Modulation of Nitrosative Stress by S-Nitrosoglutathione
Reductase Is Critical for Thermotolerance and Plant
Growth in Arabidopsis
Ung Lee,aChris Wie,aBernadette O. Fernandez,bMartin Feelisch,b,1and Elizabeth Vierlinga,2
aDepartment of Biochemistry and Molecular Biophysics, University of Arizona, Tucson, Arizona 85721
bWhitaker Cardiovascular Institute, Boston University School of Medicine, Boston, Massachusetts 02118
Nitric oxide (NO) is a key signaling molecule in plants. This analysis of Arabidopsis thaliana HOT5 (sensitive to hot tem-
peratures), which is required for thermotolerance, uncovers a role of NO in thermotolerance and plant development. HOT5
encodes S-nitrosoglutathione reductase (GSNOR), which metabolizes the NO adduct S-nitrosoglutathione. Two hot5 missense
alleles and two T-DNA insertion, protein null alleles were characterized. The missense alleles cannot acclimate to heat as dark-
grown seedlings but grow normally and can heat-acclimate in the light. The null alleles cannot heat-acclimate as light-grown
plants and have other phenotypes, including failure to grow on nutrient plates, increased reproductive shoots, and reduced
fertility. The fertility defect of hot5 is due to both reduced stamen elongation and male and female fertilization defects. The hot5
null alleles show increased nitrate and nitroso species levels, and the heat sensitivity of both missense and null alleles is
associated with increased NO species. Heat sensitivity is enhanced in wild-type and mutant plants by NO donors, and the heat
sensitivity of hot5 mutants can be rescued by an NO scavenger. An NO-overproducing mutant is also defective in thermotol-
erance. Together, our results expand the importance of GSNOR-regulated NO homeostasis to abiotic stress and plant de-
Nitric oxide (NO) is a short-lived, endogenously produced radical
2006; Besson-Bard et al., 2008). Despite its deceivingly simple
structure, the rich chemistry of NO in biological systems gives
rise to multiple secondary and tertiary reaction products, greatly
complicating our mechanistic understanding of NO-related ef-
fects (Stamler and Hausladen, 1998; Mancardi et al., 2004;
Ridnour et al., 2004). Directly and via its various chemical trans-
formations, NO not only accomplishes signaling functions but
ing other radical reactions) and pro-oxidant (through the pro-
duction ofreactive nitrogen species;RNS) properties. Inaddition
to effects on redox status, the formation of RNS leads to nitro-
sation, nitrosylation, and nitration reactions with other mole-
cules. Most of the regulatory effects of NO are thought to be
ing heme nitrosylation, Tyr nitration, Cys nitrosation, and even
Wang et al., 2006b; West et al., 2006; Zaninotto et al., 2006).
In plants, NO is believed to be produced via two different
enzymatic pathways (Guo et al., 2003; Crawford, 2006). In one
pathway, it is generated by nitrate reductase through the suc-
cessive reduction of nitrate to nitrite and further to NO. In the
other pathway, L-Arg, plus oxygen and NADPH, is converted to
NO and citrulline by the action of a NO synthase, although the
actual existence and identity of plant NO synthase is currently
unresolved (Crawford et al., 2006; Guo, 2006; Zemojtel et al.,
2006). In some cases, NO is also produced by a nonenzymatic
mechanism in which NO2?is converted to NO under acidic pH
conditions in the plant apoplast (Bethke et al., 2004a). NO has
been demonstrated to be involved in many different physiolog-
ical processes in plants. These include seed germination (Beligni
and Lamattina, 2000; Bethke et al., 2004b, 2006), plant defense
responses (Zeidler et al., 2004; Zeier et al., 2004; Delledonne,
2005; Modolo et al., 2005; Mur et al., 2006), leaf senescence
(Corpas et al., 2004; Guo and Crawford, 2005), stomatal move-
2005), hormonal signaling (Guo et al., 2003; Huang et al., 2004),
and flowering (He etal.,2004; Simpson, 2005).NO hasalso been
implicated in responses to wounding and a number of abiotic
Because of the multitude of possible chemical transitions and
targets of NO, a precise determination of the mechanism of NO
action in any of these important plant processes remains a
challenge. Therefore, it is imperative to improve our understand-
ing of NO metabolism in plants.
NO-derived RNS readily react with the major cellular antiox-
idant GSH to form S-nitrosoglutathione (GSNO). The main reac-
1Current address: Warwick Medical School, University of Warwick,
Gibbet Hill Campus, Coventry CV4 7AL, UK.
2Address correspondence to email@example.com.
The author responsible for distribution of materials integral to the
findings presented in this article in accordance with the policy described
in the Instructions for Authors (www.plantcell.org) is: Elizabeth Vierling
WOnline version contains Web-only data.
The Plant Cell, Vol. 20: 786–802, March 2008, www.plantcell.org ª 2008 American Society of Plant Biologists
group to other cellular thiols to form longer-lived nitrosothiols
(SNOs), an exemplary transnitrosation reaction. Endogenous
GSNO has been proposed to be a significant player in NO regu-
a process termed S-nitrosylation (Ji et al., 1999; Liu et al., 2001).
This modification is sometimes referred to as ‘‘the new phos-
phorylation,’’ although it is not known to be enzymatically cat-
alyzed or otherwise protein-mediated. Increasing numbers of
plant proteins are reported to be reversibly nitrosated on Cys
residues (Perazzolli et al., 2004; Lindermayr et al., 2005, 2006;
Belenghi et al., 2007). Such modifications often result in the
inhibition of enzyme activity or alteration in protein function. In
analogy to the concept of oxidative stress, an accumulation of
nitroso species as a result of either the enhanced production of
NO/RNS or the decreased clearance of nitrosated products has
been termednitrosative stress(Ridnour etal.,2004).Althoughby
now it is an established part of NO metabolism in mammalian
cells, little is known about the occurrence and consequences of
of GSNO to transfer NO to protein thiols implies that GSNO
affect many regulatory processes.
It is now recognized that an evolutionarily conserved, GSH-
dependent formaldehyde dehydrogenase (FALDH), a type III
alcohol dehydrogenase, has activity as a GSNO reductase
(GSNOR) (Jensen et al., 1998; Liu et al., 2001). In fact, it has
been proposed that the major role of GSNOR/FALDH is in con-
trolling GSNO and SNO levels rather than in detoxifying formal-
dehyde in living cells. GSNOR metabolizes GSNO to a mixture of
products depending on conditions, including GSSG, hydroxyl-
amine, NH3, and GSH sulfinic acid (Jensen et al., 1998). The
overall result is a reduction of GSNO and a decrease in the like-
lihood of enhanced protein nitrosation reactions.
In plants, there have been limited studies of GSNOR either
from the perspective of its formaldehyde-detoxifying activity
(Uotila and Koivusalo, 1979; Giese et al., 1994; Martı ´nez et al.,
1996; Dixon et al., 1998; Achkor et al., 2003) or from that of its
function in GSNO reduction (Sakamoto et al., 2002; Feechan
et al., 2005; Ruste ´rucci et al., 2007). In Arabidopsis thaliana,
(At5g43940) previously named ALCOHOL DEHYDROGENASE2
out the plant, downregulated by wounding and jasmonic acid,
and upregulated by salicylic acid (Diaz et al., 2003). Sakamoto
et al. (2002) have demonstrated that Arabidopsis GSNOR is
capable of reducing GSNO using Escherichia coli extracts ex-
pressing recombinant protein.
Information about the phenotypes associated with a loss of
GSNOR function is scarce. A T-DNA insertion mutant of the
single copy GSNOR gene in Arabidopsis was recently isolated
(designated gsnor1-3) (Feechan et al., 2005), and transgenic
Arabidopsis plants that overexpress or produce <50% wild-type
levels of GSNOR have been generated (Ruste ´rucci et al., 2007).
Studies of the disease susceptibility of these plants have yielded
contradictory results. Feechan et al. (2005) reported that the
gsnor1-3 null mutant was compromised in both R-mediated and
basal disease resistance, failing to mount a defense response
through the salicylic acid signaling network. By contrast, trans-
genic Arabidopsis plants with reduced GSNOR displayed en-
hanced resistance to Peronospora parasitica (Ruste ´rucci et al.,
2007). Furthermore, systemic acquired resistance and PR1 gene
expression were enhanced in antisense plants and impaired in
overexpression plants. While the disparity in these results re-
mains to be resolved, there is no doubt that GSNOR plays a role
in response to pathogens. No growth or developmental pheno-
types were reported associated with the absence or reduction of
GSNOR, with the exception of reduced root growth (Espunya
et al., 2006), but both groups found an approximate doubling of
total cellular SNO species, consistent with the role of GSNOR in
We now report that GSNOR activity is necessary for the
acclimation of plants to high temperature and for normal devel-
opment and fertility under optimal growth conditions. Our results
demonstrate that GSNOR has an important role in the homeo-
stasis of NO and its metabolites, affecting not only abiotic stress
but also plant developmental processes.
The Thermotolerance-Defective Mutant hot5
We identified an Arabidopsis thermotolerance-defective mutant,
hot5-1, in a screenof ethyl methanesulfonate–mutagenized seed-
lings using a hypocotyl elongation assay that was described pre-
viously (Hong and Vierling, 2000). Dark-grown, 2.5-d-old hot5-1
seedlings are completely blocked in hypocotyl elongation after
150 min of 458C heat treatment, even following a pretreatment
at 388C, which allows wild-type seedlings to survive (Figure 1B).
The hot5-1 mutant was backcrossed to the wild-type ecotype
Columbia (Col) for standard genetic analysis. F2 backcrossed
lines showed that the thermotolerance-defective phenotype seg-
regated as a single recessive trait (data not shown). Using estab-
lished map-based cloning methods (see Methods), the hot5-1
mutation was located toward the bottom of chromosome 5, be-
tween BAC clones F6B6 and MLN1. We sequenced all annotated
genes in the mapped region using genomic DNA from hot5-1
mutant plants. Sequence analysis revealed a single G-to-A muta-
tion, resulting in a Glu-to-Lys substitution at amino acid 283 in
the seventh exon of the GSNOR gene (At5g43940) (Figure 1A).
Glu-283 is 100% conserved in GSNOR from plants and other
organisms, including bacteria and human (see Supplemental
Figure 1 online).
To confirm that GSNOR is indeed the gene responsible for the
observed hot5-1 phenotype, we isolated additional alleles of the
GSNOR gene. A second missense mutation (hot5-3) was iso-
lated from available Tilling lines (Col erecta background) (Till
et al., 2003). The hot5-3 mutation leads to the substitution of a
conserved amino acid also in exon 7 (G288R), five amino acids
from hot5-1 (Figure 1A; see Supplemental Figure 1 online). Two
T-DNA insertion alleles were also obtained, hot5-2 (Col back-
ground), which is located in exon 1 and is identical to gsnor1-3
reported by Feechan et al. (2005), and hot5-4 (Wassilewskija
[Ws] background) in exon 4 (Figure 1A). The hot5-1, hot5-2, and
hot5-3 mutant alleles were backcrossed to the wild-type Col
GSNOR in Thermotolerance 787
ecotype, and the hot5-4 allele was backcrossed to the Ws
ecotype, two times to remove background mutations.
The hot5 mutants were tested for their ability to acquire heat
tolerance in comparison with the null mutant of Heat-Shock
Protein101 (Hsp101; hot1-3), which has an established heat-
sensitive phenotype (Hong and Vierling, 2001). When tested in
the hypocotyl elongation assay for acquired heat tolerance, the
phenotype of hot5-3 was equivalent to that of hot5-1 (Figure 1B),
and both mutants had a less severe phenotype than hot1-3.
However, we were unable to perform the hypocotyl elongation
assay on the T-DNA insertion alleles, because although both
germinated on plates in the dark, they failed to elongate hypo-
cotyls or develop further; we have only been able to grow these
homozygous mutants effectively in the light on soil. Therefore, to
test the heat acclimation of the hot5 insertion alleles, we devel-
oped a new thermotolerance assay, using leaf discs punched
from the fourth or fifth leaves of 25-d-old plants (see Methods).
The ability of 25-d-old leaf tissue to acquire thermotolerance
differed dramatically between the hot5 missense and T-DNA
insertion mutants (Figure 1C). The hot5-2 and hot5-4 mutants
failed to acquire thermotolerance at this stage; they rapidly lost
chlorophyll and turned yellow, exhibiting a phenotype as severe
behaved like wild-type plants, remaining green. We conclude
that the two missense mutations (hot5-1 and hot5-3) are rela-
tively weak alleles of GSNOR compared with the insertion alleles
(hot5-2 and hot5-4). In total, this analysis confirms that mutation
of GSNOR prevents the normal development of acquired ther-
motolerance in plants.
GSNOR Is Not Heat Induced, and HSPs Are Normally
Expressed in Mutant Plants
To determine how the hot5 mutant alleles and high temperature
affect the abundance of GSNOR protein, protein blot analysis
was performed on total proteins extracted from leaf discs as
used for the experiment in Figure 1C. Arabidopsis GSNOR anti-
bodies detected an;40-kD band, consistent with the predicted
molecular mass of the GSNORcoding sequence (40,697 D).This
polypeptide was present at approximately the same abundance
in both control and heat-stressed wild-type leaves (Figure 2A).
As determined by protein gel blotting of a dilution series of total
leaf protein compared with purified recombinant Arabidopsis
GSNOR, the HOT5 protein represents ;0.02% of total dark-
grown wild-type seedling protein (0.01% in leaf protein; see
Supplemental Figure 2A online). The hot5-1 missense allele had
approximately half the protein amount as the wild type, and the
decreasing to about one-third or one-quarter the level seen in
wild-type plants (Figure 2A; see Supplemental Figure 2B online).
By contrast, the hot5-3 protein accumulated to wild-type levels.
Both T-DNA insertion alleles, hot5-2 and hot5-4, had no detect-
able GSNOR protein, indicating that these are protein null alleles
and confirming the specificity of our antibody for the GSNOR
protein (Figure 2A).
We further confirmed previous observations of the ubiquitous
Dolferus et al., 1997) by protein gel blot analysis. Samples were
Figure 1. hot5 Mutants Are Defective in the Acquisition of Thermotol-
(A) Location of the hot5 missense alleles, hot5-1 and hot5-3, and the
T-DNA insertion alleles, hot5-2 and hot5-4, on the GSNOR gene
(At5g43940). aa, amino acids.
(B) Ability of wild-type and hot5 mutant seedlings to elongate after the
indicated heat treatments in comparison with the wild type and the heat-
sensitive Hsp101 null mutant hot1-3. Seedlings were grown on plates in
the dark for 2.5 d and treated at 228C only (room temperature [RT]), at
388C for 90 min, or at 388C for 90 min followed by 2 h at 228C (acclimation
treatment) and then by 90, 120, or 150 min at 458C. Wild-type seedlings
continue to elongate after 458C treatment, but hot5 missense mutations
show growth arrest.
(C) Acquired thermotolerance of leaf discs. Leaf disc samples (5 mm in
diameter) were punched from rosette leaves of 25-d-old wild-type or
mutant plants and then floated on 2 mL of 10 mM MES-KOH buffer, pH
6.8, on 12-well microplates. Heat treatments were performed as de-
scribed for (B). Leaf discs were returned to 228C under 12 h of light/12 h
of dark and photographed 5 d later.
788The Plant Cell
isolated from mature seeds, 2.5-d-old dark-grown hypocotyls,
and different organs of mature plants. GSNOR protein was pres-
ent in all organs tested, including dried seeds (see Supplemental
Figure 2C online). These data indicate that the loss of GSNOR
activity could affect phenotypes through the plant life cycle.
We next measured the effect of the hot5 mutations on GSNOR
enzyme activity in total plant extracts (Figure 2B). In leaves of
25-d-old plants, the GSNO reduction activity of wild-type plants
was similar to values reported previously (Feechan et al., 2005),
with 12.1 6 1.3 or 11.1 6 1.5 nM GSNO-dependent NADH
oxidation?min?1?mg?1total protein seen in wild-type Col and
wild-type Ws, respectively. In the missense alleles, activity com-
pared with the wild type was 33.8% in hot5-1 and 58.8% in
hot5-3 plants. The hot5-1 protein is likely to have a similar spe-
cific activity to the hot5-3 protein, considering that it is of lower
abundance in the mutant plants (Figure 2A). The absence of phe-
notype in 25-d-old seedlings of the missense mutants suggests
that this level of activity is sufficient for wild-type growth. The two
null alleles had negligible activity; the low activity detected pre-
sumably represents nonspecific GSNO-stimulated oxidation of
NADH. These data are consistent with the more severe heat-
stress phenotype of light-grown plants carrying the null alleles.
We also measured activity in 2.5-d-old dark-grown seedlings
of wild-type Col and the two missense alleles (Figure 2B).
Expressed per milligram of total protein, GSNOR activity was
actually higher for all seedling samples than in leaves, but this
appears to reflect the higher levels of GSNOR protein per milli-
gram of total protein in seedlings versus leaves (see Supplemen-
talFigure2Aonline).Surprisingly,theactivityin missense mutant,
dark-grown seedlings, expressed as a percentage of wild-type
Figure 2. GSNOR Protein Accumulation and Enzyme Activity.
(A) Accumulation of GSNOR and HSPs in wild-type and hot5 mutant plants. Total protein was isolated from control (C; 228C) or heat-stressed (H; 388C
for 90 min, followed by 2 h at 228C) 25-d-old leaf discs and analyzed with the indicated Arabidopsis HSP and GSNOR antisera. Equal quantities of total
protein (0.5 mg for Hsp101 antibodies, 5 mg for GSNOR and sHSP antibodies) from each of the mutants or the wild type were separated on 7.5%
(Hsp101), 10% (GSNOR), or 15% (sHSP) SDS-PAGE gels. Protein blot analysis with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibodies
confirmed the presence of similar protein levels.
(B) GSNOR enzyme activity in wild-type and hot5 mutant plants.
GSNOR in Thermotolerance 789
values, was similar to the activity seen in 25-d-old plants (33.7%
of wild-type values for hot5-1 and 64.8% of wild-type values for
hot5-3). We also measured changes in GSNOR activity after heat
stress in the wild type and GSNOR missense mutants. We found
no statistically significant change in GSNOR activity under heat-
stress conditions in the wild type or mutants, or when protein
extraction was performed plus or minus DTT, or when seedlings
were grown for 2.5 d in the light instead of the dark (data not
shown). The reason that dark-grown seedlings of hot5-1 and
hot5-3 have a heat-stress phenotype in the dark, despite having
apparently reasonable GSNOR activity at this stage, is not ob-
activity is required in the dark for proper growth after heat stress,
that in light-grown seedlings other factors are present that com-
pensate for the reduced GSNOR activity, or that the missense
alleles of GSNOR have altered regulation in the dark that is not
preserved by our extraction and measuring conditions.
Because HSP expression is known to be an important com-
ponent of acquired thermotolerance, we also assayed the ac-
cumulation of different HSPs by protein blot analysis in the hot5
mutants (Figure 2A). All of the hot5 alleles showed wild-type
levels of Hsp101, which is essential for heat tolerance (Hong
and Vierling, 2000), as well as cytosolic small HSPs of the class I
and II types (Lee et al., 2005). Therefore, we conclude that hot5
mutantsare notcompromised in signaling mechanismsthat lead
to the expression of HSPs and that the absence of HSPs is not
the cause of the hot5 thermotolerance defect.
hot5 Null Mutants Have Pleiotropic Phenotypes
In addition to their inability to grow following germination on nu-
trient medium plates in the dark, we also found that the GSNOR
null mutants, hot5-2 and hot5-4, had severely reduced seed
yields and abnormal growth habits. To investigate GSNOR mu-
tant phenotypes in more detail, we observed the entire life cycle
of all hot5 mutant alleles during growth in three different photo-
periods, 16 h/8 h, 12 h/12 h, and 8 h/16 h light/dark cycles. The
two hot5 missense mutants grew as well as wild-type plants
during the whole life cycle under all three light conditions, con-
sistent with our conclusion that these are mild alleles (data not
shown). However, the hot5 null mutants showed pleiotropic
medium (Haughn and Somerville, 1986) containing 0.5% su-
crose. The mutant seed germinated, but growth was arrested
right after some root elongation and emergence of small coty-
ledons, which failed to green, and the seedlings eventually died
(Figure 3A). This mutant phenotype was not recovered in the
absence of sucrose, on higher sucrose concentrations (1, 2.5,
and 5%), by germination directly on water-saturated filter paper,
or when ammonium succinate was used to replace all other
nitrogen sources in the medium (see Supplemental Figure 4A
online). Thus, the basis of this phenotype is unresolved. The hot5
null mutants, however, could be recovered on soil, as shown in
Figure 3B, allowing further study of growth phenotypes.
When hot5-2 was grown under long-day conditions (16 h of
were shorter compared with plants grown under 8 or 12 h of light
(data not shown). Indeed, under long days, the chlorophyll
content of hot5-2 was only 62% of that of wild-type plants (see
Supplemental Figure 3A online). After bolting, the hot5 null
mutants were highly branched and semidwarf under all light
conditions (Figure 3C). The roots of hot5 -2 were also reduced in
Figure 3. Pleiotropic Phenotypes of hot5 Null Mutants under 12-h/12-h
Light/Dark Growth Conditions.
(A) Ten-day-old seedling plants on nutrient plates.
(B) Twenty-five-day-old soil-grown plants.
(C) Forty-five-day-old mature, soil-grown plants.
(D) Flower phenotype. For these photographs, one sepal, petal, and
stamen were detached.
(E) Full-grown silique. From left to right: Col wild type, hot5-2, Ws wild
type, and hot5-4. The ruler at left shows millimeters.
790The Plant Cell
length compared with those of wild-type plants (see Supple-
root length reported previously for GSNOR antisense plants is
not clear (Espunya et al., 2006). The plants were also long-lived,
continuing to produce leaves for as long as 25 d after wild-type
plants had senesced.
The most dramatic phenotype of the hot5 null mutants was
reduced fertility. Leaf numbers before bolting were not altered in
hot5-2 compared with the wild type under either long or short
days. Under 12 h of light, the hot5 homozygous null mutants
produced many flowers and siliques but set very few seeds per
plant. The mutants showed normal floral organ formation, with a
wild-type number of sepals and petals and normal pistil forma-
tion. However, petals of mutant flowers were somewhat shriv-
eled and smaller than wild-type petals, and stamens did not
elongate normally, although pollen was produced at wild-type
produce seeds; consequently, the siliques did not elongate
normally (Figure 3E).
There is no doubt that the failure of the hot5-2 and hot5-4
stamens to elongate properly contributes to the severely re-
duced fertility of these mutants. To determine whether the pollen
and stigma of hot5-2 function normally for fertilization, we
performed manual self-pollination and reciprocal test crosses
between hot5-2 and wild-type plants(Table 1).Self-pollination of
hot5-2 produced only 17.4 6 6.2 (SD) seeds/silique, in contrast
with 73.2 6 5.3 seeds/silique for the wild type. In the reciprocal
crosses, 23.1 6 8.2 seeds/silique were generated using hot5-2
as the female with wild-type pollen, and 42.4 6 7.7 seeds/silique
were generated with hot5-2 pollen and wild-type females. These
data indicate that in addition to reduced anther length, loss of
HOT5 functioncompromisesboth the maleand female functions
required for fertilization and/or seed development.
GSNOR Affects Intracellular NO/Nitrosation Levels
By metabolizing GSNO, a cytoplasmic reservoir of NO and a
nitrosating species, GSNOR potentially modulates cellular NO
status. To determine whether the absence of GSNOR indeed
affects NO/nitroso levels, and how this is further affected by
29,79-difluorescein diacetate (DAF-FM DA) (Arnaud et al., 2006).
For staining, protoplasts were prepared from leaves of 25-d-old
wild-type and hot5-2 mutant plants either before or after heat
stress. NO-dependent fluorescence signals were dramatically
higher in the cytosol and chloroplasts of hot5-2 protoplasts
compared with wild-type protoplasts from untreated leaves; in
fact, no significant DAF-FM DA staining was observed in wild-
type plants (Figure 4A). The same high levels of DAF-FM DA
staining were also observed in protoplasts of hot5-4 (data not
and hot5-4 leaf tissues, heat treatment led to only a minor
increase in NO-related fluorescence in the wild type, and no
apparent change was seen in the mutant when heat stress was
performed prior to protoplast isolation. We were unable to visu-
alize intact cells when protoplasts were heat stressed after
isolation and stained, so we could not test for rapid or transient
heat-induced changes in DAF-FM DA staining in protoplasts.
Feechan et al. (2005) reported that the hot5-2 mutant (named
gsnor1-3 by this group) has increased SNO species compared
with the wild type, and increased SNO levels were also reported
for plants in which GSNOR was reduced using an antisense
species in leaves from wild-type and hot5 null plants using gas-
hot5 null mutants were found to have approximately double the
amount of nitroso species compared with the wild type, consis-
tent with previous reports. Heat stress did not significantly
change nitroso species levels in either the mutant or the wild
type. Unexpectedly, nitrate levels were also markedly higher in
fact that all plants had been grown at the same time on the same
soil. Theincrease in nitrate content appears to be correlated with
the increase in nitroso species concentration, suggesting a link
Collectively, these resultssuggestthat hot5 null mutantshavean
increased basal NO tone, which translates into a higher level of
nitrosative stress. They further indicate that GSNOR is likely
required to prevent excessive nitrosation of intracellular targets
and that the effects of heat stress are minor compared with the
effects of GSNOR mutation.
Endogenous NO Status Affects Heat Tolerance
The high levels of NO and nitroso products in the hot5 null
mutants suggest that this phenotype is causally linked to the
acquired thermotolerance defects. To test this hypothesis, leaf
discs of wild-type and hot5-2 mutant plants were floated on
MES-KOH buffer containing either of two different NO donors,
sodium nitroprusside (SNP) or DETA/NO, or the NO scavenger
CPTIO, and then treated at 458C for 2 h following pretreatment at
388C (see Methods) (Figure 5A). Under heat stress, SNP led to
severe yellowing and cell death in the wild type and further
enhanced the hot5-2 phenotype. In comparison, when leaf discs
disc yellowing was not observed. Treatment with 10 mM of the
1993]), also increased leaf yellowing in heat-stressed wild-type
plants. Consistent with the involvement of NO/nitroso products
in the heat-sensitive phenotype, 100 mM CPTIO treatment
not only partially restored the appearance of hot5-2 leaf discs
Table 1. Seed Production in Test Crosses with hot5-2
F1, Female Stigma 3 Male PollenSeed No./Silique
Col 3 Col
hot5-2 3 hot5-2
Col 3 hot5-2
hot5-2 3 Col
73.2 6 5.3
17.4 6 6.2
23.1 6 8.2
42.4 6 7.7
Five or six siliques were counted for each F1 hybrid. The data shown are
average values with standard error from the means 6 SE indicated.
GSNOR in Thermotolerance 791
to that of the wild type but also was able to block the effect of
SNP (Figure 5A).
Results of SNP and CPTIO treatments were also quantified by
the measurement of chlorophyll content over time after heat
stress in leaf discs from 25-d-old plants (Figure 5B). Four days
after heat treatment, buffer-treated hot5-2 retained only ;30%
of chlorophyll and SNP-treated hot5-2 was fully bleached. By
buffer alone, and when treated with SNP they retained;60% of
their chlorophyll after 4 d. Treatment with CPTIO dramatically
rescued the hot5-2 chlorophyll loss, with 75% of initial chloro-
phyll content remaining at 4 d after heat treatment. To show that
the effect ofCPTIO was specific tothe hot5 mutantand notjusta
general effect of NO scavenging, we also tested the ability of
CPTIO to rescue the heat sensitivity of the Hsp101 null mutant,
hot1-3. In contrast with hot5-2, the thermotolerance defect of
hot1-3 was not rescued by the NO scavenger, indicating that
the heat-sensitive defect of hot5 is unique and distinct from the
defect in the hot1-3 mutant (Figure 5B).
We next determined whether the phenotypes observed for
the wild type and hot5-2 in the presence of the exogenous
NO scavenger or NO donor correlated with cellular NO status.
Protoplasts were isolated at 2 h after heat treatment from leaf
discs exposed to CPTIO or SNP. Treatment with CPTIO dramat-
ically decreased the level of DAF-FM DA fluorescence in hot5-2
(Figure 5C) compared with buffer alone (Figure 4A). In addition,
the DAF-FM DA fluorescence in the wild type was significantly
increased by SNP treatment compared with buffer alone (Figure
4A). Thus, the heat-sensitive phenotype and NO/nitrosation
levels are correlated.
To confirm that excess NO or metabolites could also explain
the thermotolerance defect of the weak hot5 missense muta-
tions, we examined the effect of treatment with the NO donors
and scavenger on the hypocotyl elongation of heat-treated,
dark-grown hot5-1 seedlings (Figure5D). Treatment ofseedlings
with these agents just before heat stress produced quantitative
differences in subsequent elongation in the dark, consistent with
the results with hot5-2 leaf discs. CPTIO very clearly enhanced
the thermotolerance ofhot5-1seedlings,while SNP, but not KCN,
increased the heat sensitivity of wild-type and hot5-1 seedlings,
and addition of CPTIO with SNP reversed this effect. DETA/NO
treatment also impaired the heat tolerance of both the wild type
and hot5-1, although in addition it reduced hypocotyl growth at
room temperature. In total, these data demonstrate the involve-
ment of excess NO and/or nitrosative stress in the heat-sensitive
phenotype of the missense mutations, confirming that the control
of endogenous NO status is critical for survival of heat stress.
NO Status in the hot5 Missense Mutations Correlates with
The fact that the missense hot5 mutations showed a heat-
sensitive phenotype only as dark-grown seedlings prompted us
to compare the NO status of dark-grown seedlings and 25-d-old
plants of the missense mutants. We first visualized DAF-FM DA
fluorescence in hot5-1 and hot5-3 root tips of seedlings grown
and loaded with dye in complete darkness (Figure 6A; hot5-3
data not shown). Compared with the wild type, both missense
Figure 4. Endogenous NO Status, Total Nitroso Species, and Nitrate
Levels in hot5 Null Mutants.
(A) DAF-FM DA staining for NO and its metabolites. Staining was
performed in the Col wild type and the hot5-2 null mutant from leaves
that were maintained at room temperature or heat-treated before pro-
toplast isolation. NO production and the associated potential for nitro-
sation were visualized in protoplasts stained with DAF-FM DA by
confocal microscopy. Chlorophyll autofluorescence ([a] to [d]), DAF-
FM DA staining ([e] to [h]), and merged images ([i] to [l]) are shown.
Bars ¼ 10 mm. RT, room temperature.
(B) and (C) Total nitroso species (B) and nitrate (C) from wild-type and
hot5 null mutant plants. Values were normalized against total protein
amounts. Data are means of three independent experiments (n ¼ 3 to 4).
792 The Plant Cell
mutants showed much higher levels of NO-related fluorescence
after growth in the dark. The DAF-FM DA fluorescence in hot5-1
was also eliminated by pretreatment of seedlings with CPTIO, as
expected for fluorescence generated from NO (Figure 6A). Fur-
thermore, protoplasts from light-grown, 25-d-old hot5-1 plants
had wild-type, basal fluorescence levels, correlated with the
wild-type heat tolerance phenotype of the missense mutants at
this growth stage (Figure 6B). Light-grown, 2.5-d-old hot5-1 and
hot5-3 seedlings also showed wild-type levels of DAF staining.
Thus, the endogenous NO status of the hot5 missense alleles,
Figure 5. The Relationship of NO Status and Thermotolerance.
(A) As in Figure 1C, leaf discs from the wild type or the hot5-2 mutant were floated on the indicated compounds and either kept at room temperature (RT)
or heat-stressed. The photograph was taken at 5 d after heat stress.
(B) Decline in total chlorophyll in leaf discs of the indicated genotypes following heat treatment with no addition or the addition of CPTIO or SNP as
discussed in the text. At least six leaf discs from separate plants at each sampling time were used.
(C) NO-related fluorescence of protoplasts isolated from leaves treated with CPTIO or SNP.
(D) Intracellular NO status affects the hypocotyl elongation of the hot5-1 mutant grown in the dark after heat treatment. After growth for 2.5 d in the dark,
seedlings were treated with the agents indicated and then heat-stressed for the times shown (after pretreatment at 388C). After an additional 2.5 d in the
dark, hypocotyl lengths were measured and expressed as a percentage of the unheated sample.
GSNOR in Thermotolerance 793
in both the light and dark, correlates with the heat-sensitive
Although we were unable to determine the heat sensitivity of
the hot5-2 and hot5-4 null alleles as dark-grown seedlings, to
determine whether they had the same high DAF-FM DA staining
phenotype as the missense alleles when grown in the dark, null
mutant seeds were grown in the dark to generate root material
(Figure 6C). When stained with DAF-FM DA, these null mutant
roots also showed very high levels of fluorescence (Figure 6D).
Light-grown seedlings of the same age also had high levels of
DAF staining (data not shown). Thus, the missense and null
alleles of hot5 share the inability to regulate NO status with dark-
grown seedlings, further confirming that this phenotype results
from the hot5 mutations.
The NO-Overproducing nox1 Mutant Shows a
Thermotolerance Defect Correlated with NO Status
The observation that endogenous NO status affects acquired
thermotolerance predicts that mutants that overaccumulate NO
would be heat-sensitive. This possibility was tested using the
NO-overproducing mutant nox1 (also known as cue1) (He et al.,
2004), grown both in the dark and in the light, compared with
hot5-1 and hot1-3 as references. When tested for hypocotyl
elongation in the dark, nox1 does not show any defect even after
150 min of 458C heat treatment, although nox1 has a short
hot5-1 plants (Figure 7A). Consistent with the absence of a heat
phenotype, dark-grown nox1 seedlings also did not stain with
DAF-FM DA (Figure 7B). In contrast with this dark-grown phe-
notype, 10-d-old light-grown seedlings of nox1 were defective in
acquired thermotolerance (Figure 7C). Like the hot5 null mutants
(Figure 4A), protoplasts from light-grown nox1 also showed
increased DAF-FM DA levels in the absence or presence of heat
treatment, correlated with the thermotolerance defect (Figure
7D). These data further support the connection between excess
NO-related nitrosation and plant heat sensitivity.
We also tested thermotolerance in the noa1 mutant (formerly
nos1), which produces less endogenous NO (Crawford et al.,
2006), and a nitrate reductase-deficient mutant, nia1/nia, which
exhibits minimal nitrate reduction and must be grown on an
alternative nitrogen source (Wang et al., 2004). Both 2.5-d-old
dark-grown seedlings and 10-d-old light-grown seedlings were
indistinguishable from wild-type seedlings in their heat tolerance
(see Supplemental Figure 4 online).
By analyzing both missense and null mutations of the gene
encoding GSNOR, we have uncovered an important role for this
enzyme in modulating cellular NO levels and nitrosation status in
plants. Specifically, we demonstrated that GSNOR function is
required for acclimation to high temperature and for normal
plant growth and fertility. Previous studies supported the con-
clusion that GSNOR, a type III alcohol dehydrogenase originally
associated with the detoxification of formaldehyde (Uotila and
Koivusalo, 1979; Giese et al., 1994; Martı ´nez et al., 1996; Dixon
et al., 1998; Achkor et al., 2003), acts in plants as well as other
Figure 6. hot5-1 Shows Increased DAF-FM DA Staining Only as Dark-
(A) NO-related fluorescence in roots of dark-grown hot5-1 seedlings with
or without treatment with CPTIO. RT, room temperature; MES, buffer only.
(B) Protoplasts from control or heat-stressed leaves of 25-d-old hot5-1
plants exhibit wild-type, basal levels of NO-related fluorescence.
(C) hot5-2 growth phenotype on nutrient medium in the dark. Seedlings
of hot5-2 were grown in the dark for 3.5 d before staining with DAF-FM
DA in the dark. The first three seedlings are wild type and second three
are hot5-2. The distance between lines is 13 cm.
(D) hot5-2 also has elevated DAF-FM DA staining in roots after growth in
the dark compared with the wild type.
794 The Plant Cell
organisms to metabolize GSNO (Sakamoto et al., 2002; Feechan
et al., 2005; Ruste ´rucci et al., 2007). GSNOR is a potentially
significant playerin themodulation ofcellular NOstatusbecause
it effectively removes GSNO, a compound with NO-generating
and thiol-nitrosating (NOþ-transferring) potential, from the cellu-
lar pool. GSNOR will also act to regulate the availability of GSNO
for glutathiolation reactions, in which it acts by modifying other
cellular thiols, including those on proteins, to form mixed disul-
fides (R-SSG). This reaction has the potential to affect the redox
status and activity of proteins; in addition, it gives rise to the
action profile distinct from that of NO (Fukuto et al., 2005). Thus,
the effects we describe on thermotolerance, plant growth, and
fertility may be mediated by several different pathways or by
multiple mediators acting in concert. Although GSNOR does not
directly act on S-nitrosated protein substrates, GSNOR knock-
levels (Liu et al., 2001, 2004; Feechan et al., 2005; Ruste ´rucci
et al., 2007). Our studies confirm and extend these results,
indicating that GSNO modulates cellular nitrosation status. Con-
sistent with this notion, the nitroso content of leaves from the
hot5 mutants was about twice that of wild-type leaves, and the
fluorescence signal obtained with the NO probe, DAF-FM DA,
was clearly higher in the mutants compared with the wild type.
chemistry following the oxygen-dependent conversion of NO
into RNS and the chemical conversion of the weakly fluorescent
precursor into a more highly fluorescent molecule (Rodriguez
et al., 2005). Thus, a higher fluorescence signal is not neces-
sarily indicative of the presence of free NO but is an integrated
readout of cellular nitrosation chemistry (Rodriguez et al., 2005).
Figure 7. Thermotolerance and NO Status Phenotypes of the nox1 Mutant.
(A) The nox1 mutant exhibits wild-type thermotolerance as 2.5-d-old dark-grown seedlings. The asterisks indicate no growth after heat stress. RT, room
(B) nox1 shows wild-type levels of NO-related fluorescence in roots in the dark. Light microscopy ([a] and [b]), DAF-FM DA staining ([c] and [d]), and
merged images ([e] and [f]) are shown.
(C) The nox1 mutant is defective in acquired thermotolerance as 10-d-old seedlings grown in the light.
(D) NO-related fluorescence in nox1 is high in protoplasts isolated from light-grown plants. Chlorophyll autofluorescence ([a] and [b]), DAF-FM DA
staining ([c] and [d]), and merged images ([e] and [f]) are shown.
GSNOR in Thermotolerance795
as affecting processes controlled by NO-related pathways in
The direct cause of the heat sensitivity of the hot5 mutants is
not known. Assessment of the levels of major HSPs indicated
that GSNOR mutants were not defective in the production of
these protective proteins. The connection of heat sensitivity to
excess nitrosation, however, is demonstrated by several obser-
vations. First, intense NO-related fluorescence staining was
observed in dark-grown seedlings of the HOT5 missense mu-
tants (hot5-1 and hot5-3), which is where the heat-sensitive
phenotype is exhibited, and not in light-grown seedlings, which
are not heat-sensitive. Second, decreasing NO levels with the
NO-scavenger CPTIO partially rescued the heat-sensitive phe-
notype of both dark-grown hot5-1 and hot5-3 and light-grown
SNP and DETA-NO increased the heat sensitivity of wild-type
seedlings and leaves. Finally, the NO-overproducing nox1/cue1
mutant showed NO-correlated thermotolerance defects. These
observations support the hypothesis that elevated levels of
GSNO enhance heat sensitivity due to the perturbation of path-
ways sensitive to reactive oxygen species/RNS, which are likely
already under strain due to heat stress.
Although both the hot5-1 and hot5-3 missense mutants had
reduced GSNOR activity compared with the wild type, it is very
interesting that we did not see a significant difference in GSNOR
activity in the missense mutants when comparing dark-grown
seedlings and 25-d-old plants. The missense mutants accumu-
lated DAF-FM DA–staining species and showed the thermotol-
erance defect only as dark-grown seedlings, suggesting that
GSNOR activity might be lower in the dark than in the light in
dark-grown hot5-2 and hot5-4 null mutants further supports the
idea that this phenotype results from reduced GSNOR activity.
We suggest several possible reasons for the apparent discrep-
ancy between the significant GSNOR activity detected in total
extracts of dark-grown missense mutants and their DAF-FM DA
staining. First, it is possible that there are overall higher levels of
GSNO production/flux in dark-grown seedlings and, therefore,
higher GSNOR activity is required in the dark for the removal of
these species to enable proper growth after heat stress. It is also
possible that in light-grown seedlings other components are
present that compensate for the reduced GSNOR activity in the
mutants and limit the accumulation of excess nitroso species.
Another hypothesis is that the hot5-1 and hot5-3 proteins have
altered regulatory properties compared with the wild-type pro-
tein, being inactive in the dark, and that our extraction conditions
relieve this inactivation (e.g., dissociation of an inhibitor or
removal of a labile inhibitory modification). This interesting phe-
of the tissue- and environment-specific regulation of nitroso
species in plants.
Our data do not suggest that GSNOR is a regulatory player in
thermotolerance; we have no evidence that NO is involved in
heat stress signaling. Although there is one previous report that
NO levels increased during heat stress (Gould et al., 2003), we
did not observe a major heat-dependent increase in NO-related
fluorescence staining in isolated protoplasts. However, we ap-
plied heat stress before protoplast isolation and, therefore,
cannot rule out a transient NO increase in response to heat
stress. We were unable to maintain intact protoplasts when cells
were heat stressed after isolation and stained. Some increase in
DAF-FM DA staining following heat stress could be observed in
roots of dark-grown seedlings, consistent with the previous
report (Gould et al., 2003), but staining was transient and signif-
icantly lower than in the mutants in the absence of stress (see
Supplemental Figure 5 online).
The general importance of GSNOR in plants is emphasized by
its ubiquitous presence throughout the plant. While we have
documented the expression of GSNOR by protein gel blot
analysis in all organs examined, others have visualized its pres-
ence using immunocytochemistry (Barroso et al., 2006; Espunya
et al., 2006). Based on protein blot analysis compared against a
content of leaves accounts for ;0.01% of total protein (see
Supplemental Figure 2 online), and we saw no evidence for a
during development, as well as a range of stresses, is supported
absence of mRNA induction during heat stress has been con-
firmed by our own microarray studies (Larkindale and Vierling,
2008).The lack ofevidence fora significant regulation of GSNOR
at the transcriptional level or the level of protein abundance
suggests that GSNOR is regulated primarily at the posttrans-
criptional level of enzyme activity. We currently hypothesize that
some manner of redox regulation through Cys modification is
one mechanism likely to control GSNOR activity. Notably,
GSNOR has a structural zinc atom coordinated by four fully con-
served Cys residues (see Supplemental Figure 1 online), a redox
regulatory feature of other proteins. In addition, we note that
from E. coli and human, Cys-370 and Cys-284, the latter directly
adjacent to the hot5-1 missense mutation. There are also three
other Cys residues outside of the active site that are common to
eukaryotic GSNOR and that might serve a redox-regulatory role.
Despite the significant role that GSNOR may play in the regu-
lation of nitrosative stress, studies devoted to understanding the
to date has been on the role of GSNOR in pathogen defense
pathways and formaldehyde metabolism (Martı ´nez et al., 1996;
Dixon et al., 1998; Sakamoto et al., 2002; Achkor et al., 2003;
Feechan et al., 2005; Ruste ´rucci et al., 2007). Although Feechan
et al. (2005) worked with the identical null mutation we used in
this study, hot5-2 (gsnor1-3 in their report), they did not report
any morphological or developmental phenotypes of the mutant.
Our observations indicate that balanced GSNO metabolism and
cellular NO/nitrosative status is critical not only for thermotol-
erance but also for normal growth and development under
optimal growth conditions. The most dramatic phenotype we
observed was reduced fertility. In fact, it was most effective to
individual progeny of the heterozygotes were genotyped to
identify homozygous plants for physiological experiments. We
found all pleiotropic phenotypes in the backcrossed hot5-2 null
mutant as well as in a second null mutation, hot5-4, confirming
796The Plant Cell
The hot5 null mutants have more than one defect that leads to
anthers remain below the stigma surface at the time of anthesis
(Figure 2D). Furthermore, results of self-pollination and recipro-
cal crosses to the wild type indicate poor function of both the
male and female gametophytes of hot5 null mutants (Table 1).
Interestingly, release of NO has been proposed as one signal
involved in pollen tube repulsion from the ovule after fertilization
(Johnson and Lord, 2006). This repulsion prevents penetration of
theovule bymorethanonepollentubeandisacritical stepinthe
fertilization process. McInnis et al. (2006) also recently reported
significant levels of NO in pollen and suggested that pollen-
derived NO is important in the pollen–stigma interaction. We
surmise that increased NO and its metabolites in the pollen and/
or ovule lead to this defect, perhaps by interfering with pollen
tube guidance to the ovule.
Consistent with its constitutive expression throughout the
plant, HOT5 appears to be required during the entire life of the
plant. In addition to the fertility defect, hot5 null mutants had an
increased number of flowering stalks, multiple short shoots, and
were long-lived plants that continued to produce leaves even
after wild-type plants had senesced. These phenotypes may be
linked in some way to the reduced fertility. Furthermore, while
hot5 null mutant seeds could germinate on plant growth medium
plates, all further growth was arrested, although seeds could be
germinated in soil to produce mature plants. We were unable to
rescue the germination phenotype of the hot5-2 null mutant on
plant growth medium, including the NO scavenger CPTIO (100
mM to 1 mM) (data not shown). We also observed that hot5-2
was significantly less vigorous and had reduced chlorophyll
when grown under long days (16 h light) (growth conditions used
by Feechan et al.  in studying pathogen resistance). NO is
reported to accumulate in chloroplasts and to stimulate photo-
synthetic electron transport (Zhang et al., 2006). Thus, NO
accumulation could be affected by differences in photoperiod
and might alter chloroplast development and chlorophyll bio-
synthesis. PreviousstudiesindicatethatNObroadly participates
in the plant life cycle, from germination to seedling and mature
plant growth (Beligni and Lamattina, 2000; Bethke et al., 2006;
Zhang et al., 2006), and then decreases in senesced leaves
(Corpas et al., 2006). Thus, GSNOR activity can be expected to
have an effect on all of these processes.
NO is also reported to delay flowering through effects on
both photoperiod and autonomous flowering time determinants
(He et al., 2004). The nox1/cue1 mutation disrupts a chloro-
plast phosphoenolpyruvate/phosphate translocator. The mu-
tant overproduces NO apparently due to high accumulation of
L-Arg, a substrate for NO production (He et al., 2004). The NO-
overproducing nox1/cue1 mutant had a delayed-flowering phe-
notype. By contrast, plants carrying a mutation in NOA1, which
produce less endogenous NO, flowered earlier than wild-type
plants. This evidence suggests that flowering time is altered by
endogenous NO levels (Heetal., 2004).However,wedid notfind
evidence for the alteration of flowering time, despite the obvious
alteration in NO status in GSNOR null plants. Leaf numbers
before bolting were not altered in hot5-2 compared with wild-
type plants under either long or short days. This result may be
due to differences in the NO-derived species present in the
GSNOR mutants compared with the previously studied mutants
with altered NO status.
In addition to increased total NO adducts and dramatically
enhanced NO-related fluorescence staining, an unexpected
finding was that the leaves of hot5 mutants have roughly twice
the concentration of nitrate compared with wild-type leaves.
Thus, the pattern of nitrate content under basal conditions mir-
rors that of levels of nitroso species, suggesting a link between
protein nitrosation and nitrate assimilation. Since the plants
used in the biochemical studies were all grown on the same
cannot beattributedto differences innutrient availability. Rather,
they must result from differences in uptake and transport or in
nitrate consumption along the nitrate assimilation pathway.
However, given the complexity of nitrogen metabolism, its mul-
tilayered regulation and connection to other metabolic pathways
(Stitt et al., 2002; Lamattina et al., 2003), how cellular nitrosation
status may be coupled to nitrate assimilation requires further
Although the major role for GSNOR is now proposed to be in
in formaldehyde detoxification. In plants, one-carbon (C1) me-
tabolism can generate formaldehyde, which can react sponta-
neously with GSH to form S-hydroxymethylglutathione. GSNOR
acting as a FALDH oxidizes S-hydroxymethylglutathione to
S-formylglutathione. It is clear that GSNOR exhibits FALDH
in the aerobic soil bacterium Paracoccus denitrificans is critical
for methyltrophic growth (Ras et al., 1995). A yeast FALDH
deletion mutant (sfa1) showed impaired growth in the presence
in Arabidopsis, overexpression of the FALDH gene (GSNOR)
conferred a high resistance to formaldehyde (Achkor et al.,
2003). However, we do not think that the absence of formal-
dehyde detoxification activity leads to the heat-tolerance defect
of GSNOR mutants, based on the following observations.
S-Formylglutathione produced by GSNOR must subsequently
be hydrolyzed to GSH and formic acid by S-formylglutathione hy-
drolase (SFGH) (Jensen et al., 1998). We analyzed a T-DNA knock-
out mutant (SALK_002548) of the At SFGH gene (At2g41530)
(Kordic et al., 2002). As the SFGH knockout mutation is blocked
in the formaldehyde detoxification pathway, this mutant should
be heat-sensitive if formaldehyde detoxification is the problem
with the GSNOR mutant. However, the SFGH knockout mutant
showed similar phenotypes to wild-type plants in all growth con-
ditions (data not shown). Furthermore, when different concentra-
tions (0.2 mM to 0.2 mM) of formaldehyde were directly applied
to hot5-1 seedling plants in the dark or to hot5-2 leaf discs in the
light, no additional phenotypic defects were observed (data not
shown). These results suggest thatthe heat-sensitivephenotypes
of hot5 are not caused by defects in formaldehyde metabolism.
In addition to the heat-stress phenotype of hot5 null mutants,
photoperiod-dependent phenotypes suggestthatthesemutants
may be sensitive to other abiotic stresses. However, when either
hot5-1 dark-grown hypocotyls or hot5-2 light-grown leaf discs
were treated in salt, cold, or high osmotic conditions, the hot5
mutants were not more sensitive than wild-type plants (data not
GSNOR in Thermotolerance 797
of NO status may be important under stress conditions not yet
In summary, GSNOR regulates cellular nitrosation levels by
metabolizing GSNO, which is a mobile reservoir of NO in plant
cells. Therefore, understanding the mechanisms that regulate
the activity of GSNOR is a critical aspect of the study of the
overall regulation of NO-related signaling and nitrosative stress
Plant Materials and Thermotolerance Assays
Arabidopsis thaliana seeds of the indicated genetic backgrounds and
genotypes were surface-sterilized, planted on nutrient medium plates
(Haughn and Somerville, 1986) containing 0.5% (w/v) sucrose, and kept
at 48C for a minimum of 3 d. Plates were prepared with exactly 10 mL of
medium on a leveling table to ensure even heat treatment and were sealed
after planting with Parafilm to prevent desiccation. Plants were grown in
illuminated growth chambers (;100 mmol?m?2?s?1) on a 22/188C, 12- or
8-h day/night cycle for analysis of growth phenotypes. To obtain mature
hot5-2 and hot5-4 plants, heterozygous seeds were sown directly on soil,
and all plants were genotyped by PCR to identify the homozygotes. Note
thatbecause ofthereducedvigor of nullmutants under longdays,material
for allstressand other physiologicalassaysofthesemutantswas obtained
from plants grown under 12 h of light. For thermotolerance assays, 2.5-d-
old dark-grown and 10-d-old light-grown seedlings were treated as de-
scribed (Hong and Vierling, 2000). For the leaf disc assay, discs (5 mm in
diameter) were punched from third to fifth fully expanded rosette leaves of
25-d-old soil-grown plants and then floated on 2 mL of 10 mM MES-KOH
only, at 388C for 90 min, or at 388C for 90 min, followed by 2 h at 228C and
2.5 h at 458C. Leaf discs were returned to 228C under 12 h of light/12 h of
dark and photographed 5 d later.
We also tested thermotolerance in the noa1 mutant (formerly nos1),
reductase–deficient mutant, nia1/nia, which exhibits minimal nitrate re-
2004). Both 2.5-d-old dark-grown and 10-d-old light-grown mutant
seedlings were indistinguishable from the wild type in their heat tolerance
(see Supplemental Figure 4 online). The nox1 (He et al., 2004) and noa1
(Crawford et al., 2006) mutants were obtained from Z.-M. Pei (Duke
University) and N.M. Crawford (University of California at San Diego),
respectively. The nia1-1/nia2-5 seeds (Col background) were obtained
from the ABRC (CS6512).
NO-Related Chemical Treatments
The DETA/NO (DETA NONOate; diethylenetriamine nitric oxide adduct;
half lifeofNOrelease¼56 hat228Cand20 hat378C)wassynthesized by
Katrina Miranda as described previously (Hrabie et al., 1993).
The NOdonor DETA/NO aswell asthe NOdonor SNP andthe NOscav-
enger CPTIO [2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidadazoline-1-
oxy-3-oxide] were dissolved in 10 mM PBS, pH 7.4, and used to treat
seedlings on plates at final concentrations from 1 mM to 10 mM. KCN (100
For the hypocotyl elongation test, seeds were sterilized and plated in
rows on 2-mL nutrient medium plates containing 0.5% (w/v) sucrose on
3.5-cm circular plates, which were wrapped in foil. Plates were incubated
at 48C for a minimum of 3 d and then placed in a vertical position at 228C
for 3d.Onehour before heat treatment (388C for 90 min followed by 2hat
228C and then 2.5 h at 458C), plates were briefly opened under dim green
light and treated with the NO-related chemicals. Two milliliters of solution
was added on plates, which were placed in a horizontal position for 1 h at
228C. For heat treatment, the remaining solutions were poured out and
the plates were rewrapped in foil. Hypocotyl lengths were measured after
an additional 2.5 d in the dark.
Identification of hot5 Mutant Alleles
The hot5-1mutantwas originally isolated fromanethylmethanesulfonate
Vierling, 2000). For genetic mapping of the hot5-1 mutation, 1024 plants
showing the hot5-1 mutant phenotype were selected. For fine mapping,
different markers on the bottom of chromosome 5 were developed for
simple sequence polymorphism, cleaved amplified polymorphic se-
quence, and single nucleotide polymorphism analyses.
Single knockout mutants of the HOT5 gene were obtained from the
ground) and FLAG (Versailles Genomic Resource Center; FLAG_298F11;
hot5-4 in the Ws background) T-DNA collections using the accession
number of HOT5 (At5g43940) in the database (http://signal.salk.edu/).
Homozygous mutants were identified by PCR analysis using the recom-
recovered from the null mutants, routine experiments were performed by
identifying the homozygous mutants by PCR from among the progeny of
Tilling analysis (in the Col ecotype, carrying the erecta mutation) was
performed on the HOT5 gene, encompassing approximately amino acid
The hot5-3 mutant was recovered as a hypocotyl thermotolerance-
defective mutant from a total of eight missense mutations analyzed.
Both the homozygous hot5-1 and hot5-3 missense mutant alleles were
backcrossed to Col wild-type plants, and one homozygous F3 line for
each mutation was used for phenotypic analyses. For the T-DNA null
mutant alleles, heterozygous hot5-2 or hot5-4 plants were backcrossed
wild-type, heterozygous mutant, and homozygous mutant plants. All
homozygous plants were finally obtained after two backcrosses.
Purification of the HOT5 Protein
A HOT5 cDNA was cloned to the pJC20 expression vector and trans-
formed to BL21(DE3) Escherichia coli cells. HOT5 was overexpressed
with 0.05 mM isopropylthio-b-galactoside in 0.05 mM ZnCl–containing
Luria-Bertani medium. Cells were harvested after overnight induction at
308C and then extracted in 2 mM DTT, 1 mM phenylmethylsulfonyl
fluoride, 0.05 mM ZnSO4, 50 mg/mL DNase I, 2 mM MgCl2, 100 mg/mL
lysozyme, 1% (v/v) Triton X-100, and 20 mM Tris, pH 8.0. Cell extracts
were stirred at 48C for 30 min and thensonicated. The protein was loaded
onto a HiPrep 16/10 DEAE FF column and eluted with a gradient of NaCl
(0 to 200 mM) in 20 mM Tris buffer, pH 8.0. Following concentration, the
HOT5 protein fractions were loaded onto a HiPrep Sephacryl S-100
column equilibrated with 100 mM NaCl containing 100 mM Tris buffer.
Eluted HOT5 fractions were subjected to a final concentration step for
antibody production. Polyclonal antibodies were generated in rabbits by
SDS-PAGE and Protein Blot Analysis
and GAPDH antibodies) were separated by SDS-PAGE on 7.5%
(Hsp101), 10% (GSNOR and GAPDH), or 15% (small HSP) acrylamide
gels and processed for protein gel blot analysis. Protein amounts were
798 The Plant Cell
2003) with BSA as a standard. Protein blots were probed with rabbit
antiserum against HOT5, Hsp101, or the small HSPs Hsp17.6C-I and -II
were probed for cytosolic GAPDH using a GAPC antibody (a gift of Ming-
Che Shih, University of Iowa) as described (Chan et al., 2002). Blots were
incubated with goat anti-rabbit horseradish peroxidase, and bands were
visualized with the enhanced chemiluminescence protein gel-blotting
detection reagent (Amersham International) and BioMax film (Kodak).
Measurement of Nitrate, Nitroso Species, and Chlorophyll Content
Leaf extracts were prepared by homogenization of 150 mg of material in
0.5 mL of PBS containing 10 mM N-ethylmaleimide and 2.5 mM EDTA
and then either immediately centrifuged for 5 min or snap-frozen and
stored in liquid nitrogen for later analysis. The concentration of nitrate in
these leaf extracts was determined by ion chromatography using a
dedicated HPLC system for the simultaneous detection of nitrite and
nitrate (ENO-20; Eicom) following methanol precipitation (1:1, v/v). The
content of nitroso species (comprising SNO and N-nitroso products) in
extracts from hot5-2, hot5-4, and wild-type plants was quantified by
iodine/iodide in glacial acetic acid with subsequent detection of the
released NO by gas-phase chemiluminescence reaction with ozone, as
2006a). Molar concentrations of nitrate and total nitroso species were
normalized for protein content.
Chlorophyll was extracted from individual leaf discs by boiling in 95%
methanol. Chlorophyll concentration was normalized to the fresh weight
of the leaf discs and calculated as described (Lichtenthaler, 1998).
Imaging of NO Status in Arabidopsis Protoplasts
The NO status of seedling roots or protoplasts was visualized by staining
with DAF-FM DA (Arnaud et al.,2006) and confocal microscopy. The third
to fifth fully expanded rosette leaves of 25-d-old Arabidopsis plants were
used for the preparation of protoplasts according to an established
method (Sheen, 1995), with minor modifications (Lee et al., 2007).
Protoplasts were resuspended in 150 mL of 25 mM DAF-FM DA, 0.4 M
mannitol, 15 mM MgCl2, and 4 mM MES/KOH, pH 5.7, and allowed to
incubate for 15 min at 228C in the dark. DAF-FM DA treatment of
minimize light exposure. Plants were treated at 228C only, at 388C for 90
min, or at 388C for 90 min followed by 2 h at 228C and then 2 h at 458C. All
samples for NO visualization were isolated at 2 h after heat treatments
and compared with room temperature treatments.
Leaf tissues and protoplasts were visualized by confocal laser scanning
with a Plan-Apo 633 1.4 lens (numerical aperture). NO-related fluores-
cence after DAF-FM DA loading was captured following excitation at 488
nm and detection at 505 to 570 nm (BP505-570 infrared filter). Auto-
fluorescence of chlorophyll was detected at 645 nm (LP 615 filter) (Lee
et al., 2007). The Zeiss LSM Image Browser 3.2 program was employed
for image acquisition, and Photoshop 6.0 (Adobe Systems) was used for
Measurement of HOT5 Enzyme Activity
GSNOR activity was measured by monitoring the decomposition of
NADH (Jensen et al., 1998; Sakamoto et al., 2002). Oxidation of NADH,
dependent on the presence of the substrate GSNO, was determined
were prepared in 100 mL of 0.05 M HEPES buffer (20% glycerol, 10 mM
MgCl2, 1 mM EDTA, 1 mM EGTA, 1 mM benzamidine, and 1 mM
e-aminocaproic acid, pH 8.0), centrifuged to remove insoluble material,
and then clarified with a desalting column (Zeba desalting column;
Pierce). Enzyme activity was determined at 258C by incubating the
desalted fraction (10 mL) in 180 mL of 0.1 M phosphate buffer containing
10 mL of 6 mM NADH as cofactor and 10 mL of 6 mM GSNO as substrate.
GSNOR activity was monitored for 1min after the addition of NADH using
an Agilent 8453 UV spectrophotometer. The rates were corrected
for background NADH decomposition of each extract containing no
GSNO. Rates were averaged over selected intervals during which the
absorbance decline was linear. Final NADH decomposition values were
normalized against total protein amount. Data are means of three inde-
Sequence data from this article can be found in the Arabidopsis Genome
Initiative and GenBank/EMBL data libraries under accession numbers
At5g43940 and AAB06322 (GSNOR gene).
The following materials are available in the online version of this article.
Supplemental Figure 1. Amino Acid Sequence Alignment of GSNOR
from Arabidopsis (Accession Number AAB06322), Rice (Accession
Number BAD21999), Maize (Accession Number CAA71913), E. coli
(Accession Number NP_414890), and Human (Accession Number
Supplemental Figure 2. Accumulation of GSNOR Protein in the Wild
Type and hot5 Missense Mutants.
Supplemental Figure 3. Chlorophyll Content Depends on Photope-
riod in the Wild Type and hot5-2, and Root Growth Phenotype of hot5
Supplemental Figure 4. Thermotolerance Assay of noa1 and nia/nia2
Supplemental Figure 5. NO-Related Fluorescence after Heat Stress.
We are indebted to Nathan S. Bryan and John Celenza for their assis-
tance during pilot work related to this project, to Maria-Francisca Garcia-
Saura for help with the nitrate measurements, and to Katrina Miranda for
the synthesis of the DETA NONOate as well as for many helpful discus-
sions concerning NO chemistry. We also thank William Montfort and
members of his laboratory for investigations of GSNOR structure and
biochemistry that helped inform this work and Ronan Sulpice of the Max
Planck Institute for Molecular Plant Physiology (Potsdam/Golm, Germany)
for advice on the assay of GSNOR activity in whole plant extracts. This
work was supported by USDA National Research Initiative Competitive
Grants Program Grant 3510014857 to E.V.
Received May 2, 2007; revised February 4, 2008; accepted February 15,
2008; published March 7, 2008.
Achkor, H., Diaz, M., Fernandez, M.R., Biosca, J.A., Pares, X., and
Martinez, M.C. (2003). Enhanced formaldehyde detoxification by
GSNOR in Thermotolerance799
overexpression of glutathione-dependent formaldehyde dehydrogen-
ase from Arabidopsis. Plant Physiol. 132: 2248–2255.
Aracena-Parks, P., Goonasekera, S.A., Gilman, C.P., Dirksen, R.T.,
Hidalgo, C., and Hamilton, S.L. (2006). Identification of cysteines
involved in S-nitrosylation, S-glutathionylation, and oxidation to disul-
fides in ryanodine receptor type 1. J. Biol. Chem. 281: 40354–40368.
Arnaud, N., Murgia, I., Boucherez, J., Briat, J.F., Cellier, F., and
Gaymard, F. (2006). An iron-induced nitric oxide burst precedes
ubiquitin-dependent protein degradation for Arabidopsis AtFer1 fer-
ritin gene expression. J. Biol. Chem. 281: 23579–23588.
Barroso, J.B., Corpas, F.J., Carreras, A., Rodriguez-Serrano, M.,
Esteban, F.J., Fernandez-Ocana, A., Chaki, M., Romero-Puertas,
M.C., Valderrama, R., Sandalio, L.M., and del Rio, L.A. (2006).
Localization of S-nitrosoglutathione and expression of S-nitrosoglu-
tathione reductase in pea plants under cadmium stress. J. Exp. Bot.
Belenghi, B., Romero-Puertas, M.C., Vercammen, D., Brackenier, A.,
Inze, D., Delledonne, M., and Van Breusegem, F. (2007). Metacas-
pase activity of Arabidopsis thaliana is regulated by S-nitrosylation of a
critical cysteine residue. J. Biol. Chem. 282: 1352–1358.
Beligni, M.V., and Lamattina, L. (2000). Nitric oxide stimulates seed
germination and de-etiolation, and inhibits hypocotyl elongation, three
light-inducible responses in plants. Planta 210: 215–221.
Besson-Bard, A., Pugin, A., and Wendehenne, D. (2008). New insights
into nitric oxide signaling in plants. Annu. Rev. Plant Biol. 59: 21–39.
Bethke, P., Badger, M., and Jones, R. (2004a). Apoplastic synthesis of
nitric oxide by plant tissues. Plant Cell 16: 332–341.
Bethke, P.C., Gubler, F., Jacobsen, J.V., and Jones, R.L. (2004b).
Dormancy of Arabidopsis seeds and barley grains can be broken by
nitric oxide. Planta 219: 847–855.
Bethke, P.C., Libourel, I.G., and Jones, R.L. (2006). Nitric oxide
reduces seed dormancy in Arabidopsis. J. Exp. Bot. 57: 517–526.
Chan, C.S., Peng, H.P., and Shih, M.C. (2002). Mutations affecting
light regulation of nuclear genes encoding chloroplast glyceralde-
hyde-3-phosphate dehydrogenase in Arabidopsis. Plant Physiol. 130:
Corpas, F.J., Barroso, J.B., Carreras, A., Quiros, M., Leon, A.M.,
Romero-Puertas, M.C., Esteban, F.J., Valderrama, R., Palma,
J.M., Sandalio, L.M., Gomez, M., and del Rio, L.A. (2004). Cellular
and subcellular localization of endogenous nitric oxide in young and
senescent pea plants. Plant Physiol. 136: 2722–2733.
Corpas, F.J., Barroso, J.B., Carreras, A., Valderrama, R., Palma,
J.M., Leon, A.M., Sandalio, L.M., and del Rio, L.A. (2006). Constitutive
arginine-dependent nitric oxide synthase activity in different organs of
pea seedlings during plant development. Planta 224: 246–254.
Crawford, N.M. (2006). Mechanisms for nitric oxide synthesis in plants.
J. Exp. Bot. 57: 471–478.
Crawford, N.M., Galli, M., Tischner, R., Heimer, Y.M., Okamoto, M.,
and Mack, A. (2006). Response to Zemojtel et al. Plant nitric oxide
synthase: Back to square one. Trends Plant Sci. 11: 526–527.
Delledonne, M. (2005). NO news is good news for plants. Curr. Opin.
Plant Biol. 8: 390–396.
Diaz, M., Achkor, H., Titarenko, E., and Martinez, M.C. (2003). The
gene encoding glutathione-dependent formaldehyde dehydrogenase/
GSNO reductase is responsive to wounding, jasmonic acid and
salicylic acid. FEBS Lett. 543: 136–139.
Dixon, D.P., Cummins, L., Cole, D.J., and Edwards, R. (1998).
Glutathione-mediated detoxification systems in plants. Curr. Opin.
Plant Biol. 1: 258–266.
Dolferus, R., Osterman, J.C., Peacock, W.J., and Dennis, E.S. (1997).
Cloning of the Arabidopsis and rice formaldehyde dehydrogenase
genes: implications for the origin of plant ADH enzymes. Genetics
Espunya, M.C., Diaz, M., Moreno-Romero, J., and Martinez, M.C.
(2006). Modification of intracellular levels of glutathione-dependent
formaldehyde dehydrogenase alters glutathione homeostasis and
root development. Plant Cell Environ. 29: 1002–1011.
Fan, L.M., Zhao, Z., and Assmann, S.M. (2004). Guard cells: A
dynamic signaling model. Curr. Opin. Plant Biol. 7: 537–546.
Feechan, A., Kwon, E., Yun, B.W., Wang, Y., Pallas, J.A., and Loake,
G.J. (2005). A central role for S-nitrosothiols in plant disease resis-
tance. Proc. Natl. Acad. Sci. USA 102: 8054–8059.
Feelisch, M., Rassaf, T., Mnaimneh, S., Singh, N., Bryan, N.S.,
Jourd’Heuil, D., and Kelm, M. (2002). Concomitant S-, N-, and
heme-nitros(yl)ation in biological tissues and fluids: implications for
the fate of NO in vivo. FASEB J. 16: 1775–1785.
Fernandez, M.R., Biosca, J.A., Torres, D., Crosas, B., and Pares, X.
(1999). A double residue substitution in the coenzyme-binding site
accounts for the different kinetic properties between yeast and human
formaldehyde dehydrogenases. J. Biol. Chem. 274: 37869–37875.
Fukuto, J., Switzer, C., Miranda, K., and Wink, D. (2005). Nitroxyl (HNO):
Chemistry, biochemistry, and pharmacology. Annu. Rev. Pharmacol.
Toxicol. 45: 335–355.
Garcia-Mata, C., Gay, R., Sokolovski, S., Hills, A., Lamattina, L., and
Blatt, M.R. (2003). Nitric oxide regulates Kþand Cl?channels in
guard cells through a subset of abscisic acid-evoked signaling path-
ways. Proc. Natl. Acad. Sci. USA 100: 11116–11121.
Giese, M., Bauer-Doranth, U., Langebartels, C., and Sandermann,
H., Jr. (1994). Detoxification of formaldehyde by the spider plant
(Chlorophytum comosum L.) and by soybean (Glycine max L.) cell-
suspension cultures. Plant Physiol. 104: 1301–1309.
Gould, K.S., Lamotte, O., Klinguer, A., Pugin, A., and Wendehenne,
D. (2003). Nitric oxide production in tobacco leaf cells: A generalized
stress response? Plant Cell Environ. 26: 1851–1862.
Grun, S., Lindermayr, C., Sell, S., and Durner, J. (2006). Nitric oxide
and gene regulation in plants. J. Exp. Bot. 57: 507–516.
Guo, F.Q. (2006). Response to Zemojtel et al. Plant nitric oxide syn-
thase: AtNOS1 is just the beginning. Trends Plant Sci. 11: 527–528.
Guo, F.Q., and Crawford, N.M. (2005). Arabidopsis nitric oxide syn-
thase1 is targeted to mitochondria and protects against oxidative
damage and dark-induced senescence. Plant Cell 17: 3436–3450.
Guo, F.Q., Okamoto, M., and Crawford, N.M. (2003). Identification of a
plant nitric oxide synthase gene involved in hormonal signaling.
Science 302: 100–103.
Haughn, G.W., and Somerville, C. (1986). Sulfonylurea-resistant mu-
tants of Arabidopsis thaliana. Mol. Gen. Genet. 204: 430–434.
He, Y., et al. (2004). Nitric oxide represses the Arabidopsis floral
transition. Science 305: 1968–1971.
Hong, S.W., Lee, U., and Vierling, E. (2003). Arabidopsis hot mutants
define multiple functions required for acclimation to high tempera-
tures. Plant Physiol. 132: 757–767.
Hong, S.W., and Vierling, E. (2000). Mutants of Arabidopsis thaliana
defective in the acquisition of tolerance to high temperature stress.
Proc. Natl. Acad. Sci. USA 97: 4392–4397.
Hong, S.W., and Vierling, E. (2001). Hsp101 is necessary for heat
tolerance but dispensable for development and germination in the
absence of stress. Plant J. 27: 25–35.
Hrabie, J.A., Klose, J.R., Wink, D.A., and Keefer, L.K. (1993). New
nitric oxide-releasing zwitterions derived from polyamines. J. Org.
Chem. 58: 1472–1476.
Huang, X., Stettmaier, K., Michel, C., Hutzler, P., Mueller, M.J., and
Durner, J. (2004). Nitric oxide is induced by wounding and influences
jasmonic acid signaling in Arabidopsis thaliana. Planta 218: 938–946.
Jensen, D.E., Belka, G.K., and Du Bois, G.C. (1998). S-Nitrosogluta-
thione is a substrate for rat alcohol dehydrogenase class III isoen-
zyme. Biochem. J. 331: 659–668.
800 The Plant Cell