relative to that by the diamide target, it was apparent that
expression of the putative Al-inducible genes was generally higher under conditions of oxidative than Al3+ stress.
In 34% (28 out of 83) of such cases, expression under
conditions of Al3+ stress, relative to that detected under
oxidative stress, was very low (0.0±0.2) (Fig. 3B).
Similarly, for the remainder of the genic fragments within
this category, 28% (23 out of 83) and 38% (32 out of 83)
displayed low (0.2±0.4) to approximately equivalent (0.4±
0.6) relative expression levels, respectively, under Al3+
stress. In no instances did expression under Al3+ stress
substantially exceed that under oxidative stress.
Furthermore, the results of these reverse northern hybridization studies indicated that the three ESTs with
con®rmed Al3+-induced expression patterns (1F9, 1E10
and 1H12) (Fig. 2) were responsive to both Al3+ and
oxidative stress, with one (1F9) demonstrating substantially higher expression under the latter.
Discussion
Sugarcane is susceptible to the negative effects of
Al3+
The two primary progenitor species to modern sugarcane
cultivars, namely, S. of®cinarum and S. spontaneum, are
reported to have different degrees of tolerance to Al3+, with
the latter being the more susceptible (Landell, 1989).
However, commercial genotypes (Saccharum spp hybrids)
Al-responsive genes in sugarcane 1171
Fig. 3. Comparative analysis of relative expression levels of Al3+-responsive genes under Al3+ and oxidative stress. Comparison of the relative
hybridization signal strength, relative to the controls, for each probe after querying with each of the two target total cDNA populations (+Al3+ and
+diamide) was facilitated by means of QuantArrayâ MicroArray Analysis Software (Version 3.0, Packard Biosciences). (A) The inner pie chart
indicates the percentage of presumed Al3+-responsive clones to which species within the total cDNA population derived from unstressed root tips
did (+C) or did not hybridize (±C). The outer circumference panel details the percentage of probes hybridizing to cDNA species within each of the
three total populations: diamide- (D only) or Al3+-challenge-derived (Al only), or a combination thereof (Al+D; C+D; C+Al; C+Al+D).
(B) Proportion contributed to total signal by hybridization of species within Al3+- and diamide-derived total cDNA populations for clones showing
up-regulated expression in response to both Al3+ and oxidative stress. Proportions were calculated by dividing hybridization signal contributed by
Al3+ by the total signal value.
are generally regarded as tolerant of Al (Hetherington
et al., 1986): a phenotype that may have been inadvertently
selected in the extensive breeding that has culminated in
modern cultivars (Drummond et al., 2001). Nevertheless,
despite this apparently high degree of tolerance, the extent
and severity of soil acidi®cation that arises from sugarcane
husbandry, particularly on sandy soils under intensive
cultivation (Schroeder et al., 1994), suggests that even
slight susceptibility to the metal may result in perceptible
economic losses. In fact, exposure to 221 mM Al3+ resulted
in a relative inhibition of root elongation by approximately
43% in N19 (Fig. 1A), a cultivar rated within the South
African sugar industry as tolerant of the metal and hence,
widely grown on acid soils. On sandy, acidic soils within
the industry, Al concentrations of between 2 and 5 mM
have been reported (Turner et al., 1992), although the
proportion of phytotoxic species prevailing under such
conditions is unclear. Hence, given the observed susceptibility of a tolerant cultivar to Al3+ (Fig. 1A) and the
severity of soil acidi®cation within the industry, it is likely
that Al phytotoxicity accounts for substantial yield losses.
When compared to Al3+ dose±response studies conducted on other plant species (Hamel et al., 1998; Ezaki
et al., 2000), the results of this investigation revealed
notable differences and similarities amongst wheat,
Arabidopsis and sugarcane. A similar incapacity of Al3+
to curtail root elongation completely (Fig. 1A) has been
reported for wheat (Hamel et al., 1998). In that study, a
maximum RGI of 70% was observed above threshold
concentrations of 50 mM and 500 mM Al3+ for susceptible
(cv. Frederick) and tolerant (cv. Atlas-66) wheat geno-
types, respectively. For both the current investigation and
that of Hamel et al. (1998), it is unlikely that the observed
plateaus in RGI were due to saturation in the chemical
availability of the phytotoxic Al3+ species above the
threshold concentrations, as the challenge media were
formulated to deliver speci®c levels of the phytotoxic Al3+
ion according to the predictions of ion speciation software
(MINTEQA2/PRODEFA2) (Allison et al., 1990). Hence,
it is reasonable to assume that the persistent root elongation at high Al3+ does not re¯ect an experimental limitation
but rather a capacity to tolerate the metal. In this regard,
recent work conducted on wheat near-isogenic lines
demonstrated that genetic variation in Al tolerance resides
at multiple loci that segregate independently (Tang et al.,
2002). Consequently, allelic inheritance at more than one
locus may contribute to variations in tolerance, including
low-level tolerance in genotypes characterized as susceptible. In contrast to sugarcane (Fig. 1A), a complete
cessation of root growth at 800 mM Al3+ was observed in
Arabidopsis (Ezaki et al., 2000). Despite this difference,
the performance of an Al-tolerant Arabidopis ecotype
(Ler-0) (Ezaki et al., 2000) and sugarcane cv. N19
(Fig. 1A) were similar in the lower Al3+ concentration
range of 200±221 mM, where inhibition was between 40%
and 50%. Thus, given the apparent sensitivity of this
reportedly tolerant sugarcane cultivar to the negative
effects of Al3+ and the potential diversity of tolerance
characteristics within sugarcane germplasm (Landell,
1989), the manipulation of Al3+ tolerance through the
application of molecular technologies would appear to be a
viable strategy.
1172 Watt
Nature of Al3+-responsive gene expression in
sugarcane
In the light of the increasing body of evidence suggesting
that the degree of plant tolerance to abiotic and biotic stress
resides in variations within signalling events and the
regulation of gene expression (Scheel and Wasternack,
2002), it seems likely that the same may hold true for Al3+induced stress. Thus, the success of attempts to manipulate
the genetically complex trait of Al3+ tolerance (Aniol and
Gustafson, 1984; Berzonsky, 1992) may depend on the
availability of comprehensive information regarding the
way in which plants perceive and respond to harmful levels
of the metal in the rhizosphere.
Previous attempts to identify Al3+-induced genes in
roots have been successful, in that several research groups
have isolated and characterized cDNAs that show upregulated expression in response to exposure to the metal
(Snowden and Gardner, 1993; Snowden et al., 1995; CruzOrtega et al., 1997; Hamel et al., 1998). However, due to
the objectives of those studies and approaches used, rare
transcripts, such as those participating in cell signalling
and the regulation of gene expression, were not targeted.
Consequently, in an attempt to address this apparent gap in
knowledge, the current investigation deployed SSH to
capture and subsequently enrich such rare transcripts, with
a view to isolating those involved in the perception and
transmission of Al3+-induced stress signals. Examination
of the putative identities and characteristics assigned to 21
of the cDNAs synthesized from the isolated transcripts
supports the general success of the approach (Table 1) in
that, of these sequences, 13 were assigned putative
functions associated either directly or indirectly with cell
regulatory and signalling events. However, the majority of
the isolated cDNAs (28 out of 50) demonstrated no
homology to sequences lodged in the international electronic data bases, against which the searches were
conducted and may represent either novel or unique
genes involved in the response of sugarcane to the imposed
stress. Also of particular note were the strong associations
observed amongst the Al3+-responsive cDNAs captured
from sugarcane root tips and ESTs derived from stressed
maize, rice and sorghum (Table 1), supporting prior
evidence of substantial overlap in the way in which plants
perceive and respond to diverse abiotic stressors (Snowden
et al., 1995; Ezaki et al., 2000).
Apparent commonalities exist between Al3+- and
oxidative-stress induced gene expression in sugarcane
The existence of a relationship between Al3+ phytotoxicity
and oxidative stress has been inferred from the identity of
genes up-regulated in response to challenge by the metal
(Ezaki et al., 1996; Hamel et al., 1998; Richards et al.,
1998). This link may emanate from the capacity of Al3+ to
facilitate an Fe-mediated free radical chain reaction at the
plasmalemma (Gutteridge et al., 1985). Hence, to assess
whether the transcripts captured by SSH in this study were
responsive to oxidative stress, comparative expression
analysis was conducted under conditions of Al3+ and
oxidative stress. According to scienti®c convention, the
physiological effects of Al3+ and the oxidative agent,
diamide, on roots were assessed through quanti®cation of
root elongation (Fig. 1A) and GSH content (Table 2),
respectively. Challenge periods of 24 h and 4 h facilitated
the detection of signi®cant effects of Al3+ and diamide,
respectively, and, hence, roots subjected to these conditions served as material for reverse northern hybridization
analyses.
Of the transcripts not detectable in unstressed root tips,
the vast majority (175 out of 210: 83%) displayed higher
expression in response to oxidative than Al3+ stress, with
53% (92 out of 175) of these emitting undetectable
hybridization signal upon querying with total cDNA
populations derived from Al3+-challenged root tips
(Fig. 3). This apparent anomaly within the expression
pro®les may have resulted from differences between the
severity of stress imposed on roots by Al3+ and diamide, as
the stress regimes employed were selected according to
detectable effects of the stressors on the roots rather than
on gene expression data. The six genic fragments identi®ed
by reverse northern hybridization analysis as being
expressed exclusively upon Al3+ challenge (Fig. 3) are
worthy of further investigation as they may provide
insights into possible divergence between the Al3+ and
oxidative stress responsive pathways. Nevertheless, the
results of this study clearly demonstrated that the vast
majority of the genes represented by the transcripts
captured from sugarcane root tips were responsive to
both Al3+ and oxidative stress.
This investigation has provided insights into the nature
and induction behaviour of Al3+-reponsive genes in
sugarcane roots and represents the ®rst study speci®cally
to target rare transcripts, such as those participating in cell
signalling and the regulation of gene expression, elicited in
roots in response to challenge by a demonstrably
phytotoxic level of the metal. The presumed identities of
a subset of the genic fragments isolated indicated that Al3+stress results in the up-regulation of genes involved in
regulatory events. Furthermore, the patterns of Al3+- and
oxidative-stress-induced gene expression observed are in
accordance with evidence from other studies, indicating
that parallels exist in the way in which roots respond to
different abiotic stresses. These results will guide future
exploration of the potential for manipulating the Al3+tolerance status of this important tropical crop: approaches
that will focus on the contributions of cell signalling and
regulatory events to the tolerance phenotype.
Al-responsive genes in sugarcane 1173
Acknowledgements
I thank Drs Barbara Huckett, Deborah Carson and Stuart Rutherford
for invaluable advice and encouragement and Alistair McCormick
for his technical contributions to the study. The manuscript was
subject to critical evaluation by Professor Paula Watt and Dr Barbara
Huckett, for which I am truly grateful, as am I for the helpful
comments provided by two anonymous reviewers.
References
Allison JD, Brown DS, Novo-Gradac KJ. 1991. MINTEQA2/
PRODEFA2: a geochemical assessment model for environmental
system. Version 3.0 user manual. Athens, Georgia, USA: United
States Environmental Protection Agency.
Altschul SF, Madden TL, SchaÈffer AA, Zhang J, Zhang Z,
Miller W, Lipman DJ. 1997. Gapped BLAST and PSI-BLAST:
a new generation of protein database search programs. Nucleic
Acids Research 25, 3389±3402.
Aniol A, Gustafson JP. 1984. Chromosome location of genes
controlling aluminum tolerance in wheat, rye, and triticale.
Canadian Journal of Genetics and Cytology 26, 701±705.
Baker AB, Cerniglia GJ, Zaman A. 1990. Microtiter plate assay
for the measurement of glutathione and glutathione disul®de in
large numbers of biological samples. Analytical Biochemistry
190, 360±365.
Berzonsky WA. 1992. The genomic inheritance of aluminum
tolerance in `Atlas 66' wheat. Genome 35, 689±693.
Bradford MM. 1976. A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Analytical Biochemistry 72,
248±254.
Cairney J, Xu N, Pullman GS, Ciavatta VT, Johns B. 1999.
Natural and somatic embryo development in loblolly pine.
Applied Biochemistry and Biotechnology 77±79, 5±17.
Carson DL, Botha FC. 2000. Preliminary analysis of expressed
sequence tags for sugarcane. Crop Science 40, 1769±1779.
Carson DL, Huckett BI, Botha FC. 2002. Sugarcane ESTs
differentially expressed in immature and maturing internodal
tissue. Plant Science 162, 289±300.
Cruz-Ortega R, Cushman JC, Ownby JD. 1997. cDNA clones
encoding 1,3-beta-glucanase and a ®mbrin-like cytoskeletal
protein are induced by Al toxicity in wheat roots. Plant
Physiology 114, 1453±1460.
Delhaize E, Ryan PR. 1995. Aluminum toxicity and tolerance in
plants. Plant Physiology 107, 315±321.
Diatchenko L, Lau Y-FC, Campbell AP, et al. 1996. Suppression
subtractive hybridization: a method for generating differentially
regulated or tissue-speci®c cDNA probes and libraries.
Proceedings of the National Academy of Sciences, USA 93,
6025±6030.
Drummond RD, GuimaraÄes CT, Felix J, Ninamango-CaÂrdenas
FE, Carneiro NP, Paiva E, Menossi M. 2001. Prospecting
sugarcane genes involved in aluminum tolerance. Genetics and
Molecular Biology 24, 221±230.
Ezaki B, Yamamoto Y, Matsumoto H. 1995. Cloning and
sequencing of the cDNAs induced by aluminum-treatment and
Pi-starvation in cultured tobacco cells. Physiologia Plantarum 93,
11±18.
Ezaki B, Tsugita S, Matsumoto H. 1996. Expression of a
moderately anionic peroxidase is induced by aluminum
treatment in tobacco cells: possible involvement of peroxidase
isozymes in aluminum ion stress. Physiologia Plantarum 96,
21±28.
Ezaki B, Sivaguru M, Ezaki Y, Matsumoto H, Gardner RC.
1999. Acquisition of aluminum tolerance in Saccharomyces
cerevisiae by expression of the BCB or NtGDI1 gene derived
from plants. FEMS Microbiology Letters 171, 99±105.
Ezaki B, Gardner RC, Ezaki Y, Matsumoto H. 2000. Expression
of aluminium-induced genes in transgenic Arabidopsis plants can
ameliorate aluminium and/or oxidative stress. Plant Physiology
122, 657±665.
Gutteridge JMC, Quinlan GJ, Clark I, Halliwell B. 1985.
Aluminium salts accelerate peroxidation of membrane lipids
stimulated by iron salts. Biochimica et Biophysica Acta 835,
441±447.
Hamel F, Breton C, Houde M. 1998. Isolation and
characterization of wheat aluminium-regulated genes: possible
involvement of aluminium as a pathogenesis response elicitor.
Planta 205, 531±538.
Hetherington SJ, Asher CJ, Blamey FCP. 1986. Tolerance of
sugarcane to Al in soil and solution culture. Proceedings of the
Australian Society of Sugar Cane Technologists Congress 63±68.
Hewitt EJ. 1966. Sand and water culture mehods used in the study
of plant nutrients, 2nd edn. Farnham Royal, Bucks, UK:
Commonwealth Agricultural Bureaux.
Ibrahim RK, Cavia E. 1975. Acrylamide gel electrophoresis of
proteins from intact and cultured plant tissues. Canadian Journal
of Botany 53, 517±519.
Ingelbrecht IL, Mandelbaum CI, Mirkov TE. 1998. Highly
sensitive northern hybridization using a rapid protocol for
downward alkaline blotting of RNA. BioTechniques 25, 420±425.
Jaakola L, PirttilaÈ M, Hohtola A. 2001. cDNA blotting offers an
alternative method for gene expression studies. Plant Molecular
Biology Reporter 19, 125±128.
Kirch H-H, Phillips J, Bartels D. 2002. Dehydration-stress signal
transduction. In: Scheel D, Wasternack C, eds. Plant signal
transduction. New York City, NY: Oxford University Press Inc,
140±164.
Kochian LV. 1995. Cellular mechanisms of aluminium toxicity and
resistance in plants. Annual Review of Plant Physiology and Plant
Molecular Biology 46, 237±260.
Kocsy G, Galiba G, Brunhold C. 2001. Role of glutathione in
adaptation and signalling during chilling and cold acclimation in
plants. Physiologia Plantarum 113, 158±164.
Kocsy G, Szalai G, VaÂguÂjfalvi A, SteÂhli L, Orosz G, Galiba G.
2000. Genetic study of glutathione accumulation during cold
hardening in wheat. Planta 210, 295±301.
Kosower NS, Kosower EM, Wertheim B. 1969. Diamide, a new
reagent for the intracellular oxidation of glutathione to the
disul®de.
Biochemical
and
Biophysical
Research
Communications 37, 593±596.
Kurth J, Varotto C, Pasaresi P, Biehl A, Richly E, Salamini F,
Leister D. 2002. Gene-sequence-tag expression analyses of 1800
genes related to chloroplast functions. Planta 215, 101±109.
Landell MGA. 1989. Comportamento da cana-de-acËuÂcar
(Saccharum spp.) frente a nõÂveis de aluÂmõÂnio, em solucËaÄo
nutritiva. FCAV/UNESP, PhD thesis. In: Drummond RD,
GuimaraÄes CT, Felix J, Ninamango-CaÂrdenas FE, Carneiro NP,
Paiva E, Menossi M. 2001. Prospecting sugarcane genes involved
in aluminum tolerance. Genetics and Molecular Biology 24, 221±
230.
Leustek T. 2002. Sulfate metabolism. In: Somerville CR,
Meyerowitz EM, eds. The Arabidopsis book. Rockville, MD:
American Society of Plant Biologists, 1±17.
Matsumoto H. 2000. Cell biology of aluminium toxicity and
tolerance in higher plants. International Review of Cytology 200,
1±46.
Newman T, de Bruijn FJ, Green P, et al. 1994. Genes galore: a
summary of methods for accessing results from large-scale partial
1174 Watt
sequencing of anonymous Arabidopsis cDNA clones. Plant
Physiology 106, 1241±1255.
Richards KD, Schott EJ, Sharma YK, Davis KR, Gardner RC.
1998. Aluminum induces oxidative stress genes in Arabidopsis
thaliana. Plant Physiology 116, 409±418.
Scheel D, Wasternack C. 2002. Signal transduction in plants:
cross-talk with the environment. In: Scheel D, Wasternack C, eds.
Plant signal transduction. New York City, NY: Oxford
University Press Inc, 1±5.
Schroeder BL, Robinson JB, Wallace M, Turner PET. 1994.
Soil acidi®cation: effects and occurrence in the South African
sugar industry. Proceedings of the South African Sugar
Technologist's Association 68, 70±74.
Schumann AW, McArthur D, Meyer JH. 1999. Further revision
of lime recommendations used in the South African sugar
industry. Proceedings of the South African Sugar Technologist's
Association 73, 58±73.
Snowden KC, Gardner RC. 1993. Five genes induced by
aluminum in wheat (Triticum aestivum L.) roots. Plant
Physiology 103, 855±861.
Snowden KC, Richards KD, Gardner RC. 1995. Aluminiuminduced genes. Induction by toxic metals, low calcium, and
wounding, and pattern of expression in root tips. Plant Physiology
107, 341±348.
Tang Y, Garvin DF, Kochian LV, Sorrells ME, Carver BF.
2002. Physiological genetics of aluminum tolerance in wheat
cultivar Atlas 66. Crop Science 42, 1541±1546.
Taylor GJ. 1991. Current views of the aluminum stress response:
the physiological basis of tolerance. Current Topics in Plant
Biochemistry and Physiology 10, 57±93.
Tsang SS, Yin X, Guzzo-Arkuran C, Jones VS, Davison AJ.
1993. Loss of resolution in gel electrophoresis of RNA: a problem
associated with the presence of formaldehyde gradients.
BioTechniques 14, 38±381.
Turner PET, Meyer JH, Wood RA. 1992. Some results of ®eld
experiments to test lime and phospho-gypsum for aluminium
amelioration. Proceedings of the South African Sugar
Technologists' Association 66, 44±49.
Wilson DO. 1984. Determination of aluminium in plant tissue
digests using a catechol violet colorimetric method.
Communications in Soil Science and Plant Analysis 15, 1269±
1279.
Xiang C, Oliver D. 1998. Glutathione metabolic genes
coordinately repond to heavy metals and jasmonic acid in
Arabidopsis. The Plant Cell 10, 1539±1550.
Xiong L, Zhu J-K. 2002. Salt-stress signal transduction in plants.
In: Scheel D, Wasternack C, eds. Plant signal transduction. New
York City, NY: Oxford University Press Inc, 165±197.