genetic differences in seed longevity of various arabidopsis mutants

Genetic differences in seed longevity of various Arabidopsis mutants

Emile J. M. Clerkxa, Hetty Blankestijn-De Vries

a, Gerda J. Ruys

a, Steven P. C. Groot

band Maarten Koornneef

a,*

aWageningen University, Laboratory of Genetics, Wageningen, The NetherlandsbPlant Research International, Wageningen, The Netherlands*Corresponding author, e-mail: [email protected]

Received 5 November 2003; revised 26 February 2004

Seeds gradually lose their viability during dry storage. Thedamage that occurs at the biochemical level can alter the seed

physiological status and is affected by the storage conditions

of the seeds. Although these environmental conditions control-

ling loss of viability have been investigated frequently, littleinformation is available on the genetics of seed longevity.

Using Arabidopsis mutants in defined developmental or bio-chemical pathways such as those affected in seed coat compo-sition, seed dormancy, hormone function and control of

oxidative stress, we tried to gain insight into the genes and

mechanisms controlling viability of stored seeds. Mutations

like abscisic acid insensitive3 (abi3) as well as abscisic aciddeficient1 (aba1) show reduced longevity, which may be

partially related to the seed dormancy phenotype of these

mutants. Mutants with seed coat alterations, especially

aberrant tests shape (ats), showed a stronger reduction in

germination percentage after storage, indicating the import-ance of a ‘functional’ seed coat for seed longevity. A specific

emphasis was placed on mutants affected in dealing with

Reactive Oxygen Species (ROS). Because several pathways

are involved in protection against ROS and because generedundancy is a common feature in Arabidopsis, ‘double’mutants were generated. These ‘double’ mutants and the

corresponding single mutants were subjected to a controlleddeterioration test (CDT) and a germination assay on hydro-

gen peroxide (H2O2) after prolonged storage at two relative

humidities. CDT and germination on H2O2 affected all geno-

types, although it appears that other effects like geneticbackground are more important than the deficiencies in

the ROS scavenging pathway. Explanations for this limi-

ted effect of mutations affecting ROS scavenging are

discussed.

Introduction

Seeds of good quality are undamaged and have a highgermination percentage prior to and after storage, andtherefore will produce vigorous seedlings without defectsunder various environmental conditions (Dickson 1980).Very little is known about the genetic basis of differencesin seed quality because this trait is strongly affected byenvironmental factors, during seed formation, harvestand storage. This can be illustrated by geneticallyidentical seed lots in which individual seeds, even whengrown under identical conditions or even when comingfrom the same plant, may lose their viability at differentintervals after harvest. Genetic studies require differenceswithin the same species. The most obvious difference inlongevity can be found between orthodox and recalci-trant seeds. Both orthodox and recalcitrant seeds canbe found within the genus Acer (Greggains et al. 2000),although a tree species is not very attractive for geneticresearch. Differences are also found among orthodox

seeds. One of the earliest reports on the genetics of seedquality is by Lindstrom (1942), who found differences inseed longevity in different F1 hybrids of maize. Geneticdifferences for longevity after open storage can beobserved in maize varieties where hard flint and dentvarieties remain viable longer then starchy or sweet vari-eties. However, in closed storage, at fairly constantmoisture contents, few differences were evident (Bewleyand Black 1994). Lyall et al. (2003) found that nearisogenic lines of pea, only differing in the rugosus alleles,differed in longevity. A more comprehensive report onthe genetics of seed quality is by Dickson (1980). Heconcluded that many of the traits influencing seed vigourare of a quantitative genetic nature.Seed quality can be reduced on the parental plant due

to adverse environmental conditions, premature germina-tion (Coolbear 1995) and pathogens (McGee 2000). Phy-siological damage can be of different types, e.g. short-term

PHYSIOLOGIA PLANTARUM121: 448–461. 2004 doi: 10.1111/j.1399-3054.2004.00339.x

Printed in Denmark – all rights reserved Copyright#PhysiologiaPlantarum2004

448 Physiol. Plant. 121, 2004

deterioration in the field is different from long-term dete-rioration during storage, which in turn is different fromsustained mechanical damage (McDonald 1999). All seedorgans can suffer physiological damage: the seed coat,which is of maternal origin, the embryo and the endo-sperm. Damage can be sustained by the chemical consti-tuents of seeds by the way these compounds interact toform biological structures. The integrity of DNA, proteinsand membranes is especially important for maintainingseed viability.Tolerance to desiccation is a survival mechanism

adopted by orthodox seeds (Wilson 1995) and developsduring the early phases of seed maturation after morpho-genesis is completed (Bewley and Black 1994). Duringthe last period of seed maturation, well beyond themoment desiccation tolerance has been acquired, seedsstill gain in longevity (Hay et al. 1997, Jalink et al. 1998).Upon dry storage, seeds gradually lose their viability.Seeds that have initiated germination processes, becauseof vivipary or during seed priming, also exhibit reducedlongevity (Alvarado and Bradford 1988, Buitink et al.2000). Mutants that have a defective seed developmentsuch as leafy cotyledon1 and 2 (lec1, lec2), fusca3 (fus3)and extreme alleles of abi3 exhibit an extremely shortlongevity, since they lose their viability within a fewweeks of dry storage. In general the speed of viabilityloss depends very much on the storage conditions; espe-cially seed moisture content, which is dependant on therelative humidity (RH) and temperature of the environ-ment (Ellis et al. 1990, Bewley and Black 1994). Hay et al.(2003) predicted that in dry and cold conditions, 5%moisture content and �20�C, Arabidopsis seeds shouldbe storable for approximately 2000 years.Although the environmental conditions that control

loss of viability and vigour have been investigatedfrequently, little information is available on the exactmechanisms that control this loss of viability. Most datain the literature concerning this process comes fromstudies where loss of viability was correlated with biochem-ical changes and also by assuming that certain cellularstress generating processes are important (Coolbear 1995).The use of mutants related to defined developmental

or biochemical pathways may be helpful in elucidatingwhich processes are more and which are less importantfor seed longevity. Here we have used a number ofArabidopsis mutants to study seed quality during storage.Mutants were chosen based on their effect on seeddormancy, hormone biosynthesis or hormone signaltransduction, on their effect on accumulating or avoidingoxidative stress, or because they have a defect in theirseed coat. Hormones may influence germination/dormancy as shown for abscisic acid, gibberellin, andethylene. Furthermore, the response of plants toenvironmental stress is influenced by hormones, asshown by the response to oxidative stress by jasmonicacid and ethylene (Overmyer et al. 2003). Seed coatmutants were included in this analysis because Debeaujonet al. (2000) described reduction in seed longevity for suchmutants.

A specific emphasis was put on mutations affectingoxidative stress because it is often reported that a com-mon cause of reduced seed longevity might be the pro-duction of free radicals (Hendry 1993, McDonald 1999),which damage cellular components. Free radicals can begenerated by carbon, sulfur, phosphorus, nitrogen, ironand manganese metabolism and may owe their biologicalorigin ultimately from the radical promoted transfer ofelectrons to and from oxygen (Hendry 1993). ReactiveOxygen Species (ROS) are superoxide and hydroxylmolecules and, although not a true radical, H2O2,because it can provide hydroxyl molecules through theFenton reaction. ROS are generated in both stressed andunstressed cells, and in various cellular compartments.They are generated endogenously during certain devel-opmental transitions such as seed maturation and as aresult of normal, unstressed photosynthetic and respira-tory metabolism (Grene 2002). In seeds the most likelyorigins are the peroxisomes, the mitochondria, autoxida-tion in the cytosol, and chlorophyll, which is normallydegraded during maturation. ROS could directly, or viasubsequent lipid peroxidation, lead to mitochondrialdysfunction, enzyme inactivation, membrane perturba-tion and genetic damage (Coolbear 1995). In additionto the amount of ROS, the final damage is alsodetermined by the efficiency by which the cells in whichthese reactive molecules are formed, are able to scavengeand dispose of them.The antioxidant defence system in plants has both

enzymatic and nonenzymatic components. It should berealized that the enzymatic antioxidant systems can onlybe active under conditions of sufficient water. In thequiescent period when seeds are dehydrated, only mole-cular antioxidants (e.g. glutathione, ascorbate, polyols,carbohydrates, peroxyredoxin, tocopherol and pheno-lics) can alleviate oxidative stress (Hoekstra et al. 2001).Superoxide dismutase (SOD) (Kliebenstein et al. 1998,

Grene 2002) catalyses the conversion of superoxide radi-cal to H2O2. Hydrogen peroxide can be disposed of bycatalase and ascorbate peroxidase. Catalases convertsH2O2 to water and oxygen, ascorbate peroxidase formswater and dehydroascorbate from ascorbic acid (vitaminC) and H2O2 (Willekens et al. 1995, Noctor and Foyer1998, Blokhina et al. 2003). Furthermore H2O2 removalvia ascorbate peroxidase requires glutathione becauseascorbate is reduced by dehydroascorbate reductaseusing glutathione (GSH) as a reducing substrate. Thisprocess is known as the ascorbate–glutathione cycle(Noctor and Foyer 1998). Other antioxidants such asvitamin E (a-tocopherol) act against phospholipid radi-cals (Grene 2002). Phenolic compounds, among whichflavonoids, are abundant in plant tissues (Grace andLogan 2000) and their antioxidant properties arise fromtheir high reactivity as a hydrogen or electron donor andfrom the ability of the polyphenol-derived radicals tostabilize and de-localize the unpaired electron. Anotherfunction lies in their ability to chelate transition metalions and thus terminate the Fenton reaction (Rice-Evanset al. 1997). The seed coat itself, which in its mature

Physiol. Plant. 121, 2004 449

state, contains tannins that are oxidized flavonoid poly-mers, may play a role in depriving the embryo of oxygenduring storage and thus preventing ROS formation (Cor-bineau and Come 1993).The study of ROS mediated damage and repair in seeds

is complex, especially when they are in a dehydrated state.Adding water to dry seeds could start germinationprocesses, which subsequently activates the antioxidantdefence mechanism. In a large number of publicationscorrelative evidence is provided that both the amount ofoxidants and the antioxidant activity correlates with seedquality. Senaratna et al. (1988), showed that aged soybeanaxes had a low antioxidant potential when imbibed, indi-cating that the ageing process was associated with expo-sure to oxidative stress. Puntarulo et al. (1991), showed anincrease in both oxidants and antioxidants during germi-nation of soybean and similar observations were madeduring germination in Pinus pinea L. (Tommasi et al.2001), wheat (Cakmak et al. 1993), and Zea mays L.(Leprince et al. 1994). Recalcitrant seeds, which are desic-cation intolerant, contain higher amounts of ascorbic acidcompared to orthodox seeds, which might also be anindication for the importance to scavenge free radicals toprevent seed ageing. Apparently moist recalcitrant seedsneed a constant high level of protection while dry ortho-dox seeds have a low metabolic activity and hence arelatively low production of hydrogen peroxide (Tommasiet al. 1999). Bailly et al. (1996) showed that the fasterdeterioration of sunflower seeds during acceleratedageing, compared to nonaged controls, was related to adecrease in enzymes involved in ROS scavenging.Plant hormones play a role in programmed cell death

(PCD) mediated by oxidative stress and the systems usedto elucidate the role of hormones in this process are thehypersensitive response to pathogens and the exposure toozone (O3) (Overmyer et al. 2003). Plant hormones likeethylene and salicylic acid enhance the accumulation ofROS in PCD and in ROS dependent lesion propagation,whereas jasmonic acid is involved in ROS containment,as was concluded from the observation that ethyleneresistant1 (etr1) mutants are more tolerant, while jasmonicacid resistant1 (jar1) mutants are more sensitive to ozone(Overmeyer et al. 2003). However, it has also beenreported that ethylene, salicylic acid and abscisic acid areinvolved in protection of plants against heat inducedoxidative damage because mutants like etr1, abi1 and thesalycilic acid deficient (nahG) mutant are more sensitive tothis stress (Larkindale and Knight 2002). Abscisic acid,besides its involvement in heat tolerance, plays a centralrole in stress signalling in both biotic and abiotic stresses(Xiong et al. 2002).Here we report the study of mutant seeds that have

been stored under ‘normal’ laboratory conditions (ambienttemperature and RH) for 4 years and in addition, a sub-group has been tested with a Controlled DeteriorationTest (CDT), developed for Arabidopsis by Tesnier et al.(2002) that mimics natural ageing. The effect of ROS wastested using the frostbite1 (fro1) mutant, recently describedas having high constitutive levels of ROS (Lee et al. 2002)

and by germination on H2O2. Superoxide radicals arebelieved not to be able to pass biological membranes(Grene 2002) and are very short lived. Therefore, a germi-nation assay on H2O2 was used; H2O2 is normally formedas a result of SOD action and is capable of diffusing acrossmembranes (Grene 2002).A specific problem that we encountered in this genetic

approach is that the various mutants are in differentgenetic backgrounds. Different accessions show differentseed quality properties as was shown by their response toa CDT indicating the existence of genetic variation forthe response to this test (Bentsink et al. 2000, Tesnieret al. 2002, Clerkx et al. 2004).

Materials and methods

Plant material and selection of ‘double’ mutants

The defects, their respective wild types and references forthe mutants used in this study are presented in Table 1.F1 seeds of Ler� ats and ats�Ler were generated bycrossing the respective mutant and wild-type line, thecontrol Ler and ats seeds, used in the same experiments,were all derived by hand pollination. The frostbite 1(fro1) mutant was isolated in the RD29A::LUC back-ground this reporter line emits bioluminescence inresponse to low-temperature, ABA or NaCl treatment.The fro1 mutant was isolated from an EMS treatedRD29A::LUC M2 population because it showed alower level of luminescence under low temperature treat-ment. Characterization of the fro1 mutant showed aconstitutive accumulation of ROS (Lee et al. 2002).‘Double’ mutants, with oxidative stress scavenger

defects, were obtained by crossing the individual mutantsamong each other. ‘Double’ mutants were selected from atleast 200 individual F2 plants derived from these crosses.Selection of the ‘double’ mutants was either done basedon their phenotype (tt4-1 and ats) or with the use ofmutant allele specific PCR-markers. Table 2 contains allmarker information. A CAPS marker identifying CATA-LASE 1 was developed based on the CATALASE 1sequence (catalase 1:GenBank accession no. U43340).The CATALASE 3 primers are described by Frugoliet al. (1996). The catalase double mutant has a largedeletion (R. McClung, personal communication), whichenables the identification of this mutant by the absence ofamplification of both a PCR products for CAT1 andCAT3, while in the wild type both products are present.For the glutathione deficient mutant cad2-1, a CAPSmarker was developed on the basis of the published muta-tion (Cobbett et al. 1998), which amplified part of thegene and for which a subsequent digestion, with BslI at55�C, distinguished the mutant from the wild-type allele.No CAPS marker could be made to differentiate betweenthe vtc1-1 and VTC1 alleles and therefore the point muta-tion (Conklin et al. 1999), was used to develop a dCAPSmarker (Neff et al. 1998), for which the enzyme BamHIwill cleave in the product amplified in the vtc1-1 mutantsbut not that from wild type plants.


The DNA extraction procedure used for mutantspecific marker analysis made use of flower buds, whichwere harvested in liquid and frozen nitrogen andthereafter ground. DNA was isolated according to theBernatzky and Tanksley (1986) protocol, adapted for

rapid extraction of small quantities. Extraction solutionwas made of 125ml extraction buffer (0.35M Sorbitol.100mM Tris, 5mM EDTA, pH7.5 (HCl)) together with175ml lysis buffer (200mM Tris, 50mM EDTA, 2M NaCl,2% (w/v) cetyl-trimethyl-ammonium bromide), to which

Table 2. Primer names and primer sequences used to identify mutants with no obvious phenotype.

Primer name primer 1 (50? 30) primer 2 (50? 30) Ta (�C) endonuclease used type

cat1 ccgagactctcagagatc atcaaggatcgtgcgtctg 54 CAPSvtc cttgagaccattgactctcagga gaggaccagcggtacctagtgg 60 BamHI dCAPScad aggtgacaagatcattggtc caaacctataccagataagaac 54 Bs I CAPS

cat1 is used to identify the catalase mutant, vtc and cad primers are used to identify, respectively, the ascorbic acid (vtc1-1) and glutathionedeficient (cad2-1) mutants. Ta: annealing temperature used for specific amplification, CAPS markers were amplified in 35 cycles of 30 s 94�C,30 s Ta and 90 s 72�C. The vtc dCAPS was amplified in 35 cycles of 30 s 94�C, 30 s Ta and 60 s 72�C.

Table 1. Mutants used, their genetic background and reference

Mutant geneGeneticbackground Gene encoded/defect Phenotype Reference

fro1 (frostbite1) C24 NADH dehydrogenasesubunit of mitochondrialrespiratory chain complex

constitutive levels of ROS Lee et al. 2002

RD29A::LUC C24 – control for fro1 Lee et al. 2002

aba1-5 Col zeaxanthin epoxidase ABA deficient Leon-Kloosterziel et al.,1996b; Rock and Zeevaart1991, Meyer et al. 1984

C3-7-1 Col unknown reduced dormancy Raz unpublishedetr1 Col ethylene receptor with

histidine kinaseactivity

ethylene resistant Bleecker et al. 1988,Chang et al. 1993

jar1-1 Col adenylate forming enzyme jasmonic acid resistant Staswick et al. 1992,Staswick et al. 2002

vtc1-1 Col GDP-mannosepyrophosphorylase/accumulates only 30%ascorbate

vitamin-C deficient Conklin et al. 1996,Conklin et al. 1999

cad2-1 Col y-glutamyl-cysteinesynthase/ accumulates only15-30% glutathione

cadmium sensitive Howden et al. 1995,Cobbet et al. 1998

abi3-1 Ler B3 domain protein withB1 and B2 domain

abscisic acid insensitive Koornneef et al. 1984,Giraudat et al. 1992


abscisic acid insensitive Ooms et al. 1993,Giraudat et al. 1992


abscisic acid insensitive Bies-Etheve et al. 1999,Giraudat et al. 1992

ats Ler unknown/lacks 2 outof 5 integuments

aberrant testa shape Leon-Kloosterziel et al. 1994

gai Ler GRAS familytranscription factor

gibberelin insensitive Koornneef et al. 1985,Peng et al. 1997

rdo1 Ler unknown reduced dormancy Leon-Kloosterziel et al. 1996a

rdo2 Ler unknown reduced dormancy Leon-Kloosterziel et al. 1996a

rdo3 Ler unknown reduced dormancy Peeters et al. 2002

tt4-1 Ler chalcone synthase deficient/lacks browntannins

transparent testa Shirley et al. 1995

tt3-1 Ler dihydroflavenol-4-reductase/lacks brown tannins

transparent testa Shirley et al. 1995

ga1-3 Ler copalyl diphosphate synthase GA deficient Koornneef and van deVeen 1980,Sun et al. 1992

cat1 cat3 Ws lacks sequence for both genes catalase deficient Salome and McClung 2002


30 ml sarkosyl (10% w/v) was added. The mixture ofground plant material and extraction solution wasincubated for 30min at 65�C with occasional shaking.Hereafter a solution of 400ml chloroform/isoamyl alcohol(24:1 v/v) was added and the suspension vortexed. Aftercentrifuging for 5min at maximum speed in an Eppendorfcentrifuge the aqueous phase was transferred to a newtube. An equal amount of isopropanol was added, andthe DNA was precipitated by carefully inverting the tubeand after 10min centrifugation at maximum speed inan Eppendorf centrifuge the water-alcohol mixture wasdiscarded and the pellet washed with 70% cold ethanol.The pellet was left to dry and was dissolved in watercontaining RNAse A and incubated 30min at 37�C,thereafter it was stored at 4�C.

Seed production, storage and germination conditions

Seeds were sown in Petri dishes on water-saturated filterpaper and incubated in a growth chamber at 25�C. After2 days of incubation, germinated seeds were transferredinto soil and cultivated in an air-conditioned greenhouse(18–23�C) in a 16 h photoperiod.For genotype comparison two sets of seeds were used;

one set consisted of seeds harvested on monogenichormone and dormancy mutants and the other set wascomposed of the progeny of F3 seeds of selected ‘double’mutant plants together with the monogenic mutants. Inboth experiments the various genotypes were growntogether in four randomized replications, each of fiveor six plants. Seeds were harvested from mature drysiliques and bulked per replication and stored for4 years in unsealed polyethylene bags at ambient labora-tory conditions (temp 15–20�C, 30–70% RH) for naturalageing. In the double mutant experiment harvestedseeds were stored at ambient temperature in unsealedpolyethylene bags for 11 months and subsequently trans-ferred to incubators above saturated solutions of differ-ent salts CaCl2 (20

�C; 32% RH) and CaNO3 (20�C; 60%

RH) to obtain two ageing regimes.Although the germination percentage of freshly har-

vested seeds was not determined, we know from experi-ence that after harvest all seeds are viable and germinatefully after removal of dormancy by cold treatment orafter-ripening. For the set of monogenic mutants 40–80seeds from each of the 4 replicates were sown in Petridishes with 1� 10�5 M GA417. Germination percentagewas scored after 5 days incubation at 4�C and subse-quent incubation for 7 days in a growth chamber (25�C,16 h light period). For the double mutant experimentseeds, which were left to age for 1 year at 32 and 60%RH, respectively, were used in a controlled deteriorationtest (CDT, see below) and sown on 0, 0.3 (¼ 100mM) and0.5% (¼ 166mM) H2O2. Germination percentages, of40–80 seeds of three replicates, from bulked seeds of5 plants, were scored after 5 days incubation at 4�Cand subsequent incubation for 7 days in a growth cham-ber (25�C, 16 h light period).

Controlled deterioration tests

Seeds from the double mutant experiment (see above)that had been stored for either 9 months at ambientconditions (CDT1) or 9 months at ambient conditionsand 2 months at 32% RH (CDT2) were used in a CDT.CDT was performed according to Tesnier et al. (2002).Seeds are equilibrated at 85% relative humidity (15�C),immediately after equilibration 0 day controls are driedback at 32% relative humidity resulting in seed moisturecontent of approximately 6% (Tesnier et al. 2002). Treat-ment consists of storing the seeds (at 85% RH) for anumber of days at 40�C (CDT 1 and CDT2, 1, 3, 5 and 7days) after this these seeds are also dried back at 32%RH (20�C) and stored at c. 4�C for CDT1 approximately2 months and for CDT2 1 week, until germination wastested. Three replications of 50 seeds were tested per linefor each day of treatment. Final germination percentagewas determined after 14 days without (CDT1) or with(CDT2) cold pretreatment of imbibed seeds (7 days 4�C).Seeds of the F1 Ler� ats and ats�Ler together with

the controls were treated for 0,1,2 and 4 days at 40�Cand 85% RH as described above. Two replicas of 60–100seeds were sown without cold treatment and finalgermination percentage was determined after 14 days.Seeds had been stored for 10 months under ambientconditions.Seeds of the fro1 mutant and controls were harvested

from mature dry siliques and stored under ambient con-ditions until use. CDT was performed with 3 replicas of40–100 seeds, seeds were treated for 0 and 1 day at 40�Cand 85% RH as described above, final germinationpercentage was determined after 14 days.

Statistical analysis

Comparisons of genotypes for their longevity was per-formed with samples that had received the treatmentsimultaneously and not between treatments. Longevitywas estimated as a single parameter: LD50 the lethal doseof treatment necessary to kill 50% of the seeds. For thisonly germination data of mutant and wild type seeds thatdid not show dormancy could be used. Germinationproportions were transformed to probit values andplotted on a time scale, in days, applying the regressionmodule of the statistical package SPSS version 11.0.1(SPSS Inc., Chicago, IL). Thereafter, these data werefurther analysed, using the general linear model moduleof the same program, and P-values were adjusted(Bonferoni correction) for multiple comparisons withinthe same data set.

Results

Longevity of seed lots stored under ambient conditions

A germination test was performed with simultaneouslyproduced seed lots of a number of monogenic mutantsand corresponding wild type lines, stored together for


4 years under ambient conditions. The seeds showeddifferences in survival (Fig. 1). All seeds had been pro-duced in the same greenhouse under the same conditionsand were harvested at the same time except the threeLandsberg erecta (Ler) seed lots, which had been har-vested a few weeks earlier. After a 4-year storage periodsome of the Ler seeds failed to complete germination. Themost severely affected Ler mutants were those with thedifferent abi3 alleles. Significantly reduced germinationpercentage was also observed for the ats mutant in Lerbackground, which confirmed earlier observations in ourlaboratory. The decreased germination percentage ofother mutant seeds like gibberellin insensitive (gai), reduceddormancy2 (rdo2) and gibberellin deficient1-3 (ga1-3) wasnot significantly different from the wild-type Ler seeds.Columbia (Col) seeds were hardly affected by the 4-yearstorage. However, in the Col background, a decreasedgermination percentage, was observed for C3-7-1 (reduceddormancy), etr1 and aba1-5, although the reduction wasonly significant for the latter genotype. These results indi-cated an important role for ABA in seed longevity, andpossibly a role for seed dormancy, which is also stronglyreduced in ABA related mutants. No effect was observedfor the two mutants affected in ROS scavenging vitamin-Cdeficient1-1 (vtc1-1) and cadmium sensitive2-1 (cad2-1), thelatter being glutathione deficient.

The effect of ROS accumulation

To investigate if ROS induced damage may be an import-ant component of reduced viability upon Arabidopsisseed storage, a CDT was performed with the fro1 mutantof which the leaves accumulate ROS constitutively. Totest what the effect of this mutation was on CDT survivalthe mutant and controls were subjected to 1 day ofcontrolled deterioration treatment (Fig. 2). It appearedthat for this mutant the total number of germinatedseeds decreased significantly (P, 0.05), and the numberof abnormal seedlings as percentage of total germinationincreased compared to the corresponding controls andwas therefore more sensitive to the treatment.

The effect of CDT on mutants and ‘double’ mutants with

antioxidant related phenotypes

To analyse the effect of various mutations on seed qualityin detail, a CDT was performed with seeds harvested frommature dry siliques that had been stored at ambient con-ditions for 9months. Because residual dormancy was notexpected, no dormancy breaking treatment was appliedprior to germination. The results however, showed that inseeds of Wassilewskija (Ws) and of the double mutantcat1 cat3, in Ws background, residual dormancy wasstill present because the germination percentage increased

Fig. 1. Average seed germination percentages (� SE) of 4 year stored seed lots stored under ambient conditions of hormone, dormancy andoxidative stress related mutants and their respective wild-types. Mutants in Ler background (left panel) and Ler wild type are depicted as greybars and mutants in Col background (right panel) and Col wild-type are depicted in black bars. Each mutant indicated with a common lettercould not be separated statistically (P, 0.05).

Fig. 2. Germination percentages (� SE) (A) and percentage of abnormal seedlings � SE of total germination percentage (B) of wild-type (C24),RD29A::LUC and fro-1 (A) after 0 (grey bars) and 1 (black bars) day CD treated seeds.


with the duration of the CD treatment (CDT1, Fig. 3A).The same effect was observed in Col and the vtc1-1 andcad2-1 mutants in Col background. To test whether thisreduced germination prior to CDT and progressivelygreater percentage germination with increasing durationof the treatment was due to residual dormancy, orinduced by the treatment, several mutants and wild typeswere tested for the effect of a cold treatment. This experi-ment showed that some dormancy was still present in anumber of these lines, but it also showed that 1-weekof cold treatment was sufficient to break the residualdormancy (data not shown). A second CDT (CDT2,Fig. 3B) was performed and a dormancy breaking treat-ment was applied by imbibing the seed for 7 days at 4�Cprior to the germination assay. Again the Ws and the cat1cat3 mutant showed reduced germination, whereas germi-nation increased with prolonged treatment to a pointwhere the germination percentage decreased due to accu-mulating damage. The initial increase was less than with-out cold treatment. The cold treatment was enough tobreak dormancy of Col and the vtc1-1 and cad2-1 mutantsin Col background. Because Ws and cat1 cat3 mutantscompleted full germination after cold treatment in a ger-mination assay prior to CDT2 (data not shown) it is likelythat the observed dormancy was induced by the treat-ment. This is probably an effect of the temporary increasein seed moisture content, since with the 0 days CDT

dormancy was induced without storage at 40�C. Induceddormancy was observed in both CDT1 and CDT2 and theeffect of the CDT was obvious for all genotypes. Theusefulness of a CDT for analysis of genetic factors influ-encing seed longevity became clear when the germinationpercentage of vtc1-1, cad2-1, Col and Ler after CDT wascompared to germination percentages after 4 years ofnormal storage (Fig. 1). The CDT reduced the viabilityof these genotypes rapidly while a 4-year storage onlyslightly decreased viability. For statistical analysis dete-rioration was estimated as a single parameter: LD50, thenumber of days of deterioration treatment required todecline to 50% germination. Some ‘double’ mutantsshowed a significant reduction in LD50 values comparedto their respective wild type or single mutant controls(Table 3, CDT1 and CDT2). As a result of difficultieswith induced dormancy, the double mutants could notbe compared to all controls.

The effect of CDT on the ats mutant

After 4 years of storage the ats mutant seeds showed asignificant reduction in germination percentage com-pared to the Ler wild-type, although in the CDT thereduction was not significant (Table 3, CDT1). To testwhether the reduction in germination percentage, in atsseeds, was embryonic or due to the maternally inherited

Fig. 3. Germination percentages (� SE) after 0, 1, 3, 5 and 7 days of controlled deterioration treatment of CDT1: (A) stored 9 months atambient conditions (CDT2) and CDT2; (B) 9 months at ambient conditions and 2 months at 32% RH. Colours indicate the geneticbackground of the various wild types mutants and ‘double’ mutants, Col: white, Ler: grey and Ws: black, mixed graphs indicate mixedbackgrounds: grey/white is Ler/Col, grey/black: Ler/Ws and white/black: Col/Ws.


seed coat defect, reciprocal F1 seeds between Ler and atswere made. The F1 and control seeds were subjected to aCD treatment. Because the seed coat is of maternalorigin, F1 seeds had either the wild type seed coat (F1

Ler� ats) or the mutant seed coat (F1 ats�Ler) whilethe embryo is heterozygous in both cases. The deteriora-tion pattern of the F1 Ler� ats seeds more resembledthat of the Ler seeds, whereas that of the F1 ats�Ler

resembled that of ats seeds (Fig. 4). Therefore it seemedlikely that the aberrant seed coat increased the sensitivityto CDT.

The effect of germination on hydrogen peroxide (H2O2)

It could be expected that mutants with a defective anti-oxidant function should be more sensitive to the inhibit-ing effect of a ROS generating compound such as H2O2.Seeds that had been stored for 11 months at ambientconditions and thereafter for one year either at 32% or60% RH were imbibed and left to germinate at twoconcentrations of H2O2 (Fig. 5A,B). Most lines showed100% or almost 100% germination upon imbibition onwater. There was a slight reduction in the cat1 cat3double mutant seeds and in seeds of lines that had theats mutant phenotype. Imbibition and subsequent germi-nation on H2O2 strongly reduced germination of allseeds, especially when 0.5% H2O2 was applied. Storageat 60% RH renders seeds more sensitive to H2O2. Theats mutant was very sensitive to H2O2, even the lowestconcentration 0.3% H2O2 (grey bars in Fig. 5A,B)entirely prevented the completion of germination. Thiseffect was also observed in all ‘double’ mutants with ats,where the effect of the aberrant testa is so strong that theeffect a second mutation cannot be seen.

Table 3. Each mutant was compared to the respective controls. �: indicates a worse germination percentage of the mutant in the columncompared to the respective genotype at the top.¼ : no difference in germination percentage could be observed,1 : higher germinationpercentage could be observed. If dormancy was present germination percentage differences were estimated based on day 5 and 7 of treatment.Statistical analysis was performed using LD50 values, *: indicates a statistical significant difference, # indicates no statistical analysis wasperformed using this data. (d) indicates dormancy was present in this line, so no statistical comparison was possible. A correction to the P-valuefor multiple tests was made CDT1 P, 0.008 and CDT2 P, 0.012. LD50 values for Ler are CDT1 3.62 SE 0.15, CDT2 2.49 SE 0.07 and Col inCDT2 6.08 SE 0.41.

LD50 SE err Col Ler Ws vtc1-1 tt4-1 cad2-1 ats cat1 cat3

CDT 1vtc1-1 ¼ (d)#cad2-1 ¼ (d)#tt4-1 3.58 0.23 ¼ats 3.05 0.43 ¼cat1 cat3 1 (d)tt4-1 vtc1-1 (1) 2.55 0.34 � (d)# � � (d) �tt4-1 vtc1-1 (2) 2.92 0.65 � (d)# � � (d) �tt4-1 cad2-1 (1) 4.88 0.18 1 (d)# 1 1 1 (d)#tt4-1 cad2-1 (2) 5.21 0.37 1 (d)# 1 1 1 (d)#ats vtc1-1 (1) 4.19 0.47 1 (d)# 1 1 (d) 1ats vtc1-1 (2) 4.72 0.13 1 (d)# 1 1 (d) 1ats cad2-1 (1) 6.27 0.56 1 (d)# 1 1 (d)# 1 *ats cad2-1 (2) 3.72 0.37 1 (d)# ¼ 1 (d)# 1ats cat1 cat3 (1) 4.88 0.62 1 1 (d) 1 1 (d)#ats cat1 cat3 (2) 2.33 0.32 � � (d) � � (d)#tt4-1ats 2.61 0.49 � � �CDT 2vtc1-1 3.95 0.28 �cad2-1 6.21 0.56 ¼tt4-1 1.92 0.20 �cat1 cat3 � (d)#cad2-1cat1 cat3 (1) 4.15 0.35 � ¼ (d)# � ¼ (d)#cad2-1cat1 cat3 (2) 4.53 0.60 � ¼ (d)# � ¼ (d)#vtc1-1 cat1 cat3 (1) 2.15 0.09 � * � (d)# � * � (d)#vtc1-1 cat1 cat3 (2) 2.70 0.20 � * � (d)# � � (d)#vtc1-1 cad2-1 2.77 0.38 � * � � *tt4-1 cat1 cat3 (1) 3.94 0.41 1 ¼ (d)# 1 * ¼ (d)#tt4-1 cat1 cat3 (2) 2.60 0.42 ¼ � (d)# 1 � (d)#

Fig. 4. Germination percentage percentages (� SE) of Ler, F1 Lerats, F1 ats Ler and ats after 0 (black bars), 1 (dark grey bars),2 (light grey bars) and 4 (white bars) days of controlleddeterioration treatment. Common letters indicate that valuescould not be separated statistically (P, 0.05).


Statistical analysis using the probit transformed germi-nation data after 0.3% H2O2 treatment; are compiled inTable 4, with analyses after 0.5% H2O2 giving similarresults. Comparing the vtc1-1, cad2-1 and cat1 cat3mutants and the respective ‘double’ mutants showed thatnone of them is significantly different from its controls.Total germination percentage of the tt4-1 mutant seedswas significantly less than of Ler seeds after storage at32% RH and imbibition on 0.3% H2O2. A decrease ingermination percentage was also observed after 60% RHstorage, although not significant. The overall tendency for‘double’ mutants, which combine tt4-1 with the ROSscavenging deficiency mutants, suggested an additiveeffect, although only one of the tt4-1 cat1 cat3 lines wassignificantly different from its respective controls after32%RH storage.

Discussion

In the present study we analysed seeds of different geno-types. For comparison of the genotype effect, the seedswere produced simultaneously under the same environ-mental conditions and stored together under the same

conditions. Deterioration of seed was achieved byprolonged storage in ambient conditions, by storage athigher RH and by the application of a CDT, developedfor Arabidopsis by Tesnier et al. (2002), because it allowsa rapid deterioration of genotypes that under naturalstorage conditions remain viable for at least 4 years.Poor longevity of ABA related mutants, previouslyreported by Tesnier et al. (2002) after CDT, was in agree-ment to the poor longevity we observed after storage atambient conditions. The ats mutant seeds that showed areduced longevity upon storage at ambient conditions(Fig. 1), did not show a significantly reduced longevityafter the CDT1 treatment (Table 3), However, a CDTwas able to able to discriminate in the maternal effect ofthe ats mutation on longevity through modification of thetesta structure (Fig. 4). But again relative loss of longevitydue to the ats mutation is stronger during ambient storageconditions compared to a CDT. This indicates that aCDT can have predictive value for seed longevity, butdoes not completely mimics deterioration during storageat ambient conditions. The reason for this might be thatthe kinetics of seed ageing and the importance ofdegradation processes involved, vary in relation to the

Fig. 5. Germination percentages (� SE) of controls, single mutants and ‘double’ mutants after 11 months storage at ambient conditions andstorage for 1 year at 32% RH (A) and 60% RH (B) and subsequent germination percentage on 0% H2O2 (black bars), 0.3% H2O2 (grey bars)and 0.5% H2O2 (white bars).


temperature and the moisture content of the seeds (Walters1998). The mutations may affect processes that are moresensitive to one ageing condition compared to another.Comparing the longevity of Ler seeds in CDT1 and

CDT2, shows a lower LD50 value in CDT1, the sameholds for the Col seeds in CDT1 and CDT2. This dis-crepancy might be due to unnoticed differences duringthe CD treatments such as slight differences in RH dur-ing equilibration of the seeds or the amount or RH of theair trapped with the seeds in the sealed foil bags duringthe 40�C treatment. This underscores the importance ofcomparing simultaneously treated seed samples. Com-parisons within CDT1, CDT2 and the ambient storageconditions showed that Ler seeds showed in all threetests less longevity than Col seeds.The mutants were selected for this evaluation because

of their role in seed germination, their effect on hormonebiosynthesis or hormone action or because of their rolein the protection against oxidative stress. The resultsindicated that both after 4 years of ambient storage andafter CDT considerable differences in germination per-centages exist in comparison to their corresponding wildtypes. For extreme abi3 alleles, like abi3-5 (Ooms et al.1993, Bies-Etheve et al. 1999) poor longevity has beendescribed before. However, poor longevity has not beenreported previously for the leaky abi3 alleles, like abi3-1and abi3-7 (Bies-Etheve et al. 1999), which were moresensitive than wild type to a CDT (Tesnier et al. 2002,Clerkx et al. 2003). Other mutants that showed reducedlongevity are ats, aba1-5 and possibly rdo2 and C3-7-1,

which have, in common with the abi3 alleles, a reductionin seed dormancy. During dormancy metabolism is low,hence a low production of detrimental products mightimprove longevity. Bentsink et al. (2000) and Clerkx et al.(in press) showed, by QTL mapping, that the moredormant Arabidopsis accession Cape Verde Islands(Cvi) and Shakdara had a greater survival after a CDT.Another indication of a relationship between dor-

mancy and seed longevity is the induction of secondarydormancy, which in nature is a response to unfavourablegermination conditions (Bewley and Black 1994).In CDT1 in Col, vtc1-1, cad2-1, Ws and cat1 cat3 andin CDT2 in Ws and cat1 cat3, secondary dormancy isinduced. The dormancy could be induced in two wayseither through the high temperature at which the seedsare exposed during the treatment (40�C) or by re-dryingthe seeds. The first is known as thermodormancy, induc-tion of dormancy at a temperature above the optimalgermination temperature (Bewley and Black 1994).Induction of dormancy during air-dry storage was pre-viously reported for prechilled Sitka spruce seeds (Joneset al. 1998). Induction of dormancy due to a high tem-perature treatment is in our experiments not the likelycause, because control seeds (0 days treatment) are onlyequilibrated at 85% RH at 15�C and not exposed to40�C, while these seeds do show induction of secondarydormancy. The second more likely cause is humidifica-tion of the seeds at relative high humidity (85% RH),followed by drying the seeds at 32% RH, resulting in aseed moisture content (MC) of approximately 6%

Table 4. Comparison of germination percentage percentages and statistical analysis of all mutants and double mutants after storage at 32%and 60% RH and subsequent germination percentage at 0.3% H2O2. Statistical analysis using the 0.5% H2O2 gave similar results. u: indicatesthat no differences can be detected due to no germination percentage at 0.3% H2O2. At the top of the table all genotypes the comparison wasmade to: – indicates a lower germination percentage, 1 indicates a higher germination percentage germination and ¼no difference ingermination percentage could be observed. * indicates a significant difference. Statistical analysis was performed with probit transformed data,P, 0.005 (corrected for multiple tests) was considered significant.

Col Ler Ws vtc1-1 tt4–1 cad2-1 ats cat1 cat3

storage condition (%RH) 32 60 32 60 32 60 32 60 32 60 32 60 32 60 32 60

vtc1-1 ¼ 1cad2-1 � 1tt4-1 � * �ats � * � *cat1 cat3 ¼ 1cad2-1cat1 cat3 (1) � 1 � � ¼ ¼ � �cad2-1cat1 cat3 (2) � ¼ � � ¼ � � �vtc1-1 cat1 cat3 (1) ¼ 1 � 1 ¼ 1 � ¼vtc1-1 cat1 cat3 (2) � 1 � ¼ � 1 � �vtc1-1 cad2-1 � 1 ¼ 1 ¼ 1tt4-1 vtc1-1 (1) � � � � � � � �tt4-1 vtc1-1 (2) � * � � � � * � * � � *tt4-1 cad2-1 (1) � � � � � � � �tt4-1 cad2-1 (2) � � � � � � � �tt4-1 cat1 cat3 (1) � * � � * � * � � � * � *tt4-1 cat1 cat3 (2) � * � * � * � * � * � � * � *ats vtc1-1 (1) � * � * � * � * � * � * u uats vtc1-1 (2) � * � * � * � * � * � * u uats cad2-1 (1) � * � * � * � * � * � * u uats cad2-1 (2) � * � * � * � * � * � * u uats cat1 cat3 (1) � * � * � * � * u u � * � *ats cat1 cat3 (2) � * � * � * � * u u � * � *tt4-1ats � * � * � * � * u u


(Tesnier et al. 2002). All seeds are then stored at 4�C, and6% MC, until sowing of all seeds on the same day. Theinduction of secondary dormancy due to drying isreported for switchgrass seeds that had been stratified,immediately after stratification the germinability washigh but when followed by slow drying the germinationof the seeds decreased again (Shen et al. 2001).During the hypersensitive response after pathogen

infection and ozone induced cell death, involving ROS,the ethylene insensitive etr1 mutant (Bleecker et al. 1988,Chang et al. 1993), does not enhance the production ofROS, whereas the jar1 mutant, which is not able tocontain the spreading of cell death, enhances the effectsof these treatments (Overmyer et al. 2003). If thesehormones play a role in ROS accumulation during longterm seed storage one could expect that etr1 show anenhanced seed longevity whereas jar1 would show theopposite effect. We were not able to confirm this hypoth-esis because neither etr1 nor jar1 differed significantlyfrom the wild type Col seeds (Fig. 1). The effect of themonogenic mutants in which antioxidant levels arereduced is limited. It was expected that such mutantswould be affected by long term storage and CDT condi-tions because several authors have suggested that ROSdamage could be a major cause of seed damage duringstorage (Hendry 1993, McDonald 1999) especially in oilcontaining seeds where lipid peroxidation can generateROS. However, it has also been suggested that the lipidsthemselves may undergo lipid peroxidation and in thisway protect membranes from damage (Gidrol et al.1989). Damage related to ROS, and the way this isaffected by environmental and genetic factors, can beseen as the sum of the amount of damage generatedand the degree by which ROS is curtailed. An importantaspect of the latter is the effectiveness of the enzymaticand nonenzymatic antioxidant defence systems but alsothe possibility to slow down metabolism in general. CDTis known to generate free radicals (Bailly et al. 1996,Khan et al. 1996) but also has been shown to affect thelevel of antioxidants (De Vos et al. 1994, De Paula et al.1996, Bailly et al. 1997). Oxidative stress can reduce seedquality, which is suggested by the decline in germinationpercentage of the fro1 mutant (Fig. 2), known to accu-mulate ROS constitutively (Lee et al. 2002). A problemwith the interpretation of the single mutant phenotypes,especially when related to the antioxidant effects is thefunctional redundancy of the different antioxidant sys-tems. This relates to the multitude of both enzymaticantioxidant systems, which are only functional whensufficient water is available (Hoekstra et al. 2001), andnonenzymatic antioxidant systems, but also to the factthat most of the mutants used are ‘leaky’, which meansthat a residual amount of antioxidants is present. Thecat1 cat3 double mutant still has a functional CATA-LASE 2 gene, the ascorbic acid deficient mutant (vtc1-1) still has 30% of the wild type levels (Conklin et al.1996) and the glutathione deficient mutant (cad2-1)still contains 15–30% of glutathione (Howden et al.1995).

For a better understanding of the importance of theantioxidant systems we generated ‘double’ mutants,assuming that a synergistic effect might be observed insome ‘double’ mutants when the total level of importantantioxidants would drop below a certain threshold.Although some of the ‘double’ mutants might show areduced survival in both the CDT and the hydrogen per-oxide germination, they are not extremely sensitive. Threepossible explanations for this limited effect can be thatfirstly ROS accumulation and the level of antioxidants isof minor importance in generating damage during seedstorage, when no excessive ROS is generated, which isprobably the case in the fro1 mutant. Secondly, themutant combinations studied still have antioxidant levelsabove the critical threshold, especially in seeds where theseantioxidants have not been analysed. Measuring ascor-bate and glutathione levels, more particularly their redoxstate, could provide insight. Thirdly, other protectionmechanisms might be equally, or more, important. Thesemay be other antioxidant systems such as a-tocopherol,the level of SOD (reviewed by Grene 2002) and AtPer1,which is a seed specific peroxiredoxin maintained in thedry seed (Haslekas et al. 1998). Indeed, seeds deficient invitamin E (a-tocopherol) do show a reduced longevity(Dellapenna, personal communication).Important factors might also be structural, influencing

the effectiveness by which membranes and other macro-molecules are protected (Wolkers et al. 1998). A role forsugars has been suggested and a low ratio of sucrose tooligosaccharides was found to correlate with long termlongevity of seeds (Obendorf 1997). Other mechanismsimportant in dry seeds are the accumulation of amphi-philic molecules such as late embryogenic abundant(LEA) proteins, fructans and the successful formationof a biological glass protecting macromolecules andstructural components (reviewed by Hoekstra et al.2001, Oliver et al. 2001). The glassy state of desiccationtolerant tissue is depending on temperature and hydra-tion state. Leprince and Hoekstra (1998) showed thatduring seed dehydration, metabolism is downregulateddue to an increasing viscosity of the cytoplasm, whichcould prevent ROS production.The reduction in germination percentage of the single

and ‘double’ testa mutants in the H2O2 germination isstronger than the effect of the antioxidant mutants.The ats mutant is characterized by a reduced number ofintegument cell-layers (Leon-Kloosterziel et al. 1994) andtt4-1 mutant seeds are characterized by the absence oftannins in their seed coat and both mutations enhancethe permeability of the testa (Debeaujon et al. 2000).These mutant seeds might therefore be much morepermeable for H2O2 compared to mutants with normalseed coats because all ‘double’ mutants with an atsmutant testa fail to germinate on the concentrationsH2O2 employed, while all tt4-1 seem to show a reductionin germination percentage. The decreased viability of atsmutant seeds is also observed after a 4-year storage(Fig. 1), indicating the importance of a ‘functional’ seedcoat for seed longevity, even during natural ageing.


However, the effect of the ats mutation in the CDT isrelatively small, whereas natural ageing has a larger effect.Besides seed production conditions, the genetic

background in which a mutation is present, plays animportant role. This is indicated by the observation thatwhen the same two mutations are combined in variouscombinations of mixed genetic backgrounds, they show adifferent reaction to the same treatment. These differencesbetween two ‘double’ mutant lines, carrying the samemutations are observed both in the CDT and in the germ-ination on H2O2. Allelic differences between accessionscould be observed when quantitative traits loci (QTL) forCD test survival were mapped in Ler/Cvi (Bentsink et al.2000) and in Ler/Sha (Clerkx et al. 2004) recombinantinbred line populations. Allelic differences in response tooxidative stress, induced by paraquat, were also observedby Abarca et al. (2001) who found that the Cvi accessionwas more tolerant to this treatment compared to Ler andCol. It was suggested that this difference in tolerance couldbe explained at least partially by a different allele of aCu/Zn SOD. These observations make studying theeffects of defects in antioxidant scavengers in the samegenetic background necessary, especially when the effectof these mutants is relatively small as found in the pre-sent study.

Acknowledgements – We would like to thank Steve Tonsor for hishelp and advice concerning the statistical analysis of the data andcolleagues at the laboratories of Plant Physiology and Genetics andthe STW supervision committee for useful suggestions and discus-sions. This research is supported by the Technology FoundationSTW, applied science division of NWO and the technology programof the Ministry of Economic Affairs.

References

Abarca D, Roldan M, Martın M, Sabater B (2001) Arabidopisthaliana ecotype Cvi shows an increased tolerance to photo-oxidative stress and contains a new chloroplastic copper/zincsuperoxide dismutase isoenzyme. J Exp Bot 52: 1417–1425

Alvarado AD, Bradford KJ (1988) Priming and storage of tomato(Lycopersicon lycopersicum) seeds. I. Effects of storage tem-perature on germination rate and viability. Seed Sci Technol 16:601–612

Bailly C, Benamar A, Corbineau F, Come D (1996) Changes inmalondialdehyde content and in superoxide dismutase, catalaseand glutathione reductase activities in sunflower seeds as relatedto accelerated ageing. Physiol Plant 97: 104–110

Bailly C, Benamar A, Corbineau F, Come D (1997) Changes insuperoxide dismutase, Catalase and glutathione reductaseactivities in sunflower seeds during accelerated ageing andsubsequent priming. In: Basic and Applied Aspects of SeedBiology (RH Ellis, M Black, AJ Murdoch, TD Hong, eds),pp 665–672. Kluwer academic Publishers, Dordrecht

Bentsink L, Alonso-Blanco C, Vreugdenhil D, Tesnier K,Groot SPC, Koornneef M (2000) Genetic analysis of seed-soluble oligosacchrides in relation to seed storability ofArabidopsis. Plant Physiol 124: 1595–1604

Bernatzky R, Tanksley SD (1986) Genetics of actin-relatedsequences in tomato. Theor Appl Genet 72: 314–324

Bewley JD, Black M (1994) Seeds: Physiology of Development andGermination. Plenum Press, New York 0-306-44747-9

Bies-Etheve N, da Silva Conceicao A, Giraudat J, Koornneef M,Leon-Kloosterziel KM, Valon C, Delseny M (1999) Importanceof the B2 domain of the Arabidopsis ABI3 protein and 2S generegulation. Plant Mol Biol 40: 1045–1054

Bleecker AB, Estelle MA, Somerville C, Kende H (1988) Insensitiv-ity to ethylene conferred by a dominant mutation in Arabidopsisthaliana. Science 241: 1086–1089

Blokhina O, Virolainen E, Fagerstedt KV (2003) Antioxidants,oxidative damage and oxigen deprivation stress: a review. AnnBot 91: 179–194

Buitink J, Hemminga MA, Hoekstra FA (2000) Is there a role foroligosaccharides in seed longevity ? an assesment of intracellularglass stability. Plant Physiol 122: 1217–1224

Cakmak I, Strbac D, Marschner H (1993) Activities of hydrogenperoxide-scavenging enzymes in germinating wheat seeds. J ExpBot 44: 127–132

Chang C, Kwok SF, Bleecker AB, Meyerowitz EM (1993)Arabidopsis ethylene-response gene ETR1: similarity of productto two-component regulators. Science 262: 539–544

Clerkx EJM, Blankenstijn de Vries H, Ruys GJ, Groot SPC,Koornneef M (2003) Characterization of green seed, an enhan-cer of abi3–1 in Arabidopsis that affects seed longevity. PlantPhysiol 132: 1077–1084

Clerkx EJM, El-Lithy ME, Vierling E, Ruys GJ, Blankestijn-DeVries H, Groot SPCD, Vreugdenhil D, Koornneef M (2004)Analysis of natural allelic variation of Arabidopsis seed qualitytraits between the accessions Landsberg erecta and Shakdara,using a new recombinant inbred line population. Plant Physiol135: 432–442

Cobbett CS, May MJ, Howden R, Rolls B (1998) The glutathione-deficient, cadmium-sensitive mutant cad2–1 of Arabidopsisthaliana is deficient in g-glutamylcysteine synthetase. Plant J16: 73–78

Conklin PL, Norris SR, Wheeler GL, Williams EH, Smirnoff N,Last RL (1999) Genetic evidence for the role of GDP-mannosein plant ascorbic acid (vitamin C) biosynthesis. Proc Natl AcadSci USA 96: 4198–4203

Conklin PL, Williams EH, Last RL (1996) Environmental stresssensitivity of an ascorbic acid deficient Arabidopsis mutant.Proc Natl Acad Sci USA 93: 9970–9974

Coolbear P (1995) Mechanisms of seed deterioration. In: SeedQuality: Basic Mechanisms and Agricultural Implications.(A S Basra, ed), pp 223–277. Food Product Press

Corbineau F, Come D (1993) The concept of dormancy in cerealseeds. Proceedings of the 4th International Workshop on Seeds,Basic and Applied Aspects of Seed Biology, Angers, France,July 20–24, pp. 581–589

De Paula M, Perez-Otaola M, Darder M, Torres M, Frutos G,Martinez-Honduvilla CJ (1996) Function of the ascorbate-glutathione cycle in aged sunflower seeds. Physiol Plant 96:543–550

De Vos CHR, Kraak HL, Bino RJ (1994) Ageing of tomato seedsinvolves glutathione oxidation. Physiol Plant 92: 131–139

Debeaujon I, Leon-Kloosterziel KM, Koornneef M (2000)Influence of the testa on seed dormancy, germination and long-evity in arabidopsis. Plant Physiol 122: 403–413

Dickson MH (1980) Genetic aspects of seed quality. Hort Sci 15:771–774

Ellis RH, Hong TD, Roberts EH, Tao K-L (1990) Low moisturecontent limits to relations between seed longevity and moisture.Ann Bot 65: 493–504

Frugoli JA, Hong Zhong H, Nuccio ML, McCourt P, McPeek MA,Thomas TL, McClung CR (1996) Catalase is encoded by amultigene family in Arabidopsis thaliana (L.) Heynh. PlantPhysiol 112: 327–336

Gidrol X, Serghini H, Noubhani B, Mocouot B, Mazliak P (1989)Biochemical changes induced by accelerated ageing in sunflowerseeds. I. Lipid peroxidation and membrane damage. PhysiolPlant 79: 591–597

Giraudat J, Hauge BM, Valon C, Smalle J, Parcy F, Goodman HM(1992) Isolation of the Arabidopsis ABI3 gene by positionalcloning. Plant Cell 4: 1251–1261

Grace SC, Logan BA (2000) Energy dissipation and radicalscanvenging by plant phenylpropanoid pathway. Phil Trans RSoc Lond 355: 1499–1510

Greggains V, Finch-Savage WE, Quick WP, Atherton NM(2000) Putative desiccation tolerance mechanisms in ortho-dox and recalcitrant seeds of the genus Acer. Seed Sci Res10: 317–327


Grene R (2002) Oxidative stress and acclimation mechanismin plants. In: The Arabidopsis Book (C R Somerville,E M Meyerowitz, eds) The American Society of Plant Biolo-gists, Rockville

Haslekas C, Stacy RAP, Nygaard V, Culianez MFA, Aalen RB(1998) The expresion of a peroxiredoxin antioxidant gene,AtPer1. Arabidopsis thaliana is seed specific and related todormancy. Plant Mol Biol 36: 833–845

Hay FR, Mead A, Manger K, Wilson FJ (2003) One-step analysisof seed storage data and the longevity of Arabidopsis thalianaseeds. J Exp Bot 54: 993–1011

Hay FR, Probert RJ, Smith RD (1997) The effect of maturity onthe moisture realtions of seed longevity in foxglove (Digitalispurpurea L). Seed Sci Res 7: 341–349

Hendry GAF (1993) Oxygen, free radical processes and seedlongevity. Seed Sci Res 3: 141–153

Hoekstra FA, Golovina EA, Buitink J (2001) Mechanisms of plantdesiccation tolerance. Trends Plant Sci 6: 431–438

Howden R, Andersen CR, Goldsbrough PB, Cobbet CS (1995)A cadmium sensitive, glutathione-deficient mutant of Arabidopsisthaliana. Plant Physiol 107: 1067–1073

Jalink H, van der Schoor R, Frandas A, van Pijlen JG, Bino RJ(1998) Chlorophyll fluorescence of Brassica oleracea seeds as anon-destructive marker for seed maturity and seed performance.Seed Sci Res 8: 437–443

Jones SK, Gosling PG, Ellis RH (1998) Reimposition of conditionaldormancy during air-dry storage of prechilled Sitka spruceseeds. Seed Sci Res 8: 113–122

Khan MM, Hendry GAF, Atherton NM, Vertucci-Walters CW(1996) Free radical accumulation and lipid peroxidation intestas of rapidly aged soybean seeds: a light regulated process.Seed Sci Res 6: 101–107

Kliebenstein D, Monde R, Last RL (1998) Superoxide dismutase inArabidopsis: an eclectic enzyme family with disparate regulationand protein localization. Plant Physiol 118: 637–650

Koornneef M, Elgersma A, Hanhart CJ, Van LoenenMartinet EP, van Rijn EP, Zeevaart JAD (1985) A gibberellininsensitive mutant of Arabidopsis thaliana. Physiol Plant 65:33–39

Koornneef M, Jorna ML, Brinkhorst-van der Swan DLC, KarssenCM (1982) The isolation of abscisic acid (ABA) deficientmutants by selection of induced revertants in non-germinatinggibberellin sensitive lines of Arabidopsis thaliana (L.) Heynh.Theor Appl Genet 61: 385–393

Koornneef M, Reuling G, and. Karssen CM (1984) The isolationand characterization of abscisic acid- insensitive mutants ofArabidopsis thaliana. Physiol Plant 61: 377–383

Koornneef M, Van der Veen JH (1980) Induction and analysis ofgibberellin-sensitive mutants in Arabidopsis thaliana (L) Heynh.Theor Appl General 58: 257–263

Larkindale J, Knight MR (2002) Protection against heat stress-induced oxidative damage in Arabidopsis involves calcium,abscisic acid, ethylene and salycylic acid. Plant Physiol 128:682–695

Lee B-H, Lee H, Xiong L, Zhu J-K (2002) A mitochondrial com-plex I defect impairs cold-regulated nuclear gene expression.Plant Cell 14: 1235–1251

Leon-Kloosterziel KM, Alvarez-Gil M, Ruijs GJ, Jacobsen SE,Olszewski NE, Schwartz SH, Zeevaart JAD, Koornneef M(1996b) Isolation and characterization of abscisic acid-deficientArabidopsis mutants at two new loci. Plant J 10: 655–661

Leon-Kloosterziel KM, Keijzer CJ, Koornneef M (1994) A seedshape mutant of Arabidopsis that is affected in integumentdevelopment. Plant Cell 6: 385–392

Leon-Kloosterziel KM, van de Bunt GA, Zeevaart JAD, KoornneefM(1996a) Arabidopsis mutants with a reduced seed dormancy. PlantPhysiol 110: 233–240

Leprince O, Atherton NM, Deltour R, Hendry GAF (1994) Theinvolvement of respiration in free radical processes during lossof dessication tolerance in germinating Zea mays L. PlantPhysiol 104: 1333–1339

Leprince O, Hoekstra FA (1998) The response of cytochrome redoxstate and energy metabolism to dehydration support a role forcytoplasmic viscosity in dessication tolerance. Plant Physiol 118:1253–1264

Lindstrom EW (1942) Inheritance of seed longevity in maizehybrids. Genetics 27: 154

Lyall TW, Ellis RH, John P, Hedley CL, Wang TL (2003) Mutantalleles of rugosus loci in pea affect seed moisture sorption iso-terms and the relation between seed longevity and moisturecontent. J Exp Bot 54: 445–450

McDonald MB (1999) Seed deterioration: Physiology, repair andassessment. Seed Sci Technol 27: 177–237

McGee DC (2000) Pathology of seed deterioration. In: GeneticImprovement of Seed Quality, SH Moore, RW Yaklich, edsMadison Wisconsin: Crop Science Society of America, Ana-heim, pp 21–37

Neff MM, Neff JD, Chory J, Pepper AE (1998) dCAPS, a simpletechnique for the genetic analysis of single nucleotide poly-morphisms: experimental applications in Arabidopsis thalianagenetics. Plant J 14: 387–392

Noctoer G, Foyer C (1998) Ascorbate and glutathione: keepingactive oxygen under control. Ann Rev Plant Physiol Plant MolBiol 49: 249–279

Obendorf RL (1997) Oligosaccharides and galactosyl cyclitols inseed desiccation tolerance. Seed Sci Res 7: 63–74

Oliver AE, Leprince O, Wolkers WF, Hincha DK, Heyer AG,Crowe JH (2001) Non-disaccharide-based mechanisms of pro-tection during drying. Cryobiology 43: 151–167

Ooms JJJ, Leon-Kloosterziel KM, Bartels D, Koornneef M,Karssen CM (1993) Acquisition of desiccation tolerance andlongevity in seeds of Arabidopsis thaliana – A comparativestudy using ABA insensitive abi3 mutants. Plant Physiol 102:1185–1191

Overmyer K, Brosche M, Kangasjarvi J (2003) Reactive oxygenspiecies and hormonal control of cell death. Trends Plant Sci8: 335–342

Peeters AJM, Blankestijn-de Vries H, Hanhart CJ, Leon-Kloosterziel KM, Zeevaart JAD, Koornneef M (2002) Charac-terization of mutants with reduced seed dormancy at two novelrdo loci and a further characterization of rdo1 and rdo2.Arabidopsis Physiol Plant 115: 604–612

Peng J, Carol P, Richards DE, King KE, Cowling RJ, Murphy GP,Harberd NP (1997) The Arabidopsis GAI gene defines a signal-ing pathway that negatively regulates gibberellin responses.Genes Dev 11: 3194–3205

Puntarulo S, Galleano M, Sanchez RA, Boveris A (1991) Super-oxide anion and hydrogen peroxide metabolism in soybeanembryonic axes during germination. Biochim Biophys Acta IntJ Biochem Biophys 1074: 277–283

Rice-Evans CA, Miller NJ, Paganga G (1997) Antioxidant propertiesof phenolic compounds. Trends Plant Sci 2: 152–159

Rock CD, Zeevaart JAD (1991) The aba mutant of Arabidopsisthaliana is impaired in epoxy-carotenoid biosynthesis. Proc NatlAcad Sci USA 88: 7496–7499

Salome PA, McClung CR (2002) Charectirization of the Arabidopsiscatalase family. XIII International Conference on ArabidopsisResearch: Seville: Spain

Senaratna T, Gusse JF, McKersie BD (1988) Age-induced changesin cellular membranes of imbibed soybean seed axes. PhysiolPlant 73: 85–91

Shen Z-X, Parrish DJ, Wolf DD, Welbaum GE (2001) Stratificationof switchgrass seeds is reversed and hastened by drying. CropSci 41: 1546–1551

Shirley BW, Kubasek WL, Storz G, Bruggemann E, Koornneef M,Ausubel FM, Goodman HM (1995) Analysis of Arabi-dopsis mutants deficient in flavonoid biosynthesis. Plant J 8:659–671

Staswick PE, Su W, Howell SH (1992) Methyl jasmonate inhibitionof root growth and induction of leaf protein are decreased on anArabidopsis thaliana mutant. Proc Natl Acad Sci USA 89:6837–6840

Staswick PE, Tiryaki I, Rowe ML (2002) Jasmonate response locusJAR1 and several related Arabidopsis genes encode enzymes ofthe firefly luciferase superfamily that show activity on jasmonic,salicylic, and indole-3acetic acids in an assay for adenylation.Plant Cell 14: 1405–1415

Sun T, Goodman HM, Ausubel FM (1992) Cloning the Arabi-dopsis GA1 locus by genomic subtraction. Plant Cell 4:119–128


Tesnier K, Strookman-Donkers HM, van Pijlen JG, van der GeestALHM, Bino RJ, Groot SPC (2002) A controlled deteriorationtest fo Arabidopsis thaliana reveals genetic variation in seedquality. Seed Sci Technol 30: 149–165

Tommasi F, Paciolla C, Arrigoni O (1999) The ascorbate system inrecalcitrant and orthodox seeds. Physiol Plant 105: 193–198

Tommasi F, Paciolla C, Concetta de Pinto M, Da Gara L (2001)A comparative study of glutathione and ascorbate metabolism dur-ing germination of Pinus pinea L seeds. J Exp Bot 52: 1647–1654

Walters C (1998) Understanding the mechanisms and kinetics ofseed ageing. Seed Sci Res 8: 223–244

Willekens H, Inze D, van Montagu M, van Camp W (1995)Catalases in plants. Mol Breed 1: 207–228

Wilson DO Jr (1995) The storage of orthodox seeds. In: SeedQuality: Basic Mechanisms and Agricultural Implications (A SBasra, ed), pp 173–208 Food Product Press

Wolkers WF, Alberda M, Koornneef M, Leon-Kloosterziel KM,Hoekstra FA (1998) Properties of proteins and the glassy matrixin maturation defective mutant seeds of Arabidopsis thaliana.Plant J 16: 133–143

Xiong L, Schumaker KS, Zhu J-K (2002) Cell signaling during cold,drought, and salt stress. Plant Cell 14:

Edited by D. Campell


genetic differences in seed longevity of various arabidopsis mutants

Documents