Spaceflight-induced genetic and epigeneticchanges in the rice (Oryza sativa L.) genome areindependent of each other

Xiufang Ou, Likun Long, Ying Wu, Yingjie Yu, Xiuyun Lin, Xin Qi, and Bao Liu

Abstract: An array of studies have reported that the spaceflight environment is mutagenic and may induce phenotypic andgenetic changes in diverse organisms. We reported recently that in at least some plant species (e.g., rice) the spaceflightenvironment can be particularly potent in generating heritable epigenetic changes in the form of altered cytosine methyla-tion patterns and activation of transposable elements. To further study the issue of spaceflight-induced genomic instability,and in particular to test whether the incurred genetic and epigenetic changes are connected or independent of each other,we performed the present study. We subjected seeds of the standard laboratory rice (Oryza sativa L.) cultivar Nipponbareto a spaceflight in the spaceship Long March 2 for 18 days. We then investigated the genetic and DNA methylation stabil-ities of 11 randomly selected plants germinated from the spaceflown seeds by using two kinds of DNA markers, amplifiedfragment length polymorphism (AFLP) and methylation sensitive amplified polymorphism (MSAP). For AFLP, by using15 primer combinations, we assessed 460 genomic loci and found that the frequencies of genetic changes across the 11plants ranged from 0.7% to 6.7% with an average frequency of 3.5%. For MSAP, by using 14 primer combinations, we as-sessed 467 loci and detected the occurrence of four major types of cytosine methylation alterations at the CCGG sites,namely CG or CNG hypomethylation and CG or CNG hypermethylation. Collectively, the frequencies of the two kinds ofhypermethylation, CG (1.95%) and CNG (1.44%), are about two times higher than those of the two kinds of hypomethyla-tion, CG (0.76%) and CNG (0.80%), though different plants showed variable frequencies for each type of alteration. Fur-ther analysis suggested that both the genetic and cytosine methylation changes manifested apparent mutational bias towardsspecific genomic regions, but the two kinds of instabilities are independent of each other based on correlation analysis.

Key words: spaceflight, genetic and epigenetic changes, mutational bias, rice.

Resume : Plusieurs etudes ont montre que l’environnement des vols spatiaux est mutagene et peut induire des change-ments epigenetiques et genetiques chez diverses especes. Les auteurs ont rapporte recemment que, chez certaines especesvegetales (p. ex., le riz), cet environnement s’averait particulierement propice a la production de changements epigeneti-ques heritables sous la forme de modifications a la methylation des cytosines et de l’activation d’elements transposables.Afin d’etudier plus en profondeur la question de l’instabilite genomique induite par le vol spatial et pour tester specifique-ment si ces changements genetiques et epigenetiques sont lies ou independants, les auteurs ont realise la presente expe-rience. Des graines du riz standard utilise en recherche (Oryza sativa L. ‘Nipponbare’) ont effectue un sejour dansl’espace de 18 jours a bord du vaisseau spatial « Long March 2 ». Les auteurs ont ensuite etudie les stabilites genetiqueset de methylation de l’ADN chez 11 plantes issues des graines presentes a bord de ce vol a l’aide de deux types de mar-queurs moleculaires, les AFLP (« amplified fragment length polymorphism ») et les MSAP (« methylation sensitive ampli-fied polymorphism »). Pour l’analyse AFLP, 15 paires d’amorces ont permis d’examiner 460 locus genomiques ; lafrequence de changements genetiques chez ces 11 plantes variait entre 0,7 % et 6,7 %, pour une moyenne de 3,5 %. Pourl’analyse MSAP, a l’aide de 14 paires d’amorces, 467 locus ont ete etudies. Les quatre principaux types d’alterations a lamethylation ont ete detectes, soit l’hypomethylation des sites CG ou CNG et l’hypermethylation des sites CG ou CNG.Les frequences des deux types d’hypermethylation, 1,95 % des sites CG et 1,44 % des sites CNG, etaient environ deuxplus elevees que les frequences des deux types d’hypomethylation, soit 0,76 % des sites CG et 0,80 % des sites CNG,bien que ces frequences variaient chez differentes plantes. Une analyse plus approfondie a montre que les changements ge-netiques et de methylation des cytosines affichaient apparemment un biais favorisant certaines regions genomiques, bienque les deux types d’instabilites etaient independants l’un de l’autre sur la base d’une analyse de correlation.

Mots-cles : vol spatial, changements genetiques et epigenetiques, biais mutationnel, riz.

[Traduit par la Redaction]

Received 25 November 2009. Accepted 17 April 2010. Published on the NRC Research Press Web site at genome.nrc.ca on 17 June2010.

X. Ou,1 L. Long,1 Y. Wu, Y. Yu, X. Lin, and B. Liu.2 Key Laboratory of Molecular Epigenetics of MOE and Institute of Genetics andCytology, Northeast Normal University, Changchun 130024, China.X. Qi. Department of Agronomy, Jilin Agricultural University, Changchun 130118, China.

1These authors contributed equally to this work.2Corresponding author (e-mail: baoliu@nenu.edu.cn).

524

Genome 53: 524–532 (2010) doi:10.1139/G10-030 Published by NRC Research Press

IntroductionRecent years have witnessed much interest in investigat-

ing the possible utility and advantage of using spaceflight-induced mutations in plant genetic improvement. Severalstudies have reported the effects of spaceflight on plantgrowth or development retardation and the induction of var-ious phenotypic and genetic mutations at frequencies ex-ceeding those of ground mutagenesis (Krikorian andO’Connor 1984; Kuang et al. 1996; Li et al. 2007).Although abolishment or alteration of functional proteins asa result of spaceflight-induced mutations in some criticalcellular genes represents the most parsimonious explanationfor the phenotypic mutations (Nechitailo et al. 2005; Li etal. 2007), it becomes increasingly clear that DNA sequenceis not the only determinant of phenotypes. Rather, epigeneticchanges, which are heritable alterations in chromatin struc-ture including cytosine DNA methylation and (or) histonemodifications, may also cause transgenerational alteration ingene expression and produce novel phenotypes that do notentail a change in the primary nucleotide sequence (Yoderet al. 1997; Boyko and Kovalchuk 2008). Given that epige-netic modifications are particularly prone to environmentalperturbations, it is logical to assume that the nature of thespaceflight environment, which encompasses multiple fac-tors including cosmic irradiation, microgravity, and spacemagnetic fields, may induce concomitant epigenetic and ge-netic changes.

Indeed, we have found recently that spaceflight is highlymutagenic from an epigenetic perspective and induces highfrequencies of transgenerational alteration in DNA methyla-tion in rice (Oryza sativa L.) plants germinated from space-flown seeds (Ou et al. 2009). Moreover, transpositionalactivation of several transposable elements (TEs) that werequiescent in the ground control plants also occurred in someof the spaceflown rice plants (Long et al. 2009). However, alikely issue regarding the genetic and epigenetic mutationfrequencies estimated in these studies is that only prese-lected TEs or genic sequences were investigated, and hencean unbiased assessment of the extent to which the space-flight environment is mutagenic has not been completed.Here, we have extended our earlier studies by assessing thegenetic variation and alteration in cytosine methylation inspaceflown rice plants at a large number of randomlysampled and hence unbiased loci from a genome-wide per-spective. Specifically, we used amplified fragment lengthpolymorphism (AFLP) and methylation sensitive amplifiedpolymorphism (MSAP) markers to assess genetic and cyto-sine methylation alteration frequencies, respectively, inspaceflown rice plants relative to ground control samples.We aimed to determine whether a common mechanism orindependent mechanisms underlie genetic and epigenetic in-stabilities in response to spaceflight.

Materials and methods

Plant material and spaceflightWell-developed and healthy mature seeds of the standard

laboratory rice (Oryza sativa L.) cultivar Nipponbare wereused. The seeds were randomly divided into two groups and

packed in cotton-cloth envelopes. One group was flown onthe spaceship Long March 2 for 18 days (from 28 Septem-ber to 15 October 2004) and the other group was kept undernormal ambient conditions and later used as the ground con-trols. Both the spaceflown and ground control seeds weregerminated immediately after the flight under standard con-ditions in double-distilled water, and equal numbers (11) ofrandomly selected plants (designated as C0 and M0, respec-tively, for the ground control and spaceflown plants) weretagged and later used for this study.

AFLP and MSAP analysisGenomic DNA was isolated from expanded leaves of the

individual C0 and M0 plants using a modified cetyltrimethy-lammonium bromide (CTAB) method (Kidwell and Osborn1992). The experimental conditions for the standard AFLPemploying enzymes EcoRI and MseI were exactly as de-scribed by Wang et al. (2005). One preselective and 15 se-lective primer combinations were used (Table S1).3

MSAP is a modified version of AFLP that incorporatesmethylation-sensitive restriction enzymes in an efficient pro-cedure to reveal genome-wide DNA methylation alterationsin a locus-specific manner (Cervera et al. 2002; Dong et al.2006). HpaII and MspI are isoschizomers that recognize thesame restriction site (5’-CCGG-3’) but have differential sen-sitivity to methylation states of the two cytosines: HpaII willnot cut if either of the cytosines is fully (double-strand) me-thylated, whereas MspI will not cut if the external cytosineis fully or hemi- (single-strand) methylated. Thus, for agiven DNA sample, the full methylation of the internal cyto-sine (CG methylation) or hemi-methylation of the externalcytosine (CNG methylation) at the assayed CCGG sites canbe unequivocally resolved (Cervera et al. 2002; Dong et al.2006). One pair of preselective and 14 pairs of selective pri-mers were used (Table S2).3 The restriction enzymes EcoRI,MseI, HpaII, and MspI were purchased from New EnglandBiolabs Inc. (Beverly, Massachusetts, USA). The amplifica-tion products of AFLP and MSAP were resolved by electro-phoresis in a 5% denaturing polyacrylamide gel andvisualized by silver staining (Wang et al. 2005). Only clearand completely reproducible bands in two experiments using

Fig. 1. Tabulated frequencies of genetic changes detected by AFLPthat occurred in some or all of the 11 randomly selected spaceflownrice (cultivar Nipponbare) plants (M0). Frequencies for two types ofpatterns (loss of parental bands and gain of novel bands) as well asfor the total variant bands are indicated.

3 Supplementary data for this article are available on the journal Web site (http://genome.nrc.ca).

Ou et al. 525

Published by NRC Research Press

Table 1. Number and type (L, loss; N, novel) of variant AFLP bands detected in each of the spaceflown (M0) plants by each of 15 primer combinations.

Eleven randomly selected M0 plants germinated from spaceflown seeds of the standard rice cultivar Nipponbare

1 2 3 4 5 6 7 8 9 10 11

Primer pair (total no. of bandsscored by each primer pair) L N L N L N L N L N L N L N L N L N L N L Na1 (26) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0a2 (38) 1 3 1 3 1 3 1 3 1 0 1 1 1 1 0 3 0 3 1 0 0 1b1 (46) 2 1 2 1 1 1 1 1 0 0 0 0 0 0 1 1 2 1 0 1 0 1b7 (46) 2 2 2 2 2 2 2 1 0 1 0 1 0 1 2 1 2 1 0 1 0 1c4 (36) 1 0 1 0 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0c7 (39) 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0e1 (31) 1 0 1 0 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0e2 (26) 2 1 2 1 2 1 2 1 0 0 1 0 1 0 1 1 2 1 0 0 1 0e9 (28) 1 1 1 1 1 1 1 1 0 0 0 0 0 0 1 1 1 1 0 0 0 0f1 (19) 0 1 0 2 0 2 0 2 0 0 0 0 0 0 0 2 0 2 0 0 0 1f6 (24) 0 1 0 1 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0g1 (26) 3 0 3 0 3 0 3 0 0 0 0 0 2 0 2 0 2 0 0 0 0 0g7 (28) 2 1 2 1 1 1 1 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0h5 (26) 1 3 1 3 1 1 1 3 0 0 0 1 0 0 1 1 1 1 0 0 0 0h7 (21) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0Sum of variant bands 16 14 16 15 14 13 14 13 1 2 2 3 4 2 9 10 13 10 1 2 1 4Genetic alteration freq. (%)* 3.5 3.0 3.5 3.3 3.0 2.8 3.0 2.8 0.2 0.4 0.4 0.7 0.9 0.4 2.0 2.1 2.8 2.2 0.2 0.4 0.2 0.9Total genetic alteration freq. (%)* 6.5 6.7 5.9 5.9 0.7 1.1 1.3 4.1 5.0 0.7 1.1

*Divided by the total number of bands (460) that were scored by the 15 primer pairs used.

526G

enome

Vol.

53,2010

Publishedby

NR

CR

esearchPress

independent DNA extractions (technical replications) werescored.

Recovery and sequencing of AFLP and MSAP bandsBands showing clearly variant patterns of genetic or cyto-

sine methylation alteration in the spaceflown rice plants rel-ative to the ground controls were eluted from the silver-stained AFLP or MSAP gels and re-amplified with the ap-propriate selective primer combinations. Sizes of the PCR

products were verified by agarose gel electrophoresis, andthen the products were cloned into the pMD18-T vector (Ta-kara Biotechnology Inc., Dalian, China). The cloned DNAsegments were sequenced with vector primers by automatedsequencing. The Advanced BlastN and BlastX programs atthe NCBI Web site (http://www.ncbi.nlm.nih.gov/) wereused for homology analysis of the cloned DNA sequencesthat gave quality reads (Table S3).3

Results

Global assessment of genetic mutations in the spaceflownrice plants by AFLP

We performed AFLP analysis to get a general assessmentfrom a genome-wide perspective of the genetic stability ofthe spaceflown rice plants. By using 15 pairs of EcoRI andMseI primers (each with 3 selective nucleotides at their 3’end) (Table S1), we scored 460 clear and reproducible bandsbetween two technical replications. Thirteen primer pairsgenerated polymorphic patterns among the 11 spaceflownrice plants or between at least one spaceflown plant and theground controls, whereas 2 primer combinations producedmonomorphic banding patterns only. Taking all 15 primercombinations together, an average mutation frequency of3.5% was calculated for the 11 randomly selected space-flown rice plants. The mutation frequencies are variableamong the 11 randomly chosen spaceflown plants, rangingfrom 0.7% (#5 and #10) to 6.7% (#2) (Fig. 1 and Table 1).The variant bands can be divided into two types, i.e., loss oforiginal bands that existed in the ground controls but disap-peared from some of the spaceflown plants, and gain ofnovel bands that did not exist in the ground controls but ap-peared de novo in the spaceflown plants (e.g., Fig. 2A).When all 11 plants were considered together, the two typesof variant bands (loss and gain) occurred at more or lessequal frequencies; however, when plants were consideredon an individual basis, one type of variant band could bepredominant over the other (Fig. 1 and Table 1).

Given that the standard laboratory rice cultivar Nippon-bare is a genetically pure line and was maintained in ourhands by constant strict-selfing, the possibility of preexistinggenetic polymorphism among the ground control plants isunlikely. However, to further validate this, we randomlychose an equal number of plants germinated from theground control group of seeds (see Materials and methods)and performed a parallel AFLP analysis by using the same15 primer pairs that revealed genetic changes in the space-flown plants (above), and we detected only monomorphicpatterns across the 11 ground control plants (e.g., Fig. 2B),thus validating the genetic pure-line nature of the controlplants.

Global assessment of alterations in cytosine methylationin the spaceflown rice plants by MSAP

We performed MSAP analysis to get a general assessmentfrom a genome-wide perspective of the epigenetic stabilityof the spaceflown rice plants. With 14 selected EcoRI +HpaII/MspI primer combinations (Table S2), 467 clear andreproducible bands between two technical replications wereamplified from the same 11 randomly chosen spaceflownplant samples as used for AFLP analysis. Relative to the

Fig. 2. (A) Example of AFLP profiles showing the two types ofgenomic changes in some or all of the 11 randomly selected space-flown rice plants (M0). Arrows indicate loss of bands comparedwith control plants and arrowheads indicate gain of bands in thespaceflown plants relative to the ground control samples (one isshown). The primer combination is EcoRI + AAG / MseI + CTG.(B) Example of AFLP profiles in 11 randomly selected groundcontrol plants (C0). The primer combination is EcoRI + ACC /MseI + CAC.

Fig. 3. Tabulated frequencies of the four major types of alterationsin cytosine methylation at the 5’-CCGG sites, namely CG hypo-methylation, CNG hypomethylation, CG hypermethylation, andCNG hypermethylation, which occurred in some or all of the 11randomly selected spaceflown rice plants (M0).

Ou et al. 527

Published by NRC Research Press

ground control samples, the MSAP profiles of spaceflownplants revealed the occurrence of four major types of cyto-sine methylation alterations at the CCGG sites. These areCG hypomethylation, CNG hypomethylation, CG hyperme-thylation, and CNG hypermethylation (e.g., Fig. 3 and Ta-ble 2). Collectively, the frequencies of the two kinds ofhypermethylation, CG (1.95%) and CNG (1.44%), arehigher than those of the two kinds of hypomethylation, CG(0.76%) and CNG (0.80%) (Fig. 3 and Table 2). Similar tothe case of genetic variations detected by AFLP, the fre-quencies for each type of methylation alteration are also var-iable among the 11 spaceflown plants (Fig. 4A).Specifically, 4 plants (#1, #3, #5, and #11) showed moreCNG than CG hypomethylation, while 5 plants (#2, #4, #7,#9, and #10) showed more CG than CNG hypomethylation(Fig. 3 and Table 2).

As in the AFLP analysis, a parallel MSAP analysis of thesame 11 ground control plants by all 14 primer pairs foundno evidence for preexisting cytosine methylation polymor-phisms among ground control individuals of the rice cultivarNipponbare (e.g., Fig. 4B). This is consistent with our ear-lier analysis by methylation-sensitive Southern blotting (Ouet al. 2009).

Both genetic and epigenetic changes suggest ‘‘mutationalhotspots’’ in the spaceflown rice plants

Spaceflight is a process including various factors such ascosmic irradiation, microgravity, and space magnetic fields(Mashinsky and Nechitailo 2001). Given that relative tomost ground mutagens (e.g., ionizing acute irradiation), theduration of a spaceflight journey is long (18 days in this ex-periment) and the environment within the space shuttle ishighly uniform, it is probable that the spaceflight-inducedgenetic and (or) epigenetic mutational spectrum might revealsome targeting preference or ‘‘hotspots’’ in the genome dueto the persistent nature of the treatment. Indeed, a recentstudy by Li et al. (2007) showed that spaceflight-inducedmutations preferentially occurred at intrinsically polymor-

phic sites in the rice genome, thus suggesting some kind ofmutational hotspots. To test whether genetic or methylationmutational bias also existed in the genome of the space-flown rice plants used in this study, the number of changedloci that occurred in at least 6 of the 11 spaceflown plants inthe AFLP or MSAP profiles was tabulated (Table 3). Thedata showed that 19 (52.8%) of the 36 changed AFLP lociand 20 (48.8%) of the 41 changed MSAP loci occurred inmore than half of the 11 randomly chosen spaceflownplants, hence pointing to the existence of both genetic andepigenetic ‘‘mutational bias sites’’ in the rice genome withregard to their mutational lability in the spaceflight environ-ment.

Recovery and sequencing of AFLP and MSAP variantbands

To gain some information on the nature of sequencesunderlying the genetic and epigenetic changes, we isolated,cloned, and sequenced a subset of variant bands (15 fromAFLP that showed obvious mutational bias; 32 fromMSAP, of which 20 showed obvious mutational bias andthe rest showed alteration in 1 to 5 of the 11 spaceflownplants). A BlastX search at the NCBI Web site (http://www.ncbi.nlm.nih.gov/) indicated that diverse sequences are in-volved in the genetic and methylation changes induced byspaceflight. Specifically, one variant AFLP band (AF3) andtwo variant MSAP bands (MS1 and MS9) showed signifi-cant homology to genes with known or predicted functions.AF3 is a putative RAFTIN1 protein, while MS1 and MS9are, respectively, a protein kinase domain and a putativefatty acid elongase. Two variant bands (one each fromAFLP and MSAP) are related to putative transposon or ret-rotransposon proteins. Nineteen variant bands (7 AFLP and12 MSAP) showed homology to hypothetical proteins withunspecified functions. The remaining 6 and 17 variantbands, respectively, from AFLP and MSAP showed no ho-mology to the current database sequences at GenBank(Table 4 and Table S3).

Table 2. Number and type (CG hypo-, CNG hypo-, CG hyper-, and CNG hypermethylation) of variant bands de-tected by MSAP employing a pair of isoschizomers (HpaII and MspI) and 14 MSAP primer combinations in eachof 11 randomly selected spaceflown (M0) plants.

M0 plantCG hypomethylation(%), HpaII (– ? +)

CNG hypomethylation(%), MspI (– ? +)

CG hypermethylation(%), HpaII (+ ? –)

CNG hypermethylation(%), MspI (+ ? –)

1 4 (0.86) 8 (1.71) 12 (2.57) 4 (0.86)2 3 (0.64) 1 (0.21) 12 (2.57) 7 (1.50)3 2 (0.43) 6 (1.28) 13 (2.78) 11 (2.36)4 5 (1.07) 2 (0.43) 9 (1.93) 3 (0.64)5 3 (0.64) 5 (1.07) 3 (0.64) 4 (0.86)6 3 (0.64) 3 (0.64) 3 (0.64) 2 (0.43)7 4 (0.86) 1 (0.21) 6 (1.28) 9 (1.93)8 3 (0.64) 3 (0.64) 12 (2.57) 9 (1.93)9 3 (0.64) 1 (0.21) 9 (1.93) 9 (1.93)10 2 (0.43) 1 (0.21) 15 (3.21) 11 (2.36)11 6 (1.28) 10 (2.14) 6 (1.28) 5 (1.07)Average (%) 3.45 (0.76) 3.73(0.80) 9.09 (1.95) 6.64 (1.44)

Note: We did not score variant MSAP bands that occurred concomitantly in both enzyme digestions, as they might have had agenetic basis due to nucleotide sequence changes in the adaptor and (or) selective nucleotide sequences. A total of 467 bandswere amplified from the 11 randomly chosen spaceflown plants.

528 Genome Vol. 53, 2010

Published by NRC Research Press

No correlation existed between the spaceflight-inducedgenetic and epigenetic changes

The simultaneous occurrence of both genetic and DNAmethylation changes raised the question as to whether thetwo kinds of genomic instabilities are interlaced or occurredindependently. To test this, we used Pearson’s test of SPSS11.5 for Windows (SPSS Inc., Chicago) to calculate the cor-relation coefficients for paired matrices generated by each ofthe markers. The results showed no evidence for the exis-tence of a significant general correlation between the geneticchanges detected by AFLP and the cytosine methylation al-terations detected by MSAP (Table 5). To further dissect thepossible relationships, we calculated the correlations be-tween all possible pairs of each of the two variable AFLPpatterns (gain and loss of bands) and each of the variableMSAP patterns representing the four major types of methyl-ation alterations, namely CG or CNG hypermethylation and

CG or CNG hypomethylation (Fig. 3). Again, we found thatnone of the correlation coefficients between the variable pat-terns of AFLP and those of MSAP were statistically signifi-cant (n = 11, P < 0.05) (Table 5). Although a lepto-correlation (P < 0.1) was found to exist between the totalsequence change detected by AFLP and CG hypermethyla-tion (Table 5), statistical insignificance of all the other pair-wise correlations strongly suggests independency of the twokinds of genomic instabilities.

DiscussionAccumulated recent studies have reported that the space-

flight environment is mutagenic and capable of inducingvarious phenotypic and genetic changes (Krikorian andO’Connor 1984; Kuang et al. 1996; Li et al. 2007). How-ever, opposite results pointing to complete genetic stabilityin spaceflown plants have also been reported (Sychev et al.2007). One possible cause for the discrepancy is differentialsensitivity of different plant species to the spaceflight envi-ronment, a scenario consistent with what is well knownabout differences among species or even individuals in sus-ceptibility to wide-ranging mutagens and stresses.

Our previous studies have shown that spaceflight is capa-ble of inducing a high incidence of alteration in DNA meth-ylation and associated transpositional activation of TEs thatare quiescent in ground controls (Long et al. 2009; Ou et al.2009), thus pointing to the genetic and epigenetic mutagenicnature of the spaceflight environment. This possibility wasconfirmed in this study by employing two types of molecu-lar markers enabling the assessment of genetic changes andcytosine DNA methylation alterations, respectively, at multi-ple and randomly sampled loci across the rice genome in 11randomly selected spaceflown plants of the standard ricecultivar Nipponbare. We found that relative to the groundcontrols, both kinds of genomic instabilities, i.e., geneticchanges and alterations in cytosine methylation, not only oc-curred in the spaceflown plants but occurred at high fre-quencies. We noted that similar to mutations induced byground mutagens, stochasticity apparently played a role (butsee below) in both kinds of genomic changes induced by thespaceflight, as evidenced by the wide range of mutation fre-quencies across the 11 randomly chosen plants that appa-rently had experienced the same conditions. Similarfindings have been reported in other studies of spaceflight-induced genetic mutations (Nechitailo et al. 2005; Li et al.2007).

In a previous study, by using methylation-sensitive South-ern blotting with a set of preselected TEs and genic sequen-ces as probes, we found that the predominant type ofcytosine methylation alteration induced by spaceflight is hy-permethylation of CNG sites, with CG hypermethylation andboth CG and CNG hypomethylation occurring only occa-sionally (Ou et al. 2009). In this study, by analyzing a largernumber of randomly sampled loci across the rice genomebased on MSAP fingerprinting, we found that methylationalterations also occurred at CG sites in the spaceflown riceplants, thus indicating locus-specific (or sequence-specific)epigenetic responses to the spaceflight. Nonetheless, theconclusion that spaceflight-induced methylation alterationsin rice are mainly of the hypermethylation type still holds,

Fig. 4. (A) Example of MSAP profiles showing the various types ofalterations in cytosine methylation at the 5’-CCGG sites in 11 ran-domly selected spaceflown rice plants (M0). Arrows indicate loss ofbands in the spaceflown plants compared with the ground controlplants, thus representing hypermethylation of CG (in EcoRI +HpaII digest) or CNG (in EcoRI + MspI digest); arrowheads indi-cate gain of bands in the spaceflown plants compared with the con-trol plants, thus denoting hypomethylation of CG (in EcoRI +HpaII digest) or CNG (in EcoRI + MspI digest). The primer com-bination is EcoRI + ACT / HpaII (MspI) + TCT. (B) Example ofMSAP profiles in 11 randomly selected ground control plants (C0).The primer combination is EcoRI + AAG / MseI + TCG.

Ou et al. 529

Published by NRC Research Press

with the average frequencies of CG and CNG hypermethyla-tions being 1.95% and 1.44%, respectively, compared with0.76% and 0.80%, respectively, for CG and CNG hypome-thylations. This is consistent with the proposal that cytosinemethylation has evolved primarily as a genome defense de-vice (Yoder et al. 1997; Rapp and Wendel 2005), and imme-diate genome-wide hypermethylation together with locus-specific hypomethylation subsequent to a traumatic stressrepresents a protective and adaptive response (Kovalchuk etal. 2003; Boyko et al. 2007; Boyko and Kovalchuk 2008;Ou et al. 2009).

Although a widely held basic tenet of mutational theory isthe random nature of mutation with regard to both rate andspectrum, numerous recent molecular studies have chal-lenged this idea (Metzgar and Wills 2000; Rando and Ver-strepen 2007, and references therein). Several studies haveshown that mutational ‘‘hotspots’’ may exist in the genomesof higher eukaryotes, and the term refers to a locus or clus-ters of loci in a given genome that are more prone to muta-tion by a specific physical or chemical mutagen (Rando andVerstrepen 2007). These studies have shown that both therate and the spectrum of mutations are largely genetically

Table 3. Analysis of genetic (based on AFLP) and epigenetic (based on MSAP) mu-tations shared by 2 to all 11 randomly selected spaceflown (M0) plants.

No. of M0 plants thatshared a given variant band

No. (%) of variant bandsdetected by AFLP

No. (%) of variant bandsdetected by MSAP

1 2 (5.6) 4 (9.8)2 2 (5.6) 6 (14.6)3 4 (11.1) 2 (4.9)4 4 (11.1) 6 (14.6)5 5 (13.9) 3 (7.3)6 13 (36.1) 8 (19.5)7 4 (11.1) 3 (7.3)8 1 (2.8) 6 (14.6)9 0 (0) 3 (7.3)10 0 (0) 0 (0)11 1 (2.8) 0 (0)Total 36 (100) 41 (100)Mutational bias loci* 19 (52.8) 20 (48.8)

*Refers to mutations (variant bands) that occurred in at least 6 of the 11 randomly chosen M0plants.

Table 4. Functional classification of a subset of sequenced variant bandsfrom AFLP or MSAP profiles that represent genetic or epigenetic variationsin one or more of the 11 randomly selected plants germinated from space-flown seeds of rice cultivar Nipponbare.

CategoryNo. (%) of variantbands from AFLP

No. (%) of variantbands from MSAP

Known-function protein 1 (6.7) 2 (6.3)Putative protein-coding gene 7 (46.7) 12 (37.5)Transposon or retrotransposon 1 (6.7) 1 (3.1)No similarity 6 (40.0) 17 (53.1)Total 15 (100) 32 (100)

Note: Detailed information on these variant bands is given in Table S3.

Table 5. Correlation coefficients (r values) between genetic variations de-tected by AFLP and cytosine methylation alterations detected by MSAP,using Pearson’s test.

Total sequence change Loss GainTotal methylation alteration 0.390 0.353 0.427CG hypermethylation 0.573 0.555 0.586CNG hypermethylation 0.077 0.089 0.063CG hypomethylation –0.041 –0.072 –0.006CNG hypomethylation –0.022 –0.074 –0.038

Note: Except for the existence of a lepto-correlation between ‘‘Total sequencechange’’ and ‘‘CG hypermethylation’’ (r = 0.573, P < 0.1), none of the remaining cor-relation coefficients are statistically significant (P > 0.05).

530 Genome Vol. 53, 2010

Published by NRC Research Press

controlled and modulated, suggesting that at least some ofthe mutations represent an adaptive response of the genomeand hence should be nonrandom (Metzgar and Wills 2000;Rando and Verstrepen 2007). In this study, among the ob-served mutation loci detected by AFLP and MSAP in thespaceflown rice plants, 52.8% and 48.8% of the genetic andmethylation changes occurred in at least 6 of the 11 ran-domly chosen spaceflown rice plants, thus indicating thatthe same set of loci is extremely labile to genetic or epige-netic mutation and hence suggesting the existence of geneticand epigenetic ‘‘mutational hotspots’’ in the rice genome inresponse to the spaceflight environment. Apparently, thismutational bias occurred concomitantly with the generaltrend of stochasticity, discussed above.

Based on sequence homology, the loci that underwent thegenetic and DNA methylation changes are diverse, includingprotein-coding genes, transposon- and retrotransposon-relatedsequences, and sequences with as yet unknown functions, im-plying that these loci in the rice genome are more sensitive tospaceflight-induced mutagenesis. A previous study showedthat spaceflight-induced mutations preferentially occurred atpolymorphic sites existing in different rice cultivars (Li et al.2007). This is in agreement with our results in the sense thatthere exist certain groups of loci in the rice genome that areparticularly susceptible to the spaceflight environment, hencepointing to the existence of mutational hotspots.

Previous studies have shown that epigenetic variations,particularly those in the form of cytosine methylation, canbe dependent on, interlaced with, or independent of geneticvariations (Richards 2006). In this study, the statistical insig-nificance of all possible pairwise correlation coefficients be-tween genetic variations detected by AFLP and alterations incytosine methylation detected by MSAP strongly suggeststhat the spaceflight-induced genetic and epigenetic varia-tions are largely independent of each other. This accordswell with the relationship between naturally occurring ge-netic changes and methylation alterations in rice cultivars(Ashikawa 2001; Wang et al. 2004), thus suggesting thatthere might be no fundamental difference in the nature andmechanism of genetic or epigenetic mutations that occur inspace and those that occur on the ground. However, giventhat the mutation rate and spectrum were estimated basedon randomly chosen individuals and randomly sampled ge-nomic loci, the present study has provided unequivocal evi-dence that the spaceflight environment is highly mutagenicin rice and, by extension, probably also in other plants.

To conclude, it is clear from this study that the space-flight environment is indeed highly mutagenic to plantsboth genetically (generating nucleotide base changes) andepigenetically (causing alterations in cytosine methylation),with the two kinds of mutations being largely independentof each other. Although it is likely that both the mutationalmechanism and the nature of mutations induced by space-flight are not fundamentally different from those of muta-tions generated on the ground, spaceflight-basedmutagenesis likely represents a potentially efficient ap-proach for plant mutagenesis and breeding selection.

AcknowledgementsThis study was supported by the Program for Changjiang

Scholars and Innovative Research Team (PCSIRT) in Uni-

versity (No. IRT0519). We are particularly grateful to ananonymous reviewer who painstakingly read our manuscriptand all related publications and who gave numerous con-structive suggestions to improve the manuscript.

ReferencesAshikawa, I. 2001. Surveying CpG methylation at 5’-CCGG in the

genomes of rice cultivars. Plant Mol. Biol. 45(1): 31–39. doi:10.1023/A:1006457321781. PMID:11247604.

Boyko, A., and Kovalchuk, I. 2008. Epigenetic control of plantstress response. Environ. Mol. Mutagen. 49(1): 61–72. doi:10.1002/em.20347. PMID:17948278.

Boyko, A., Kathiria, P., Zemp, F.J., Yao, Y., Pogribny, I., and Ko-valchuk, I. 2007. Transgenerational changes in the genome sta-bility and methylation in pathogen-infected plants: (virus-induced plant genome instability). Nucleic Acids Res. 35(5):1714–1725. doi:10.1093/nar/gkm029. PMID:17311811.

Cervera, M.T., Ruiz-Garcıa, L., and Martınez-Zapater, J.M. 2002.Analysis of DNA methylation in Arabidopsis thaliana based onmethylation-sensitive AFLP markers. Mol. Genet. Genomics,268: 543–552. doi:10.1007/s00438-002-0772-4.

Dong, Z.Y., Wang, Y.M., Zhang, Z.J., Shen, Y., Lin, X.Y., Ou,X.F., et al. 2006. Extent and pattern of DNA methylation altera-tion in rice lines derived from introgressive hybridization of riceand Zizania latifolia Griseb. Theor. Appl. Genet. 113(2): 196–205. doi:10.1007/s00122-006-0286-2. PMID:16791687.

Kidwell, K.K., and Osborn, T.C. 1992. Simple plant DNA isolationprocedures. In Plant genomes: methods for genetic and physicalmapping. Edited by J.S. Beckman and T.C. Osborn. KluwerAcademic Publishers, Dordrecht, Netherlands. pp. 1–13.

Kovalchuk, O., Burke, P., Arkhipov, A., Kuchma, N., James, S.J.,Kovalchuk, I., and Pogribny, I. 2003. Genome hypermethylationin Pinus silvestris of Chernobyl — a mechanism for radiationadaptation? Mutat. Res. 529(1–2): 13–20. doi:10.1016/S0027-5107(03)00103-9. PMID:12943916.

Krikorian, A.D., and O’Connor, S.A. 1984. Karyological observa-tions. Ann. Bot. (Lond.), 54: 49–63.

Kuang, A., Musgrave, M.E., and Matthews, S.W. 1996. Modifica-tion of reproductive development in Arabidopsis thaliana underspaceflight conditions. Planta, 198(4): 588–594. doi:10.1007/BF00262646. PMID:11539321.

Li, Y., Liu, M., Cheng, Z., and Sun, Y. 2007. Space environmentinduced mutations prefer to occur at polymorphic sites of ricegenomes. Adv. Space Res. 40(4): 523–527. doi:10.1016/j.asr.2007.04.100.

Long, L., Ou, X., Liu, J., Lin, X., Sheng, L., and Liu, B. 2009. Thespaceflight environment can induce transpositional activation ofmultiple endogenous transposable elements in a genotype-depen-dent manner in rice. J. Plant Physiol. 166(18): 2035–2045.doi:10.1016/j.jplph.2009.06.007. PMID:19628300.

Mashinsky, A.L., and Nechitailo, G.S. 2001. Results and prospectsof studying the gravitationally sensitive systems of plants underconditions of space flight. Cosmic Res. 39(4): 317–327. doi:10.1023/A:1017993528116.

Metzgar, D., and Wills, C. 2000. Evidence for the adaptive evolu-tion of mutation rates. Cell, 101(6): 581–584. doi:10.1016/S0092-8674(00)80869-7. PMID:10892644.

Nechitailo, G.S., Lu, J.Y., Xue, H., Pan, Y., Tang, C.Q., and Liu,M. 2005. Influence of long term exposure to spaceflight on to-mato seeds. Adv. Space Res. 36(7): 1329–1333. doi:10.1016/j.asr.2005.06.043.

Ou, X.F., Long, L.K., Zhang, Y.H., Xue, Y.Q., Liu, J.C., Lin, X.Y.,and Liu, B. 2009. Spaceflight induces both transient and herita-

Ou et al. 531

Published by NRC Research Press

ble alterations in DNA methylation and gene expression in rice(Oryza sativa L.). Mutat. Res. 662(1–2): 44–53. doi:10.1016/j.mrfmmm.2008.12.004. PMID:19135069.

Rando, O.J., and Verstrepen, K.J. 2007. Timescales of genetic andepigenetic inheritance. Cell, 128(4): 655–668. doi:10.1016/j.cell.2007.01.023. PMID:17320504.

Rapp, R.A., and Wendel, J.F. 2005. Epigenetics and plant evolu-tion. New Phytol. 168(1): 81–91. doi:10.1111/j.1469-8137.2005.01491.x. PMID:16159323.

Richards, E.J. 2006. Inherited epigenetic variation — revisiting softinheritance. Nat. Rev. Genet. 7(5): 395–401. doi:10.1038/nrg1834. PMID:16534512.

Sychev, V.N., Levinskikh, M.A., Gostimsky, S.A., Bingham, G.E.,and Podolsky, I.G. 2007. Spaceflight effects on consecutive gen-erations of peas grown onboard the Russian segment of the In-ternational Space Station. Acta Astronaut. 60(4–7): 426–432.doi:10.1016/j.actaastro.2006.09.009.

Wang, Y.M., Lin, X.Y., Dong, B., Wang, Y., and Liu, B. 2004.DNA methylation polymorphism in a set of elite rice cultivarsand its possible contribution to inter-cultivar differential geneexpression. Cell. Mol. Biol. Lett. 9(3): 543–556. PMID:15332130.

Wang, Y.M., Dong, Z.Y., Zhang, Z.J., Lin, X.Y., Shen, Y., Zhou,D., and Liu, B. 2005. Extensive de novo genomic variation inrice induced by introgression from wild rice (Zizania latifoliaGriseb.). Genetics, 170(4): 1945–1956. doi:10.1534/genetics.105.040964. PMID:15937131.

Yoder, J.A., Walsh, C.P., and Bestor, T.H. 1997. Cytosine methyla-tion and the ecology of intragenomic parasites. Trends Genet.13(8): 335–340. doi:10.1016/S0168-9525(97)01181-5. PMID:9260521.

532 Genome Vol. 53, 2010

Published by NRC Research Press