Australian Biochemist
Showcase on Research
Molecular Basis of the Anaerobic Response
in Plants
R. Dolferus, E.J. Klok, K.P. Ismond, C. Delessert, S.Wilson,W.J. Peacock and E.S. Dennis
CSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601
12
Survival of plants under
low oxygen conditions
Molecular changes in plants
upon low oxygen treatment
Plants and animals are obligate aerobic
organisms. Oxygen is essential as a terminal electron acceptor in respiration and
is also required in many other biochemical reactions.Adverse weather conditions
such as heavy rainfall or flooding and
melting of snow in cold climate regions,
can cause soil waterlogging and restrict
oxygen supply to living organisms. It was
essential to evolve survival mechanisms
for these temporarily adverse conditions
and prokaryotic and eukaryotic organisms
have achieved this using a variety of different strategies (1, 2).
Plants are confronted with peculiar
problems, being highly differentiated organisms, anchored with their roots in the
soil, and having no central nervous system, nor an oxygen transporting blood
stream like animals. Plants show an amazing variability in the capacity to tolerate
low oxygen stress. Aquatic plants are morphologically and physiologically very different from land plants, but even land
plants themselves show a lot of variation
in low oxygen tolerance.
It is clear that the response to low oxygen stress in plants is complex and involves many physiological as well as morphological changes. A lot of research
effort has been invested in understanding
the physiological basis of the low oxygen
response.This work was often carried out
on plants with varying degrees of tolerance to the stress, which often complicated interpretation and extrapolation of
the data.
The aim of our laboratory is to understand the molecular basis of the response
to low oxygen conditions, with the longterm aim of designing strategies to improve flooding-tolerance of important
crop plants. We have chosen Arabidopsis
as a model because of the molecular biology resources available for this plant.
One way to understand the physiological response of plants to low oxygen conditions is to identify the genes that are
activated under these conditions.Two-dimensional electrophoresis (2D-EF)
showed that about 20 anaerobic proteins
(ANPs) are expressed in maize roots under low oxygen conditions (3). In contrast
to other stresses, the function of many of
these genes is known. ADH (alcohol dehydrogenase) was the first anaerobic gene
identified (4) and has served as a model
ever since (5).
The function of many other anaerobic
proteins (ANPs) were identified by
proteome analysis, a combination of 2DEF, pulse labelling, immunology and protein microsequencing techniques. Most of
the ANPs are involved in sugar metabolism and fermentation (6). Recently, 2Delectrophoresis was combined with the
power of mass spectrometry to identify
another 46 proteins involved in the
acclimation response of maize roots,
showing that proteins from a wider range
of cellular functions are also induced by
low oxygen stress (7).
Obviously, metabolic adaptation is important during the acclimation period to
low oxygen conditions. This may not be
surprising, since oxygen is an important
substrate of cell metabolism. A reduction
in oxygen supply leads to an 18-fold reduction in ATP production, so plant roots
have to compensate for this loss by accelerating sugar metabolism and glycolysis. Induction of fermentation pathway
enzymes is required to regenerate NAD+
reducing power.
different plant species. There are few examples of tolerant and intolerant varieties within one species, so a metabolic
engineering approach using transgenic
plants might be the only way to address
this question.
There is an assumption that plant species with a more active ethanol fermentation compared to lactic acid fermentation metabolism are more flooding-tolerant (8, 9) Unlike lactic acid, ethanol does
not cause cytosolic acidosis, which causes
cell death under low oxygen stress. But
recent data suggest that the origin of acidosis might not be purely metabolic, but
could involve proton pumps (9, 10).
Most higher plants induce ethanol, lactic acid and alanine fermentation pathways
de novo during low oxygen stress. These
pathways consist of only one or two
enzymatic reactions, making them a relatively easy target to carry out metabolic
engineering. We have cloned the genes
involved in these three pathways in
Arabidopsis and produced sense and
antisense overexpressing plants, in order
to find out their contribution to flooding
tolerance.
Using a survival assay for low oxygen
conditions (11), we were able to show
that overexpression of pyruvate decarboxylase, the first enzyme of the alcohol
fermentation pathway, improves survival
in Arabidopsis. We are currently investigating how overexpression of the entire
alcohol fermentation pathway, and downregulation of lactic acid fermentation, can
further improve survival.
Gene expression under low
oxygen conditions
We have mapped the promoter ele-
Can flooding tolerance be obtained ments of the ADH1 gene with the aim of
identifying the factors regulating the
by metabolic engineering?
Despite this extensive knowledge, it remains unclear what determines the difference between flooding-sensitive and
flooding-tolerant plant species.This question is very difficult to answer, because of
large morphological differences between
anaerobic response genes.The anaerobic
response element (ARE) was identified in
the maize and Arabidopsis ADH1 promoters (12, 13) (Fig. 1) and is a bipartite element consisting of GT- and GC-motifs
which are both crucial for gene expres-
Volume 31, No. 4, December 2000
Showcase on Research
Molecular Basis of the Anaerobic Response in Plants (contin.)
is activated.
The first stage in the anaerobic response obviously consists of putting in
place the regulatory machinery required
for induction of the second stage genes,
such as ADH1. This provides a metabolic
adaptation to supply energy and metabolic
components for survival of short-term
low oxygen stress.The second stage could
also help to establish a third stage in the
response: the provision of longer term
survival strategies.
One example in some plants is the development of aerenchyma, root cortical
air spaces promoting air transport from
shoot to root (17, 18). Aerenchyma formation requires the hormone ethylene.
Amongst the genes induced by low oxygen conditions are ethylene biosynthetic
enzymes (19).
Overlap between low oxygen
stress and other stress responses?
Fig. 1: Structure and organisation of the Arabidopsis ADH1 promoter. The anaerobic
response element (ARE) consists of a GT-motif and a GC-motif. In contrast to the maize ADH1
gene (12), the Arabidopsis ARE consists of single GT- and GC-motifs and the GT-motif is in the
opposite orientation. Both elements are critical for expression under low oxygen conditions, but
they are also essential for expression under other stress conditions, indicating the ARE has a
much wider function.The G-Box sequence is not required for response to low oxygen stress, but
is required by the other stresses to obtain high expression levels. Cold stress is further differentiated from dehydration, salinity and wounding stress in that it requires the G-Box sequence in an
ABA-independent manner. This was demonstrated by endogenous ABA measurements and the
use of Arabidopsis ABA mutants (6).
sion. Gel retardation experiments showed
a protein binding to the GC-motif (GCBP1; ref. 14), but we have been unable to
clone the corresponding gene.
Progress was made in the identification
of the protein binding to the GT-motif,
when we realised that this site resembles
a Myb-transcription factor-binding site.We
showed that AtMYB2, a drought and ABAinduced transcription factor indeed binds
to the GT-motif. AtMYB2 is also low-oxygen-induced and reaches maximal expression (2-4h) prior to ADH1 (4-6h).AtMYB2
is able to trans-activate ADH1 promoter
activity in transient assays (15).
We could not generate transgenic
plants with increased or decreased levels
of AtMYB2 using a strong constitutive
promoter. This could mean that manipulation of AtMYB2 levels in either direction may be lethal to the plant, an hypothesis that we are currently investigating
using inducible rather than constitutive
promoters.
Different stages in the low
oxygen response?
Treatment of plants with the protein
synthesis inhibitor cycloheximide during
anaerobic treatment prevents induction
of ADH1 mRNA. This indicates that protein synthesis is required prior to induction of anaerobic genes such as ADH1.
AtMYB2 induction does not require protein synthesis (15). Cycloheximide treatment caused up-regulation of AtMYB2
mRNA levels under both control and low
oxygen conditions.
This has been reported for several
other signal transduction components and
could mean that cycloheximide acts as a
stabiliser of mRNAs which normally turn
over quickly (16). This could mean that
AtMYB2 mRNA levels, and possibly other
regulatory factors of the anaerobic response, are regulated post-transcriptionally. Studying the regulation of AtMYB2
could therefore provide us with an understanding of how the anaerobic response
The ARE is not only required for gene
expression under low oxygen conditions,
but is also critical for response to a variety of other stresses including drought,
cold, salinity and wounding (Fig. 1).
The transcription factor AtMYB2 is induced by all the stresses that induce the
ADH1 gene, suggesting that at least some
of the regulatory factors could be shared
by different stresses. Genes containing the
ARE in their promoter must play a similar role in acclimation to various stresses.
It is possible that the metabolic adaptation seen under low oxygen conditions is
at least partially shared by other stresses.
Increased sugar metabolism caused by glycolysis and alcohol fermentation could –
certainly in the presence of oxygen – provide the energy to cope with the sudden
requirement to establish protection
against abiotic stresses.
Microarrays: the answer to
many questions?
Proteome analysis has enabled the identification of many anaerobic proteins, but
failed to reveal any of the regulatory components. This is partially due to the fact
that the technique was mainly used to
look at the acclimation stage and not the
early stage of the response, during which
most regulatory events take place. On the
other hand, because most regulatory proteins are expressed at very low levels, it
is doubtful whether those components
13
Australian Biochemist
Showcase on Research
Molecular Basis of the Anaerobic Response in Plants (contin.)
would be detectable on 2D-gels.
Other restrictions are the limited pH
range of the first dimension isoelectric
focusing gels, limiting detection of certain
transcription factors with high pI values,
and the inability to identify insoluble and
down-regulated proteins.
Molecular biologists have been dreaming of a method to look at gene expression at the genome level, and the
microarray technology allows this to be
acheived (20, 21). Up to 10,000 DNA samples (genomic DNA, cDNAs,
oligonucleotides) can be spotted on one
glass microscope slide using a robot and
analysed with fluorescent RNA probes.
We have recently started this approach
with the intention of identifying regulatory factors involved in low-oxygen-induced gene expression.
A cDNA librar y was made from
anaerobically-induced Arabidopsis roots,
and 2 sets of microarray slides were prepared: a 3.5K array, containing 1,000
anaerobic cDNA clones plus 2,500
Arabidopsis EST clones, and a 10K array
containing 10,000 random anaerobic
cDNA clones. We screened the 3.5K array using mRNA from different time
points during anaerobic treatment, and
established a list of more than 200 genes
that are up-regulated or down-regulated.
The list contains genes from a wide variety of cellular processes, including signal transduction components and transcription factors.
Microarrays will allow us to answer
many questions about the anaerobic response in plants. We will be able to look
at different time points of anaerobic treatment and compare the genes that are induced or repressed. We will also be able
to identify the genes that are shared by
other environmental stresses, by using
mRNA probes from different stress treatments. By using RNA of cycloheximidetreated plants, we will be able to identify
the earlier regulatory components that
are stabilised by cycloheximide.
We will look at plants overproducing
AtMYB2 and identify the genes that are
affected by this factor. It might also be
possible to identify quantitative and qualitative differences in the set of expressed
genes in flooding-tolerant and floodingsensitive plant species (provided there is
enough cross-hybridisation) and identify
key genes in determining tolerance.
14
The opportunities offered by microarrays
are seemingly limitless. However, the real
challenges have become the bottlenecks
of computing and database management,
and the question of how to exploit this
sudden wealth of information. This will
involve testing the relevance of several
genes by producing lots of transgenic
plants with sense and antisense over-expression constructs, or large scale screening of insertion mutant lines to find any
mutant phenotype associated with a particular gene. The speed and reliability of
the microarray approach can also be exploited to analyse transgenic or mutant
lines. This should enhance our efforts to
identify those genes relevant for the
anaerobic response.
References
1. Zitomer, R. S., and Lowry, C.V. (1992) Microbiol.
Rev. 56, 1-11
2. Bunn, H. F., and Poyton, R. O. (1996) Physiol. Rev.
76, 839-885
3. Sachs, M. M., Freeling, M., and Okimoto, R. (1980)
Cell 20, 761-767
4. Hagemann, R. H., and Flesher, D. (1960) Arch.
Biochem. Biophys. 87, 203-209
5. Freeling, M., and Bennett, D. C. (1985) Ann. Rev.
Genet. 19, 297-323
6. Dennis, E. S., Dolferus, R., Ellis, M., Rahman, M.,
Wu,Y., Hoeren, F. U., Grover, A., Ismond, K. P.,
Good,A. G., and Peacock,W. J. (2000) J. Exp. Bot.
51, 89-97
7. Chang, W. W., Huang, L., Shen, M., Webster, C.,
Burlingame, A. L., and Roberts, J. K. (2000) Plant
Physiol. 122, 295-318
8. Roberts, J. K. M., Callis, J., Jardetzky, O.,Walbot,
V., and Freeling, M. (1984) Proc. Natl. Acad. Sci.
USA 81, 6029-6033
9. Germain, V., Raymond, P., and Ricard, B. (1997)
Plant Mol. Biol. 35, 711-721
10. Ratcliffe, R.G. (1997) Ann. Bot. 79 (Suppl A), 3948
11. Ellis, M. H., Dennis, E. S., and Peacock,W. J. (1999)
Plant Physiol. 119, 57-64
12. Olive, M. R.,Walker, J. C., Singh, K., Dennis, E. S.,
and Peacock,W. J. (1990) Plant Mol. Biol. 15, 593604
13. Dolferus, R., Jacobs, M., Peacock,W.J., and Dennis,
E.S. (1994) Plant Physiol. 105, 1075-1087
14. Olive, M. R., Peacock, W. J., and Dennis, E. S.
(1991) Nucl. Acids Res. 19, 7053-7060
15. Hoeren, F., Dolferus, R., Wu,Y., Peacock, W. J.,
and Dennis, E. S. (1998) Genetics 149, 479-490
16. Yeilding, N. M., Procopio, W. N., Rehman, M. T.,
and Lee,W. M. F. (1998) J. Biol. Chem. 273, 1574915757
17. Drew, M. C., Jackson, M., Giffard, S. C., and
Campbell, R. (1981) Planta 153, 217-224
18. Drew, M.C. (1997) Ann. Rev. Plant Physiol. Plant
Mol. Biol. 48, 223-250
19. Olson, D. C., Oetiker, J. H., and Yang, S. F. (1995)
J. Biol. Chem. 270, 14056-61
20. Richmond,T., and Somerville, S. (2000) Curr. Opin.
Plant Biol. 3, 108-116
21. Schaffer, R., Landgraf, J., Perez-Amador, M., and
Wisman, E. (2000) Curr. Opin. Biotechnol. 11, 162167
DEHYDRATION, SALINITY, WOUNDING
COLD
LOW OXYGEN
ABA
ARE
CCACGTGG
G-Box
(-214 to -208)
Fig. 1
AAAACCAAA
GCCCC
GT-Motif
(-159 to -149)
GC-Motif
(-147 to -143)
TATAAAT
TATA -Box
(-31 to -23)