Proc. Natl. Acad. Sci. USAVol. 90, pp. 9310-9314, October 1993Developmental Biology

Maize mitochondrial manganese superoxide dismutases are encodedby a differentially expressed multigene family

(cDNA clone/gene expression/Zea mays)

DAHAI ZHU AND JOHN G. SCANDALIOS*Department of Genetics, Box 7614, North Carolina State University, Raleigh, NC 27695

Communicated by Clement L. Markert, July 19, 1993

ABSTRACT We have isolated maize cDNAs encodingthree manganese-containing superoxide dismutases (MnSODs)distinct from the one previously reported. Molecular analysesindicate that multiple MnSOD transcripts are encoded bydifferent, though similar, genes in the maize genome. A singleMnSOD gene has been reported in all other organisms exam-ined to date. The deduced amino acid sequences show that thesemaize MnSOD proteins have a mitochondrial transit peptideand that the first 9 amino acids (matrix-targeting sequence) inthe transit peptide are conserved. This suggests that all themaize MnSOD proteins are mitochondria-associated isozymes.RNA blot analysis demonstrated that each member ofthe maizeMnSOD multigene family is both spatially and developmentallyregulated. One gene, Sod3.3, was predominantly expressed inthe embryo late in embryogenesis. Patterns of increased Mn-SOD transcript accumulation are shown to be associated withincreased mitochondrial activity during plant growth anddevelopment. The influence of mitochondrial metabolism onthe expression of the nuclear MnSOD genes is discussed.

Active oxygen species, the ubiquitous products of single-electron reductions of dioxygen, include the superoxideradical (0O), hydrogen peroxide (H202), and the hydroxylradical (OH-). Their production in aerobic organisms canresult in the peroxidation ofmembrane lipids (1), DNA strandbreakage (2), and inactivation of enzymes (3). Aerobic orga-nisms have evolved both enzymatic and nonenzymatic an-tioxidant defense systems to minimize the deleterious effectsofoxidative stress. Superoxide dismutase (SOD) and catalaseare two such important antioxidant enzymes. SOD (super-oxide oxidoreductase; EC 1.15.1.1) catalyzes the dismutationof two superoxide radicals into oxygen and hydrogen perox-ide. Catalase (H202:H202 oxidoreductase, EC 1.11.1.6) cat-alyzes the reduction of hydrogen peroxide into oxygen andwater. SOD and catalase act in concert to prevent cellulardamage by removing the superoxide and hydrogen peroxidegenerated during electron transport and other processes.

In maize, it was reported that at least six SOD isozymes arecoded for by six nonallelic nuclear genes: Sodi, Sod2, Sod3.1(previously referred to as Sod3), Sod4, Sod4A, and Sod5(4-6). The cytosolic isozymes, SOD-2, SOD-4, SOD-4A, andSOD-5, and the chloroplast-associated SOD-1 are copper-and zinc-containing homodimeric enzymes. SOD-3, the mi-tochondria-associated isozyme, is a manganese-containingtetrameric enzyme (4, 7). The maize SOD-3 protein encodedby the Sod3.1 gene is synthesized on cytosolic polyribosomesas a precursor with a cleavable presequence of 31 aa at theamino terminus (8). Import studies using deletion mutants inthe maize SOD-3 transit peptide have verified that the initialportion (about 9 aa) of the maize SOD-3 transit peptide isrequired to direct the protein into the yeast mitochondrial

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

matrix in vivo (9) and into isolated maize mitochondria invitro (10).The role of SOD in protecting cells against the toxic, often

lethal effects of oxygen free radicals has been studied invarious organisms (11). Expression of one of the maizeMnSOD genes (Sod3.1) in a MnSOD-deficient yeast mutantrendered the transformed yeast cells resistant to paraquat-induced oxidative stress (9). In maize, the MnSOD tran-scripts increase in embryos treated with the fungal toxincercosporin (12) or abscisic acid and high osmoticum (un-published data). MnSOD from Nicotiana plumbaginifolia isinduced by ethylene, salicylic acid, and Pseudomonas syrin-gae infection (13). Overexpression of the pea chloroplastCu/ZnSOD in transgenic tobacco increased resistance tooxidative stress (14). These findings confirm that SODs playa significant role in preventing cellular damage caused byoxygen free radicals.We have isolated three more cDNAs ofmaize Sod3 genes. t

The expression patterns of each individual maize MnSOD(Sod3) gene were examined with gene-specific probes. Theresults demonstrate that maize MnSOD genes constitute amultigene family and that individual members are differen-tially expressed during maize development.

MATERIALS AND METHODSConstruction and Screening of cDNA Library. Since the

plant growth regulator abscisic acid increases the abundanceof at least two maize Sod3 transcripts in developing maizeembryos (unpublished data), excised 21-day postpollination(dpp) embryos from line W64A were treated with 0.1 mMabscisic acid on Murashige-Skoog plates at 25°C in the darkfor 24 hr. Poly(A)+ RNA was isolated from the abscisicacid-treated embryos and used for cDNA synthesis with aZAP cDNA synthesis kit (Stratagene). One maize MnSODcDNA (Sod3.1) had been previously cloned (6). The cDNAlibrary was screened with a 32P-labeled full-length Sod3.1sequence. Twenty-five positive clones were obtained. Threewith different insert sizes were designated Sod3.2, Sod3.3,and Sod3.4 and were characterized further.DNA Sequencing and Amino Acid Alignment. The Sod3.2,

Sod3.3, and Sod3.4 cDNAs were subcloned and sequencedwith plasmid DNAs by the dideoxy chain-terminationmethod (15). Amino acid alignment of the predicted maizeMnSODs was established (16).

Construction of Sod Gene-Specific Probes. A 30-mer oli-godeoxynucleotide (5'-ACTAGGAGCAGAGACAGAG-TAGCACGCAAG-3') was used as the probe for the Sod3.1gene. A 20-mer synthetic oligodeoxynucleotide (5'-CTCGA-

Abbreviations: SOD, superoxide dismutase; dpi, days postimbibi-tion; dpp, days postpollination.*To whom reprint requests should be addressed.tThe sequences reported in this paper have been deposited in theGenBank database [accession nos. L19461 (Sod3.2), L19462(Sod3.3), and L19463 (Sod3.4)].

9310

Dow

nloa

ded

by g

uest

on

May

6, 2

021

Proc. Natl. Acad. Sci. USA 90 (1993) 9311

ACTCTGAAAGGGCCG-3') was used as the Sod3.2-specificprobe. The nucleotide homology between Sod3.3 and Sod3.4is extremely high; therefore, it is impossible to find DNAfragments that could be exclusively used as gene-specificprobes for each of them. We used, however, the oligodeoxy-nucleotide (5'-GTGCGTATGCGGAGTGTGTG-3') localizedin the 5' end of the untranslated regions of Sod3.3 and Sod3.4.These two Sod3 transcripts can be distinguished because theydiffer in size (0.87 kb for Sod3.3 and 1.02 kb for Sod3.4). Theposition of each Sod3 probe is underlined in Fig. 2.RNA Preparation and RNA Blot Analysis. W64A seeds were

soaked for 10 min in 1% NaOCl and then in deionized waterfor %24 hr. Imbibed seeds were germinated in the dark onmoistened germination paper (Kimpak, Chicago) at 25°C.The scutella were excised from seedlings on day 1 to day 10postimbibition (dpi) and stored at -70°C for RNA prepara-tions. Maize kernels were collected from field-grown W64Aplants at 13-37 dpp. Scutella with embryo axes were dis-sected from the kernels and stored at -70°C until needed.The following W64A tissue samples were collected for RNApreparation: root, stem, anther, mature tassel, young tassel,mature leaf, young leaf, silk, and husk. Total RNA from eachsample was isolated by cold phenol extraction (17). TotalRNA (20 ,ug per lane) was electrophoresed in denaturing1.6% agarose gels and blotted onto nitrocellulose filters.Blots were sequentially probed with probes for Sod3.1,Sod3.2, and Sod3.3/Sod3.4 which were labeled at the 5'terminus with T4 polynucleotide kinase. Hybridization timewas 40 hr at 51°C (Sod3.2 and Sod3.3/Sod3.4) or 59°C(Sod3.1). The filters hybridized with the Sod3.2 and Sod3.3/Sod3.4 probes were washed three times with 0.1 x standardsaline citrate/0.1% SDS at 47°C for 20 min. The blotshybridized with the Sod3.1 probe were washed three timeswith 0.1 x standard saline citrate/0.1% SDS at 50°C for 20min.Genomic DNA Blot Analysis. Genomic DNA was isolated

from W64A leaves (18). DNA (10 ,g) was digested withEcoRI or BamHI restriction enzyme, electrophoresed in a0.8% agarose gels, and transferred onto nitrocellulose. Afterthe filter was hybridized with the 32P-labeled full-lengthSod3.1 DNA fragment at 65°C for 20 hr, it was washed fourtimes with 0.1 x standard saline citrate/0. 1% SDS at 65°C for30 min and visualized by autoradiography.

RESULTSMaize MnSOD (Sod3) Genes Constitute a Multigene Family.

Multiple bands can be detected under high hybridization andwashing stringencies when DNA blots of maize genomicDNA are probed with 32P-labeled full-length maize Sod3.1cDNA (Fig. 1). In this report, we demonstrate the existenceof multiple distinct maize Sod3 cDNAs. Taken together,these results confirm that maize MnSODs are encoded by amultigene family. This genome organization provides thebasis for an explanation of the differential expression of themaize MnSOD genes during development.

Isolation and Characterization of Three More MnSOD(Sod3) cDNAs. A maize MnSOD cDNA was previouslycloned and has been published as Sod3 (6). Three distinctmaize MnSOD cDNAs were isolated from a newly con-structed cDNA library (see Materials and Methods). For thesake of consistency the MnSOD sequences will be referred toas Sod3.1 (previously published Sod3), Sod3.2, Sod3.3, andSod3.4. The maize MnSOD cDNAs show a high degree ofsequence homology in the coding regions and vary in thenoncoding regions (5' and/or 3' untranslated sequences) (Fig.2A). Only one translation initiation codon (ATG) is found ineach. This start site is in-frame for encoding polypeptides of232 aafor SOD-3.2 and 233 aafor both SOD-3.3 and SOD-3.4.Kozak's translation consensus sequence (GCCATGG) (19) is

E Bkb

-23. 1

-9.4

-6 .6-4.4

-2.3-2.0

FIG. 1. Maize MnSOD (Sod3) genes comprise a multigene family.Genomic DNA (10 pg) from line W64A was digested with EcoRI(lane E) or BamHI (lane B) restriction enzyme. The restrictionfragments were electrophoresed in a 0.8% agarose gel, transferredonto a nitrocellulose filter, and then hybridized with the 32P-labeledfull-length Sod3.1 cDNA. Results in this and all subsequent figuresare representative of at least three independent experiments.

found in all maize MnSOD cDNAs. The 5' untranslatedleader regions of the Sod3 cDNA clones vary in length andnucleotide sequence. Variable numbers ofAGCG motifs arefound in the 5' untranslated region of all MnSOD cDNAclones (three in Sod3.1, five in Sod3.2 and Sod3.4, four inSod3.3). Two polyadenylylation signals are found at posi-tions 789 and 1024 in Sod3.2, but only one polyadenylylationsignal, corresponding to the signal at position 789 in Sod3.2,is present in Sod3.3 (position 827) and in Sod3.4 (position818). An additional 154 nt are found at the 3' untranslated endof Sod3.2. The deduced amino acid sequences verify that theproteins encoded by the Sod3.2, Sod3.3, and Sod3.4 cDNAcontain a mitochondrial transit peptide (29 aa) with the first9 aa being identical (Fig. 2B). These 9 aa have been shown tobe necessary to target the maize pre-SOD-3 into the yeastmitochondrial matrix in vivo (9) and into isolated maizemitochondria in vitro (10). These data suggest that the SOD-3proteins encoded by these cDNA clones are all mitochondria-associated MnSODs.

Tissue-Specific Expression of the Maize So3 Genes. Expres-sion of the maize MnSOD genes in various tissues wasexamined by RNA blot analysis in which the same filter wassequentially blotted with the various Sod3 gene-specificprobes (Fig. 3). The Sod3.1 transcript was detected in thetissues examined, but there was great variation in its steady-state level (Fig. 3A). Scutella of 4-dpi seedlings had thehighest levels of Sod3.1 transcript. Husk, silk, and stemshowed higher levels of the transcript, followed by root andscutella from 27 dpp kernels. Sod3.1 transcript was barelydetected in young leaf and mature leaf (Fig. 3A). The minordifferences in Sod3.1 transcript seen in anther, young tassel,and mature tassel were due to loading variation. When thesame blot was hybridized with the Sod3.2-specific probe, theSod3.2 transcript was detected only in the scutella from 4-dpiseedlings and 27-dpp kernels (Fig. 3B). The Sod3.3 transcriptwas detected in the scutella from 27-dpp kernels (Fig. 3C),while the Sod3.4 transcript was found mainly in the scutellafrom 4-dpi seedlings, with low levels of the transcript de-tected in scutella at 27 dpp (Fig. 3C). These results suggestthat the Sod3.2, Sod3.3, and Sod3.4 genes are expressed ina tissue-specific manner in maize.

Differential Expression of the Maize Sod3 Genes in theScutellum During Seedling Development. Multiple Sod3 tran-

Developmental Biology: Zhu and Scandalios

Dow

nloa

ded

by g

uest

on

May

6, 2

021

9312 Developmental Biology: Zhu and Scandalios

A 1)(2)(3)(4)

Proc. Natl. Acad. Sci. USA 90 (1993)

CCACGCACCCAGGAGATAC ...............---CCA ..... ... ... ... ... ... ... ... ... ... ..G . T.

...C.C.AGCG ... ... ... ... ... ... ... ... ... ... ... ...

CACACACTCCGCATACGCACCTACGAGAGTCGAGTGACAGTGAGCGAGCGAGCGAGCG----AGCC ATG GCT CTC CGC ACC CTG GCA TCG MG MC GCC CTA 103...................................................AGC .. ... ... ... ... ... ... ... ... ... ... ... ...

(1). C.. T.. ..C ..C .G T C. GCG GCG ..T.. A.A

(2) ... ... .C. --- ...

(3) TCC TTC GCG CTC GGT GGA GCG GCC CGG CCG TCG GCG GAG --- --- TCC GCG AGG GGG GTG ACGACG GTC GCG CTC CCC GAC CTC TCC 190(4) C........................................................C.--- ...

(1). A .C C C C

(3) TAC GAC TTC GGC GCG CTG GAG CCG GTC ATC TCG GGG GAG ATC ATG CGC CTG CAC CAC CAG MG MC CAC GCC ACC TAC GTC GTCMC 277(4) .. CC. C G

(1) G C.T ..AAC.C TC. C.G CG ..G

(2).. G.CC C(3) TAC AAC MG GCG CTT GAG CAG ATC GAC GAT GTC GTC GTC AAG GGC GAC GAC TCC GCT GTC GTC CAG CTC CAG GGC GCC ATC MG TTC 364

(4)..G .C C C CC

(1). C G.T. C

(2). . .G... ...C... ... ...... ... G. .A. CC.(3) MC GGC GGC GGT CAT GTG MC CAT TCA ATC TTT TGG MG MC CTC MG CCT ATT AGC GAA GGT GGT GGG GAG CCG CCACATGGG 451

(4) ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...

(1) ... ... ... ... ... ... ... ... ... ... ..G ... ... ... ... ... ... .A. ... ... ... ... ... ... ... ..C ... .. G ...(2) ... ... ... ... ... ... ...

(3) CTT GGC TGG GCC ATC GAT GAG CAT GTTT CCT TG TTT GAG GCA CTT GTA MG AGG ATG MT GCA GAA GGC GCT GCT TTA rCM GGA TCT 538(4) ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...

(3) GGA TGG GTGTGG TTA GCT TTGGAT AAA GCAG GCA AMGGTTTA GTTM ACT AC AGCT AAT CAG GA CCT CTG GTG ACT AAA GGT 625

(4). C.C.

(1)

(2)(3)

(4)

... ... ... ... ... ... ... ... ... ... ... ... ... ... ..A ... ... ... ... ... ... ... ... ... ... ... ... ... ...

... ... ... ... ... ... ... ... ... ... ... ... ..G ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...GCA AGC TTG GTT CCG CTG TTG GGG ATT GAT GTC TGG GAA CAT GCG TAC TAC CTG CAG TAC MG MT GTT AGG CCG GAT TAC CTG MC 712... ... ... ... ... ... ... ... ... ... ... ... ..G ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...

(1)................................................C......A..GA....(2) ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .......................................(3) AAC ATC TGG MG GTG ATG AAC TGG AAA TAT GCT GGA GAG GTG TAC GAA AAT GTT CTT GCT TGA ATTGTCTTAATGGACATACTCA--TCTGCG 806(4) ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .......................................

(1) CG.... C----CGG-...... CTGTCTAGCGGCTGGACCTTGTGTACATTTCACTGAGATAGACTAMTGCACGGCCTGCCGATTTTTGT(2) .... TGACCMGTGTACATTTCGCTMGATAGACTAACGCACATGGCCTGCCGATTTT a88(3) --CCGGGTTTGTTTTCGGGCTGTTTGACCATGTAATAMAGAT GGACTTGTGTACCT AMAAA

(4) ......................................................... 921

(1)(2)(4)

1036

(1) AGCTGAGCTTCCGATGTTTCTT A

(2) CCAAAGAACCATTTCTACTCTTGTCTCCATTMTAAAATCTGCCGAGGTGCCGATGTTTGCTT

B SOD3. 1 MALRTLASKKVLSFPFGGAGRPLAAAASARGVTTVTLPDLSYDFGALEPAISGEIMRLHHQKHHATYVANYNKALEOLETAVSKGDASOD3.2 A...AL...A ..S.A.L..A..V.-...N.VH. V DAV.V...A

SOD3.3 .......NA...AL A. .S.E.. A. V. N. V DDV.V ...DS0D3.4 .......NA...AL... A. .S.A.. A. V. H .G DM.A ...A

SOD3.1 SAVVQLQGAAIKFNGGGHVNHSI FWKNLKPISEGGGEPPHGKLGWAIDEDFGSFEALVKKMNAEGMLQGSGWVWLALDKEAKKVSVSOD3.2 ..................F. E..P. .L..SOD3.3 ..................V. K. G..V..SOD3.4 ................. G K. G.P.L..

SO03.1 ETTANODPLVTKGASLVPLLGIDVWEHAYYLQYKNVRPDYLNNIWKVMNWKYAGEVYENVLASOD3.2 ..............................................................soD3 .3 ..............................................................SOD3.4 ..............................................................

FIG. 2. (A) Nucleotide sequence of Sod3 cDNAs. The nucleotide sequence of Sod3.3 cDNA [fine (3)] is shown in full. The nucleotidesequences for Sod.3.1 (1), Sod3.2 (2), and Sod3.4 (4) are indicated only as they differ from the Sod3.3 sequence. The sequence-specific probesfor each are underlined. (B) Alignment of the deduced amino acid sequences for the four maize SOD-3 proteins. Gaps introduced to obtainmaximal homology are indicated by dashes. Residues conserved are indicated by dots. Transit-peptide cleavage site is underlined.

scripts accumulated in scutella during germination (Fig. 4).Sod3.2 transcript was first detected in scuteila at 2 dpi,increased dramatically by 3 dpi, and thereafter decreasedslightly by 10 dpi (Fig. 4B). The accumulation pattern ofSod3.1 and Sod3.4 transcripts was similar to that of Sod3.2(Fig. 4A and C). Trace amounts of the Sod3.3 transcript weredetected in scutella from 3 to 9 dpi (Fig. 4C). The maize Sod3transcripts were not detected in scutella at 1 dpi and were

differentially accumulated in the scutellum during earlysporophytic development.Accumulation of Maize Sod3 Transcripts in Embryos During

Kernel Development. RNA blot analysis indicated that allSod3 transcripts accumulated in embryos during embryogen-esis (Fig. 5). The Sod3.1 transcript was detected at 13 dpp andthe level remained constant from 13 to 23 dpp (Fig. SA). TheSod3.1 transcript decreased dramatically at 29 dpp and wasnot detected from 35 to 37 dpp (Fig. SA). Sod3.2 transcriptlevels remained fairly constant from 13 to 37 dpp (Fig. SB).Levels of the Sod3.4 transcript were low before 15 dpp,increased at 19 dpp, and remained constant from 21 to 37 dpp

(Fig. SC). Interestingly, the Sod3.3 transcript, which wasbarely detected in scutella during germination, was primarilyaccumulated in the maturing embryos (Fig. SC). Low levelsof Sod3.3 transcript were detected at 13 and 15 dpp, and itaccumulated dramatically from 19 to 37 dpp (Fig. SC). Thesedata suggest that expression of the Sod3.3 gene is embryo-specific, being expressed predominantly late in maize em-bryogenesis.

DISCUSSIONThe genes encoding the MnSODs of maize constitute amultigene family, unlike the single MnSOD gene reported inanimal and other plant systems. Multiple transcripts forMnSOD have been reported in human (20) and rat tissues(21), but in these cases, the multiple MnSOD mRNAs aretranscribed from a single functional gene and the multipleMnSOD transcripts result from alternative polyadenylylationin the rat (22) or alternative splicing and polyadenylylation inthe human (20). This report provides an example of multiple

TCGTCCTGCTTGCGTGCTACTCTGTCTCTGCTCCTAGTTTTTGGCATCATGTTTATGTTGAGCAAGGTGATGCCCAAGGGAAGCCATTCCCACTCTTGTCTCCATTAATATCGTCCATCCTGTTTGCGTGCCACTCCA CTTTCTGCTGCTCGTAGTGGCAGGCACATAGTATGTGCTTAACGTTAGACTAGCCGCTT AGATCACATGGTGATGCC

Dow

nloa

ded

by g

uest

on

May

6, 2

021

Proc. Natl. Acad. Sci. USA 90 (1993) 9313

1 2 3 4 5 6 7 8 9 10 11A

B

C

A-Sod3.?

B-Sod3.2

C-Sod3.4-Sod3.3V.

D-pHA2

13 15 19 21 23 26 29 31 35 37 (dpp)

-Soda2.. ..

_,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..t._..._...I......I.e.E.....E.EM;.-..

-Sod34

S ~~~~~~-Sod33

r -So3. 2

D-pHA2

FIG. 3. Tissue-specific expression of the maize MnSOD genes.RNA blot analysis of total RNA of root (lane 1), stem (lane 2), youngleaf (lane 3), mature leaf (lane 4), husk (lane 5), silk (lane 6), anther (lane7), young tassel (lane 8), mature tassel (lane 9), 4-dpi scutellum (lane 10),and 27 dpp embryos (lane 11). Total RNA samples (20 ,ug) wereelectrophoresed in a denaturing 1.6% agarose gel and blotted onto anitrocellulose filter. The same blot was sequentially hybridized withthe gene-specific probes for Sod3.1 (A), Sod3.2 (B), Sod3.3/Sod3.4(C), and pHA2 containing an 18S ribosomal sequence as a control (D).

MnSOD mRNAs generated by distinct genes. Molecularanalysis of the maize Sod3 cDNAs reveals that there are atleast four Sod3 genes in the maize genome, Sod3.1, Sod3.2,Sod3.3, and Sod3.4. The maize MnSOD cDNAs share sig-nificant coding sequence homology but differ in the 5' and 3'noncoding regions. A region of 154 nt was only found in the3' nontranslated region of Sod3.2. The nucleotide sequencesof Sod3.3 and Sod3.4 are nearly identical in both their codingand their 5' and 3' noncoding regions. Thus, unique DNAfragments for gene-specific probes could not be designed.The sizes of the two transcripts, however, are different invivo due to the consistent and reproducible presence of alonger poly(A) tail on Sod3.4. By using a common oligonu-cleotide probe it was possible to distinguish the two mRNAson the basis of size (see Materials and Methods). The twotranscripts generated by these two distinct genes also differin the number of an AGCG repeated motif within their 5'nontranslated leader sequence (four in Sod3.3 and five inSod3.4). The different length of poly(A) tails in these twotranscripts may be a reflection of dissimilar stability of themRNAs in different tissues in vivo.Accumulation of the Maize MnSOD Transcripts Is Tissue-

Specific. The various Sod3 genes are tissue-specific in theirexpression in maize, except for Sod3.1, which is variablyexpressed in all the tissues examined. These observationssuggest that the SOD-3 protein readily detected by zymogram

12 3 4A

B

C

5 6 7 8 9 10 (dpi)

-Sodw.

-Sod3.2

-Sod3.4

-Sod3.3

D,. pHA2

FIG. 4. Accumulation ofSod3 transcripts in maize scutella duringearly sporophytic development. Expression patterns of the Sod3genes in the scutella ofgerminating seedlings were examined byRNAblot analysis using Sod3 gene-specific probes. RNA was isolatedfrom scutella 1-10 dpi. Total RNA samples (20 jAg) were electro-phoresed in a denaturing 1.6% agarose gel then transferred onto anitrocellulose filter. The same filter was sequentially hybridized withthe Sod3.1 (A), Sod3.2 (B), Sod3.3/Sod3.4 (C), and pHA2 probes.

FIG. 5. Differential accumulation of Sod3 transcripts in maizeembryos during kernel development. W64A kernels were collectedfrom field-grown plants at 13-37 dpp. Total RNA samples (20 ,ug)from the dissected embryos (scutella plus axes) were electropho-resed in a denaturing 1.6% agarose gel and then blotted onto anitrocellulose filter. The same blot was sequentially hybridized withthe Sod3.1 (A), Sod3.2 (B), Sod3.3/Sod3.4 (C), and pHA2 probes.

analyses may be encoded mostly by Sod3.1 (23). The tran-scripts of Sod3.2 and Sod3.4 have been shown to be presentin scutella of both developing and germinating maize em-bryos. The Sod3.3 transcript can be predominantly detectedin the scutella of developing maize embryos. This tissue-specific expression of the Sod3 multigenes may specificallybe associated with mitochondrial biogenesis and specificrespiratory demands in the different tissues during plantgrowth and development.

Mitochondrial Metabolism and Expression of the NuclearMnSOD Genes in Scutelia During Maize Germination. Duringgermination, the growth rate of the embryonic axis increasesdramatically, and massive amounts of energy and nutrientsare required. It is assumed that the digestion and mobilizationof the storage macromolecules in the scutellum are energy-requiring processes; therefore, more ATP must be generatedby the mitochondria. The production of superoxide radicalsis rigorously coupled with the rate of mitochondrial respira-tion. The rapid increase in oxygen and respiratory activityresults in the increased production of superoxide inside themitochondria. The pattern of MnSOD RNA accumulation inthe scutella ofgerminating maize seedlings coincides with theincrease in mitochondrial respiratory activity during germi-nation. RNA blot analysis showed that levels of the maizeSod3 transcripts increased dramatically in scutella from 3 to10 dpi. Since the scutellum is a terminally differentiatedstorage tissue, this increase in Sod3 transcripts reflectsincreases in cellular metabolism and is not due to cell divisionor differentiation. Even though MnSOD transcripts cannot bedetected in scutella of 1-dpi seedlings, previous studiesshowed that active SOD-3 is present in maize scutella at thistime (7). This suggests that the MnSOD detected must be apreexisting protein inside the mitochondria. Resumption ofrespiratory activity during the early stages ofgermination hasbeen shown to depend on cytochrome-c oxidase and ATPasethat were conserved in the quiescent maize embryo ratherthan on newly synthesized or assembled enzyme complexes(24). We believe that superoxide generated at the early stagesof germination (at least the first 24 hr) similarly is scavengedby residual SOD-3 present in the mitochondria of dry seeds.Our present results imply that expression of the nuclear

MnSOD genes in maize is influenced by mitochondrial ac-tivity. It has been proposed that reactive oxygen intermedi-ates may serve as messenger molecules to modify the activityof transcription factors such as Fos and Jun (25) and NF-KB(26) in animal cells. We know very little about how expres-sion of nuclear genes is controlled by mitochondrial activityor what the signal transduction pathway between mitochon-dria and nucleus in plants might be. Though some evidence

r::

---- --- --- ---- --- ------

Developmental Biology: Zhu and Scandalios

Dow

nloa

ded

by g

uest

on

May

6, 2

021

9314 Developmental Biology: Zhu and Scandalios

in yeast supports the existence of a retrograde path ofcommunication from mitochondria to the nucleus (27), it isunknown what signals are directed from mitochondria to thenucleus, or how the nuclear genome receives the signals andresponds.

Developmental Expression of the Maize MnSOD GenesDuring Embryogenesis. An interesting observation in thisstudy is that accumulation of the Sod3.3 transcript has thecharacteristics of Lea (Late embryogenesis abundant) genesin maize and coincides with the advent of dehydration.Desiccation causes an increase in lipid peroxidation andproduction of activated oxygen free radicals in maize (28).These products of lipid peroxidation may induce oxidativestress in planta. Loss of desiccation tolerance appears to beassociated with a decline in SOD and peroxidase activities(28). In addition, abscisic acid and the antioxidant SODs areboth components of plant tolerance to some environmentalstresses, such as water stress (29). Abscisic acid-mediatedincreases in the steady-state level of the Sod3.3 transcripthave been seen in scutella of developing maize embryos(unpublished data). Therefore, we propose that the SOD-3isozyme encoded by Sod3.3 may have protective role(s) thatenable the embryo to acquire desiccation tolerance duringlate embryogenesis.Three other maize Sod3 transcripts also accumulate in

scutella during maize embryogenesis. It is conceivable thatmitochondrial biogenesis and respiration may be very activeduring seed maturation. Therefore, these physiological pro-cesses may produce more superoxide free radicals in thevarious parts of mitochondria. It is possible that a singleSOD-3 isozyme may not be sufficient to prevent cellulardamage from the increased oxidative stress and that allSOD-3 isozymes may be necessary to provide comprehen-sive protection.

Possible Implications of Multiple MnSODs Within MaizeMitochondria. All of the cloned maize Sod3 cDNAs encodehighly homologous but distinct MnSOD proteins. Structuralanalysis shows that all maize MnSODs have the followingcharacteristics: (i) a mitochondrial transit peptide and matrix-targeting sequences (the first 9 aa in the transit peptide); (ii)four conserved Mn ion ligand sites (His26, His2B, Asp163, andHis'67), and (iii) five residues (Gly75, Gly76, Phe84, Gln146, andAsp147) that are present only in MnSOD and absent fromFeSOD. These structural characteristics indicate that maizehas multiple forms of MnSOD associated with its mitochon-dria. Multiple maize Sod3 transcripts have been detected inother maize lines (unpublished data), indicating that thisphenomenon is not line-specific. We know little about whymaize has multiple forms of MnSOD in its mitochondria invivo. During evolution, the increased concentration of intra-mitochondrial superoxide in aerobic organisms may haverendered additional MnSOD an advantage for natural selec-tion. Therefore, random duplications of the maize MnSODgenes were maintained. Some ofthe-Sod3 gene copies may berequired to produce large amounts of MnSOD isozymesunder oxidative stress to provide adequate protection,whereas other genes are differentially and spatially expressedduring development and in response to different physiolog-

ical stresses. It is also possible that some of the maizeMnSOD isozymes may occupy different specific locationsinside mitochondria to provide efficient protection fromspecific environmental stresses.

We thank J. D. Williamson and L. Guan for helpful comments onthe manuscript. We are grateful to S. Ruzsa and S. Kernodle forexcellent technical assistance. This research was supported by aU.S. Environmental Protection Agency Research Grant (R814013) toJ.G.S.

1.

2.

3.

4.5.

6.

7.

8.

9.10.

11.

12.

13.

14.

15.

16.

17.

18.

19.20.21.

22.

23.

24.25.

26.

27.

28.

29.

Mead, J. F. (1976) in Free Radicals in Biology, ed. Pryor,W. A. (Academic, New York), Vol. 1, pp. 51-68.Brawn, K. & Fridovich, I. (1981) Arch. Biochem. Biophys. 206,414-419.Fucci, L., Oliver, C. N., Coon, M. J. & Stadtman, E. R. (1983)Proc. Natl. Acad. Sci. USA 80, 1521-1525.Baum, J. A. & Scandalios, J. G. (1982) J. Hered. 73, 95-100.Cannon, R. E. & Scandalios, J. G. (1989) Mol. Gen. Genet.219, 1-8.White, J. A. & Scandalios, J. G. (1988) Biochim. Biophys. Acta951, 61-70.Baum, J. A. & Scandalios, J. G. (1981) Arch. Biochem. Bio-phys. 206, 249-264.White, J. A. & Scandalios, J. G. (1987) Biochim. Biophys. Acta926, 16-25.Zhu, D. & Scandalios, J. G. (1992) Genetics 131, 803-809.White, J. A. & Scandalios, J. G. (1989) Proc. Natl. Acad. Sci.USA 86, 3534-3538.Scandalios, J. G. (1992) Molecular Biology of Free RadicalScavenging Systems (Cold Spring Harbor Lab. Press, Plain-view, NY).Williamson, J. D. & Scandalios, J. G. (1992) Plant J. 2, 351-358.Bowler, C., Alliotte, T., De Loose, M., Van Montagu, M. &Inze, D. (1989) EMBO J.I, 31-38.Sen Gupta, A., Heinen, J. L., Holaday, A. S., Burke, J. J. &Allen, R. D. (1993) Proc. Natl. Acad. Sci. USA 90, 1629-1633.Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl.Acad. Sci. USA 74, 5463-5467.Needleman, S. B. & Wunsch, C. P. (1970) J. Mol. Biol. 48,443-453.Beachy, R. N., Chen, Z. L., Horsch, R. B., Rogers, S. G.,Hoffman, N. L. & Fraley, R. T. (1985) EMBO J. 4, 3047-3053.Dellaporta, S. L., Woods, L. & Hick, J. B. (1983) Plant Mol.Biol. Rep. 1, 19-23.Kozak, M. (1983) Microbiol. Rev. 47, 1-45.Church, S. L. (1990) Biochim. Biophys. Acta 1087, 250-252.Ho, Y.-S. & Crapo, J. D. (1991) Am. J. Respir. Cell. Mol. Biol.4, 278-286.Hurt, J., Hsu, J.-L., Dougall, W. C., Visner, G. A., Burr, I. M.& Nich, H. S. (1992) Nucleic Acids Res. 20, 2985-2990.Baum, J. A. & Scandalios, J. G. (1979) Differentiation 13,133-140.Ehrenshaft, M. & Brambl, R. (1990) Plant Physiol. 93, 295-304.Abate, C., Patel, L., Rauscher, F. J., III & Curran, T. (1990)Science 249, 1157-1161.Schreck, R., Rieber, P. & A.Baeuerle, P. (1991) EMBO J. 10,2247-2258.Liao, X., Small, W. C., Srere, P. A. & Butow, R. (1991) Mol.Cell. Biol. 11, 38-46.Leprince, O., Deltour, R., Thorpe, P. C., Atherton, N. M. &Hendry, G. A. F. (1990) New Phytol. 116, 573-580.Skriver, K. & Mundy, J. (1990) Plant Cell 2, 503-512.

Proc. Natl. Acad. Sci. USA 90 (1993)

Dow

nloa

ded

by g

uest

on

May

6, 2

021