Proc. Natl. Acad. Sci. USAVol. 84, pp. 2798-2802, May 1987Developmental Biology

A maternal mRNA localized to the animal pole of Xenopus eggsencodes a subunit of mitochondrial ATPase

(determinants)

D. L. WEEKS AND D. A. MELTONDepartment of Biochemistry and Molecular Biology, 7 Divinity Avenue, Harvard University, Cambridge, MA 02138

Communicated by Joseph G. Gall, December 29, 1986 (received for review October 29, 1986)

ABSTRACT We have previously isolated several cDNAclones of mRNAs that have the unusual property of beinglocalized to either the animal pole or the vegetal pole of frogeggs. To gain insight into the function of these maternalmRNAs we have determined their DNA sequence and deducedthe sequence of the proteins they encode. Here we report thatAn2, an mRNA localized to the animal pole ofXenopus oocytesand eggs, codes for the a chain of mitochondrial ATPase.Furthermore, we compare the intracellular localization of theAn2 mRNA and mitochondria in oocytes and eggs and find thatthey do not have the same degree of localization. In the light ofthese results we discuss possible reasons for the maternallocalization of the An2 mRNA.

Embryologists have often proposed that maternal compo-nents deposited in the egg cytoplasm are responsible forspecifying cell fates during early development (1). In princi-ple, the differential distribution of cytoplasmic factors todifferent blastomeres provides a mechanism for distinguish-ing one daughter cell from another. A variety of studies,including the isolation of blastomeres from different parts ofembryos, demonstrate that different regions of the egg andembryo have distinct developmental fates. Though thesestudies provide indirect support for the existence of cyto-plasmic determinants, there are still few examples of well-characterized, localized maternal components that mightspecify cell fate (reviewed in ref. 2).We have examined the distribution of one type of potential

determinant, maternal RNAs, along the animal-vegetal axisof Xenopus eggs. We identified a rare class of localizedmRNAs and have obtained cDNA clones corresponding tosome of these messages (3, 4). Four members of this class ofmaternal RNAs, three RNAs localized to the animal pole(Anl-3) and one to the vegetal pole (Vgl) of unfertilized eggs,have been chosen for detailed study. Two of these RNAs(Anl and Vgl) are found only during oogenesis and pregas-trula development, whereas An2 and An3 are present at leastuntil the swimming tadpole stage. The isolation of thesecDNA clones confirms and extends other studies showingthat proteins (5) and translatable RNAs (6) are unevenlydistributed along the animal-vegetal axis in frog eggs.We are focusing on two aspects of these localized mRNAs.

First, what, if any, developmental function do these RNAs ortheir protein products have? Second, how are these RNAslocalized within a single cell? As a first step toward thesegoals, we have isolated and sequenced a nearly full-lengthcDNA clone for An2. The sequence of An2 reveals, some-what surprisingly, that this localized mRNA codes for the asubunit of mitochondrial ATPase.

MATERIALS AND METHODSMaterials. SP6 RNA polymerase and RQ1 RNase-free

DNase were obtained from Promega-Biotec (Madison, WI).Ribonucleases A and T1 were obtained from Sigma. Klenowpolymerase, DNA polymerase, and restriction enzymes wereobtained from Promega-Biotec and New England Biolabs.Radioisotopes were purchased from Amersham. Xenopuslaevis were purchased from Xenopus I (Ann Arbor, MI).

Sequencing. The DNA sequences of An2 cDNA cloneswere obtained by using the chain-termination method ofSanger et al. (7). Restriction fragments were cloned into M13vectors (8) and in all cases the sequence of both strands wasdetermined.

Extraction of RNA and DNA from Oocytes and Embryos.Frozen oocytes and eggs were sectioned as described (3).RNA and DNA were extracted with proteinase K digestionand phenol/chloroform extraction as described (9). LargeRNA (including mRNA) was separated from small RNAs(e.g., tRNA) and DNA (including mitochondrial DNA) byprecipitation with 4.0 M LiCl on ice for 1 hr. After centrif-ugation for 10 min at 13,000 x g, the pellet containing RNAwas washed with 70% ethanol and dissolved in water that hadbeen treated with diethyl pyrocarbonate. The DNA in thesupernatant was precipitated with ethanol, washed with 70%ethanol, and resuspended in sterile water.

Living oocytes were manually dissected with watchmak-er's forceps to obtain preparations of nuclei and cytoplasmfor analysis. In cases in which the nuclei were contaminatedwith yolk, this was eliminated by peeling off the nuclearmembrane with forceps.DNA and RNA Analysis. Mitochondrial DNA was analyzed

by cutting at a unique BamHI site prior to electrophoresis in1% agarose and Southern transfer (26) to GeneScreen (NewEngland Nuclear-DuPont). Blots were hybridized underconditions already described (10) to nick-translated mito-chondrial DNA isolated from pXIM31 (11), a clone contain-ing the full mitochondrial genome ofX. laevis. RNA analysiswas performed as described (12) using SP6-generatedantisense transcripts. The 32P-labeled RNA probes forhistone H4, An2, and Vgl transcripts were 450, 390, and 370nucleotides long, respectively. Densitometry of the autoradi-ograms was performed with an LKB Ultrascan densitometer.Computer Analysis. Sequence analysis was carried out on

an IBM-AT computer using Microgenie software from Beck-man. Protein data base from the National Biomedical Re-search Foundation was supplied with the Beckman software.

RESULTSSequence Analysis Suggests Homology to Mitochondrial

ATPase. An2 mRNA isolated from Xenopus eggs was previ-ously shown to be about 1.9 kilobases (kb) long (3). Theoriginal 1.6-kb An2 cDNA isolate was used as a probe torescreen an oocyte cDNA library for longer clones of thesame gene. One of the clones isolated, designated An2.1, is

2798

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.

Proc. Natl. Acad. Sci. USA 84 (1987) 2799

1.83 kb long or 96% of the estimated size of the full-lengthAn2 mRNA.The restriction map, DNA sequence, and protein encoded

by An2 are presented in Fig. 1. We note that the cDNA clonelacks a poly(A) tract and the consensus 3' processing signal(AATAAA), though we have shown that An2 RNA is presentin the poly(A)+ fraction of egg RNA (3). This suggests that atleast part of the nucleotides missing from the An2.1 cDNAclone are from the 3' end. The proposed protein encoded byAn2 begins at nucleotide 44 and ends at nucleotide 1679 togive a polypeptide with a deduced molecular mass of 58.9kDa. There are no other open reading frames longer than 70amino acids.The An2 protein was compared to the protein sequences

contained in the National Biomedical Research Foundationprotein data bank. The only significant matches that appearedwere with subunits of proton-translocating ATPases. Moststriking were the homologies (>50%) between An2 and the achains of F1/FO-type ATPases from tobacco chloroplast andEscherichia coli (285/491 amino acids for tobacco chloroplastand 168/287 for E. coli). In addition, there are two regionsthat share >70% amino acid homology. These two highlyconserved regions are thought to be involved in ATP binding(3), one of the activities of the a subunit of mitochondrialATPases. The specific residues conserved in ATP bindingproteins are indicated by the boxed sequences in Fig. 2a.The sequence data also reveal the presence of a putative

mitochondrial transport signal in the An2 protein. These

a

signals have no canonical amino acid sequence but seem tofollow three rules: (i) the transport signals are enriched forpositively charged residues, more or less periodically spaced,(ii) they are also enriched for hydroxylated amino acids, and(iii) they lack negatively charged amino acid residues (14).The sequence of the amino terminus ofAn2 conforms to theserules for transport into the inner mitochondrial matrix (Fig.2b). Thus, An2 contains the type of signal sequence expectedfor the a chain of mitochondrial ATPase and these data lendfurther support to our identification of the An2 protein.

In all animals so far studied, the a subunit of mitochondrialATPase is encoded by the nuclear rather than the mitochon-drial genome (14). An2 shows no sequence homology to thepublishedXenopus mitochondrial DNA sequence (15). More-over, when An2 DNA is used to probe isolated Xenopusmitochondrial DNA no signal is detected. Finally, a genomicSouthern blot using An2 as a probe confirms that An2 isrepresented in the Xenopus nuclear genome (data notshown).Comparative Localization of Mitochondria and An2 mRNA

in Oocytes and Eggs. Identification of the An2-encodedprotein as the a subunit of a mitochondrial ATPase raises thequestion of whether mitochondria are similarly localized inthe animal pole of oocytes and eggs. Coordinate localizationof mRNAs encoding mitochondrial proteins with the mito-chondria would help to explain the regional distribution ofAn2. To assay the distribution of mitochondria, frozenXenopus oocytes and unfertilized eggs were manually dis-

=1 OObpEcoRI Pst Hind 1II Cla Pvu 11 Bgl 11 EcoR

1 500 1000 1 1500 l1 1~~~~~~~~~~~~~~~~~~~~~~~~~

An2.1 cDNA -1820 bp- I

open reading frame - 544 amino acids

b

1oo 200

500 .OOI GCATGACA;AIGa[IdE1TAIT AAGGAAG6TGACAT T6T6AA6AEGAC AGGA6CIAT T6T66A1 6T6CCA46GTtG'ATGAGC GT GGT66GIC66116GGA'6C TCT66GTAACA"CTATTGATGG[AAGGGCCCTAIT GT6aCT AAAACICGCAGGAGAGTYGGTrCTAA4GGCCCCAGGCAICAT TCCCC6TPTCICIGSlle.AI aAsr AsplysLev I PI.yel I utl yAspl IIeV ILysAr;ThrG! yAI a11eValkApY^I roVa1 GIP6yApGueuLruG1 fArg92a1 Va AspAI *!ed61yAsnihr IIeAspE IltlysEI yPro I e61 y~erLysThrAr9ArgAf gVaI GI yLeuLysAI O~rdlyIIr1 e11eProArg IIe~erVa

700 800aAG66AACCCATtSCAGACtGST AT AA66 T6C'T TGAEA6TC T6ET16CCCAT T66CCGTG6TCAGC6TGAGCTCATTTAC6GG1ACAGAC4AGA T66CAAA4CCTCCl ATTGCCATCGATACCAT TATCAACIAGAGA6Ai4TCAACGAT6GGAACT SAT5AG'AAGAAGAAG TGTAC GT6AiT AT066C'CAT TGGGCAGAI Ar961 uPr owet6! nThr 61 y1!eLjsAi atal Asp~erl~euial Prol e61 yArg~ly61nArg6 uleul II el IIe61 yAspArgGInThr GIyLysThrSer 1II PA I a IIeAspThr II e 11 eAsn61 rtysArgPheAsnAsp6lyrThrAsp6luLysLysLrsLtuTyfCysl IIPylrvil al a II e6 yGI nL

Q,O) .IIvOA646AC T SAC!6A1SCCHAT6 C[AT6A6TACACCATT6TG6T6GTCIC6STACoCTC16P16CT6TCCCCT6CAA AICI1SC1 CGATPCTE T6 ITCCA 66GAGb6TAT TCAGAGACAATGGAACGCAC6 TTTCATCATGATGATp lTC AA6CA66Cl6TA6CC aCCGTCA6ATGIC IC T6ysAr 9LeuThr AspAIaAspAI atetLyslyr Thr I11 eVa IVa I Ser Ar gThrAl aSefAspA I aAI ~r ol.eu61 n IyrLeuAI O~r ol r SerGI yC ys~er Met 61 y6lul yr tek gAspAs,)61 y I r HisAAI alteu 11 el IeT yf spAspituSerL ys6lnA I OVaI Al a yr Arg ncetSerLeuteu

Q100 1200CT6CGIC1.6CC ICC IG CTGAGGCC IACCCTG66GACOTCI ICI T CAC I CCC6CtTTCT6AAAGAGCAGCCAAAA'TGACGATCAC I 666CGrC CC TGACT GE TCTC TCAT I AACCEAGGCCGGT5ATGTGTCCGCT T ACATCCCAACCAAT6GCATCTCCAICACT6ACGGACAGAT TLtuAr gAr gProPro61 ryAfGIuA I aTyrPr o6 Asp'al PhelroLeuwl sSerArgleuLtu61 uArqAI 4A IaLys~etAsnkspHj sPhe61 y61!Gt6ySefLeulhrAla~euProVaII1I PS! uhr GI1rA a61 >AspVa I SerAI aTyr11eProftfsn~ai IIIeSer II PThr AsP y61nl II ePh

* 1300 1. 400C TT 6GAGACA6A51 T6T TCTACAA6GGT AI CC6GCC TGCT AltAAC6T6G6GTC T6TC TG6TGTCCA6A61G66ATCAG[ I SC T CAP.ACCA6AGC I ATG6AAACA66TG66CC66T AC AATGAAGCTGCtAGTTG16C I CA6 T ACC6TGAA611T6CTGC I TTT GECCCA61 TC66CTCAGA!T TGGAT GETT6C IACACAACAACt CCPLeuGIu Ihr61 uLeuPhle IrLysGl 11 eAr gPrOal.11 eAsnVaIS61yLeuSer Va Ser ArgVaI GI ySer AIaA I a1 InThrAr 9A daE-t lysG nVal AI GIyrThrwetLysLeu6I uLeuAI a61nT yr ArgG I uValAl jAI O~heA I a61 nPhe6l yStrAspLeuAspAI Wh aThr61I6nLeuL

500 1607.0

I ICCTT6CTC C611AAG~~~~~~~AO60CAACACCAGGAOCIICITSAATCGTOAACCGCSG61t66C6TCT6ACT6AGTT6C'6AAGCAAGGACAATACGTTCCCAT6GGCcATC6AAGAACAG66AAqCCGAA TC T A6CC66TGTCA6G6GACATC T1 6CAA6iAT 66GCCCASCAAAAT CACAAA6 ITT6AGAGc XICGC I 1 C Ac 1 GCTA6CACACA661ItITCI PATCAGGeuAsnArgG1 yVaI0Ar9LeuThr61 uLeuLeuLy~s1In61y61InTyrVaI ProMetAI SIe1uW1u61InVa Thrval 1I1eTyrAlEI yVaI Ar96 yHsLeuAspLys~et G roSerlys I1tIhrlysPheGu~erAlaPhPLeu AI is ILysSer6In 1 I6 61

1700 18006C16AI66AAA6ATCTCC6AACAGGCC6AT6CCAAGCTCAAAGAAATCSTCCAA4C T TCCTGTCCACAT TC6AA6CATAAAI ICC6661 TC ICC66ACCASXTT616A16 CAC6TTCCCAAT16l TCC 661 ITTCATrTCT T6CIAAACT TATTGAAGCAC1T6T6CTtI TAAT6TACA6AAAI TC~TTAAT66A6AAliAsp5plyyslleSer6luGln4laAspOiaLysLeuLys61ulleValLeuAsnPheLeuSerThrphe6luAliaEnd

67AAAC7TTCATC6AAAATC66AATTC

FIG. 1. Structure of a cDNA clone coding for a localized maternal mRNA. (a) Restriction map and schematic representation of cDNA cloneAn2.1 indicating the position of the open reading frame. EcoRI sites shown are from linkers used in cDNA library construction. bp, Base pairs.(b) DNA and deduced protein sequence of An2.1.

AACATCCACCC6CAA5CASTCACCA6ACTCCCAAGCC6E46TCAYGTTGTCA6TCC6761AGCCGCCGCGCT66CCC6C6(CTTGCCTC61CA61CAGGCCT6616TCCAAGAAA6CCCTCGGT6CTGCTTTTGTTSCAACCASAAACATTCAT6CATCT6GCGTGTGGCCTCCAGAGAAGTCTGGCACTGCTGAAGTGT

Developmental Biology: Weeks and Melton

2800 Developmental Biology: Weeks and Melton

a* * * * * *

An2 185 Ser Val Arg Glu Pro Met Gin Thr Gly Ile Lys Ala Val Asp Ser Leu Val Pro Ile Gly

Esherichia coli 141 Ser Val Asp Gin Pro Val Gin Thr Gly Tyr Lys Ala Val Asp Ser Met Ile Pro Ile GlyTobacco chloroplast 142 Ser Val Tyr Gin Pro Leu Gin Thr Gly Leu Ile Ala Ile Asp Ser Met Ile Pro Ile Gly

* * * * * * * * * * * * * * * * *

Arg Gly Gin Arg Glu Leu Ile Ilc Gy1 Asp Arg Gin Thr Gly Lys Thr Ser Ile Ala Ile

Arg Gly Gin Arg Glu Leu Ile Ile Gly Asp Arg GInIThr Gly Lys Thr Arg Leu Ala Ile

Arg Gly Gin irgPlu Leu Ile Ile GlAsp Arg Gin I h Gly Lys]Thr Ala Val Ala Thr

An2 263 Tyr Thr Ile Val Val Ser Arg Thr Ala Ser Asp Ala Ala Pro Leu Gln Try Leu Ala Pro

Eshenchia coli 220 Asn Thr Ile Val Val Val Ala Thr Ala Ser Glu Ser Ala Ala Leu Gln Try Leu Ala Pro

Tobacco chloroplast 221 Tyr Thr Ile Val Val Ala Glu Thr Ala Asp Ser Pro Ala Thr Leu Gln Try Leu Ala Pro

Tyr Ser Gly Cys Ser Met Gly Glu Tyr Phe Arg Asp Asn Gly Thr His Ala Leu Ile IleMet Pro Val Ala Leu Met Gly Glu Tyr Phe Arg Asp Arg Gly Glu Asp Ala Leu Ile Ile

Tyr Thr Gly Ala Ala Leu Ala Glu Tyr Phe Met Tyr Arg Gly Arg His Thr Leu Ile Ile

Tyr Asp Asp Leu Ser Lys Gln Ala Val Ala Tyr Arg Gln Met Ser Leu Leu Leu Arg ArgTyr Asp Asp Leu Ser Lys Gln Ala Val Ala Tyr Arg Gln Ile Ser Leu Leu Leu Arg Arg

Tyr Asp Asp Pro Ser Lys Gln Ala Gln Ala Tyr Arg Gln Met Ser Leu Leu Leu Arg Arg

* * * * * *

Pro Pro Gly Arg Glu Ala Tyr Pro Gly Asp Val Phe Tyr Leu His Ser Arg Leu Leu Glu 343

Pro Pro Gly Arg Glu Ala Phe Pro Gly Asp Val Phe Tyr Leu His Ser Arg Leu Leu Glu 300

Pro Pro Gly Arg Glu Ala Tyr Leu Gly Asp Val Phe Tyr Leu His Ser Arg Leu Leu Glu 301

b

An2 Amino terminal End

Met Leu Ser ValI Val Ala Ala Ala Leu Al4{ Ala Leu PrcJ3 Gln Ser Gly Leu Val SerEy-sLy-slAla Leu Gly Ala

FIG. 2. Protein homology between An2 and ATPases. (a) Protein homology between An2, E. coli ATPase a chain, and tobacco chloroplastATPase a chain. Asterisks indicate shared homology between all three sequences. ATP binding domains (13) are enclosed in boxes. (b) Aminoterminus of deduced An2 protein. Positively charged amino acids are enclosed in boxes; hydroxylated amino acids are underlined.

sected into an animal pole third or a vegetal pole third or wereundissected. In these three fractions, mitochondrial DNAwas separated from large RNAs by precipitation of the RNAwith 4 M LiCl. Mitochondrial DNA was cut at a uniqueBamHI site, fractionated by agarose gel electrophoresis, andblotted onto GeneScreen membranes. The filters were hy-bridized with a 32P-labeled probe made by nick-translatingplasmid pXIM31, which contains the entire Xenopus mito-chondrial genome (11). RNA isolated from the variousfractions was examined by RNase protection assays using32P-labeled antisense RNA probes for histone H4, An2, andVgl mRNAs.The results presented in Fig. 3a show that An2 mRNA is

significantly more concentrated in the animal pole than aremitochondria. Quantitation of the autoradiograms by densi-

tometer tracings (Fig. 3b) shows that the mitochondria areslightly concentrated in the animal pole with an animalpole/vegetal pole ratio of 1.5 in oocytes and 3 in eggs. Thisslight gradient closely parallels the distribution of histone H4RNA. In fact, most cellular components other than yolk,including total poly(A)+ RNA (16), are slightly more concen-trated in the animal pole. In contrast, An2 mRNA shows a 4-to 8-fold animal pole localization in oocytes and a 10- to15-fold enrichment in the animal pole of eggs. The increasedlocalization of mitochondrial DNA, An2 mRNA, and histoneH4 mRNA in eggs relative to oocytes may be a consequenceof oocyte maturation. A control for these experiments, VglmRNA, has the opposite orientation for its enrichment and isabout 20 times more abundant in the vegetal pole in eggs andoocytes. We conclude that An2 mRNA, which most probably

205

160

161

Proc. Natl. Acad. Sci. USA 84 (1987)

Proc. Natl. Acad. Sci. USA 84 (1987) 2801

An2Nuc Cyt Tot

HistoneNuc Cyt Tot

_b

w

Histone

An2

...! _Mo . . . .. ...... ____

Vgl

b.DNA or RNA [Animal polel/ [vegetal polelOocytemitochondrial DNA 1 to 2histone RNA 1.2 to 2An2 RNA 4 to 8Vgl RNA < 0.05

mitochondrial DNA 3histone RNA 2 to 4An2 RNA 10 to 15Vgl RNA < 0.05

FIG. 3. Distribution of maternal components along the animal-vegetal axis in oocytes and eggs. (a) Autoradiographs of mitochon-drial DNA Southern blots and ofRNase protection assays using total(T) oocytes or eggs, animal pole (A) thirds, and vegetal pole (V) thirdsof oocytes or eggs. Southern blots were probed with 32P-labelednick-translated mitochondrial DNA. RNase protection assays usedSP6 probes homologous to histone H4, An2, or Vgl mRNAs.Southern blot analysis was carried out on 3 total and 10 sections ofoocytes or eggs. RNase protection assays were carried out on 1 totalor 3 sections of oocytes or eggs. Protected fragments are approxi-mately 360 nucleotides (H4), 370 nucleotides (An2), and 355 nucle-otides (Vgl) in length. (b) Relative distribution of mitochondrialDNA and histone, An2, and Vgl RNA as determined by densi-tometry of autoradiographs including those shown in a.

codes for a subunit of mitochondrial ATPase, is moreconcentrated in the animal pole than are mitochondria.An2 mRNA Is Not Localized in the Germinal Vesicle. We

have begun to investigate how the animal localization ofAn2RNA might be accomplished. One possibility is that the An2RNA is sequestered in the oocyte's nucleus or germinalvesicle, as has been demonstrated for specific histonemRNAs in sea urchin oocytes (17). It has been proposed thatthe nuclear location of the histone mRNAs may preventtranslation until the nuclear membrane is dissolved. Becausethe nucleus of full-grown frog oocytes lies in the animal pole,localization of An2 mRNA inside the nucleus could accountfor its animal pole localization. We have dissected stage VIoocytes and analyzed the distribution of the An2 mRNA innuclei and cytoplasm. The RNase protection assays (Fig. 4)reveal that the An2 mRNA is found almost entirely in thecytoplasm. In situ hybridization experiments (18) have al-ready shown that histone mRNA is mainly found in thecytoplasm of oocytes as opposed to the nucleus. Similarly,using RNase protection assays (Fig. 4), histone H4 mRNA ispredominantly found in the cytoplasm. We note a faint banddetected with the histone probe in the nuclear RNA. Thismay be due to unprocessed histone transcripts and/or read-through transcripts known to occur in lampbrush chromo-somes (19). In any case, we conclude that the animal polelocalization of the An2 mRNA is not due to sequestration inthe nucleus.

FIG. 4. Intracellular distribution of An2 and histone mRNAs inoocyte nuclei and cytoplasm. RNase protection assays for An2 andhistone H4 mRNAs were performed with full-grown, stage VIoocytes. RNAs were isolated from five nuclei (Nuc), cytoplasm offive enucleated oocytes (Cyt), and five whole oocytes (Tot). Expo-sure times were 48 hr for the An2 autoradiograph and 6 hr for histone.

DISCUSSIONThe sequence of the An2 cDNA allows us to identify itsprotein product as aXenopus mitochondrial ATPase a chain.The mitochondrial ATPase, of which the a chain is a part, isresponsible for the generation of ATP from ADP and Pi andalso for the generation of an ATP-driven electrochemicalgradient (20). The protein encoded by An2 is the proper sizeto be the a chain of this mitochondrial ATPase, is strikinglyhomologous to other a-chain sequences, and has a suitableamino terminus for a mitochondrial import signal sequence.Final proof of this assignment must await the production ofantibodies to assay for the An2 protein in mitochondria.The reason for the strong animal pole localization of the

An2 mRNA is not clear. As noted above, An2 mRNA islocalized in a much steeper gradient than are mitochondria.In this regard, it may be helpful to recall some details aboutthe biogenesis of mitochondria in frog development. All ofthe mitochondria found in eggs and early embryos are ofmaternal origin; indeed they are synthesized in pre-stage IVoocytes (21). There is no new synthesis of mitochondria untilthe tadpole stage (stage 32, ref. 22), some 2 days afterfertilization. In oocytes, mitochondria appear to be synthe-sized in the so-called mitochondrial cloud, which is locatedon what will eventually become the vegetal pole side of thenucleus. In addition, some mitochondria are produced else-where in the cytoplasm at discrete "foci of proliferation"(23). We do not know if the An2 mRNA is translated orlocalized in early oocytes. Nevertheless, one can speculatethat mRNAs coding for mitochondrial proteins would bepositioned near the mitochondrial cloud but excluded fromthe vegetal pole side by the mass of the mitochondrial clouditself.Another possibility is that An2 mRNA is localized to

accomplish the generation of a respiration gradient duringdevelopment. There is an animal-vegetal gradient of respi-ration in developing embryos (24). By middle to late blastula,cells in the animal pole of the embryo are more activelyrespiring then those in the vegetal pole, and dorsal cells aremore active than ventral cells. In axolotls, changes in thestructure of mitochondria, including expansion of innermitochondrial surface area during early embryogenesis, canbe correlated with the spatial differences in respirationdescribed by Brachet (24). The largest inner mitochondrialsurface areas are found in the animal pole cells (25). Regionaldifferences in respiration, and therefore ATP synthesis,correspond to the higher rate of cell division in the animalpole and also with the prospective energy needs related tomotility of these cells during gastrulation. Neither newmitochondria nor new mRNA for the a subunit of mitochon-drial ATPase (An2) is synthesized while this gradient forms(3, 22). In light of previous reports (23) and the mitochondriallocalization data presented above, differences in mass ofmitochondria seem an unlikely mechanism for the establish-

a.

T A V

PROBE

Mitochondria

T A V

Developmental Biology: Weeks and Melton

2802 Developmental Biology: Weeks and Melton

ment of the respiration gradient. We have previously shownthat the animal localization of An2 mRNA is maintainedduring the cleavage divisions and the message is preferen-tially segregated to the animal pole cells of blastula. One cantherefore speculate that the regional activation of mitochon-dna that is responsible for the respiration gradient is medi-ated by translation of localized An2 mRNA. This regionalproduction of the a subunit for the ATPase could preferen-tially activate mitochondria in the animal region.One tends to think that likely candidates for maternal

cytoplasmic determinants are factors that directly affect geneexpression. These would include factors that affect genetranscription, mRNA translation, or protein processing andactivities in a region-specific manner. And perhaps examplesof these types of factors will be identified as the character-ization of localized molecules continues. At the same time,the identification of the protein product of An2, the a subunitofthe mitochondrial ATPase, provides an example ofanothertype of molecule that may serve as a local determinant. Thesemolecules, by affecting processes like respiration, ion flow,or metabolism, might regionally alter the physiological stateof blastula cells. Such changes in cell physiology may enablethis type of localized molecule to exert an effect on prospec-tive cell fate.

We thank Jan Fassler, Richard Harvey, Chris Kintner, HeatherO'Keefe, and Ariel Ruiz i Altalba for comments on the manuscript.We also thank Guido Guidotti for helpful discussions and commentsand Nick Hopwood for assistance in the characterization of An2.1.D.L.W. acknowledges support from National Institutes of HealthPostdoctoral Fellowship GM11213-01. This work was supported bygrants from the National Institutes of Health and The ChicagoCommunity Trust/Searle Scholars Program.

1. Wilson, E. B. (1925) The Cell in Development and Heredity(Macmillan, New York).

2. Davidson, E. (1986) in Gene Activation in Early Development(Academic, New York), pp. 411-524.

3. Rebagliati, M. R., Weeks, D. L., Harvey, R. P. & Melton,D. A. (1985) Cell 42, 769-777.

4. Weeks, D. L., Rebagliati, M. R., Harvey, R. P. & Melton,D. A. (1985) Cold Spring Harbor Symp. Quant. Biol. 50,21-29.

5. Moen, T. L. & Namenwirth, M. (1977) Dev. Biol. 58, 1-10.6. King, M. L. & Barklis, E. (1985) Dev. Biol. 112, 203-212.7. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. NatI.

Acad. Sci. USA 74, 5463-5467.8. Messing, J. & Vieira, J. (1982) Gene 19, 269-276.9. Melton, D. A. & Cortese, R. (1979) Cell 18, 1165-1172.

10. Krieg, P. A. & Melton, D. A. (1985) EMBO J. 4, 3463-3471.11. Rastl, E. & Dawid, I. B. (1979) Cell 18, 501-510.12. Melton, D. A., Krieg, P. A., Rebagliati, M. R., Maniatis, T.,

Zinn, K. & Green, M. R. (1984) Nucleic Acids Res. 12,7035-7057.

13. Walker J. E., Saraste, M., Runswick, M. J. & Gray, N. J.(1982) EMBO J. 1, 945-951.

14. Hurt, E. C. & van Loon, A. P. G. M. (1986) Trends Biochem.Sci. 11, 204-207.

15. Roe, B. A., Ma, D.-P., Wilson, R. K. & Wong, J. F.-H. (1985)J. Biol. Chem. 260, 9759-9774.

16. Phillips, C. R. (1982) J. Exp. Zool. 223, 265-275.17. Showman, R. M., Wells, D. E., Anstrom, J., Hursit, D. A. &

Raff, R. A. (1982) Proc. Natl. Acad. Sci. USA 79, 5944-5947.18. Jamrich, M., Mahon, K. A., Gavis, E. R. & Gall, J. G. (1984)

EMBO J. 3, 1939-1943.19. Diaz, M. 0. & Gall, J. G. (1985) Chromosoma (Berlin) 92,

243-253.20. Scarborough, G. (1986) Proc. Natl. Acad. Sci. USA 83,

3688-3692.21. Webb, A. C. & Smith, L. D. (1977) Dev. Biol. 56, 219-225.22. Chase, J. W. & Dawid, I. B. (1972) Dev. Biol. 27, 504-518.23. Heasman, J. J., Quarmby, J. & Wylie, C. C. (1984) Dev. Biol.

105, 458-469.24. Brachet, J. (1974) in Introduction to Molecular Embryology,

Heidelberg Science Library (Springer, New York), Vol. 19,pp. 117-119.

25. Nelson, L., Lorentzon, R., Boquist, L. & Lovtrup, S. (1982)Exp. Cell Res. 137, 25-29.

26. Southern, E. M. (1975) J. Mol. Biol. 98, 503-517.

Proc. Natl. Acad. Sci. USA 84 (1987)