Expression of Chloroplast and Mitochondrial Genes during ... - NCBI

8 downloads 0 Views 1MB Size Report
Lee SLJ, Warmke HE (1979) Organelle size and number in fertile and T-cytoplasmic male-sterile corn. Am J Bot 66: 141-148. 15. Lonsdale DM, Hodge TP, ...
Received for publication April 26, 1991 Accepted December 8, 1991

Plant Physiol. (1992) 99, 396-400

0032-0889/92/99/0396/05/$01 .00/0

Expression of Chloroplast and Mitochondrial Genes during Microsporogenesis in Maize1 Fran9oise Moneger, Paul Mandaron, Marie-Fran9oise Niogret, Georges Freyssinet, and Regis Mache* Laboratoire de Biologie Moleculaire V6getale, Unit6 1178 du Centre National de /a Recherche Scientifique, Universit6 Joseph Fourier, BP 53, 38041 Grenoble Cedex, France (F.M., P.M., M.-F.N., R.M.); and Biologie Moleculaire et Cellulaire Vegetale, Rhone-Poulenc Agrochimie, BP 9163, 69263 Lyon Cedex 09, France (G.F.) ABSTRACT

chondrial genome leading to the production of variant polypeptides (at least in the case of CMS-T), which are proposed to affect development of the male flower. Lee and Warmke (14) have shown by EM studies that the number of mitochondria increases in anthers of both normal and T-sterile lines of maize. This increase takes place both in the sporogenous cells and in the tapetum of anthers. The authors suggested that the rapid multiplication of mitochondria could induce a condition of stress to which the male sterile lines of maize respond differently than the normal lines. This hypothesis emphasizes the importance of genetic events that occur in mitochondria during the normal development of pollen. Following meiosis in the male flower of maize, each microspore develops individually. After vacuole formation in the unicellular microspore, two successive haploid mitoses occur, giving rise to a large vegetative cell and two sperm cells. Starch is accumulated in the amyloplasts contained in the vegetative cell until maturation of the pollen. A method has been developed in our laboratory to isolate microspores at different steps of their development without contamination by diploid cells from anther tissues ( 17). We present here the study of the expression of some chloroplast and mitochondrial genes in isolated microspores at different steps of their development. The results show that gene transcripts from mitochondria are present at higher steady-state levels at mid-term of the microsporogenesis compared with leaves. Following the two mitoses in microspores, plastid transcripts, with the exception of a very low amount of 16S rRNA, are no longer detectable.

Mitochondrial and plastid gene expression has been examined during maize (Zea mays) microsporogenesis. Accumulation of transcripts was found for three mitochondrial genes studied (cob, atp6, and atp9) at the mid-term of pollen development. In contrast, these mitochondrial transcripts were undetectable in mature pollen. Southern and DNA gel blot experiments showed that the copy number of mitochondrial genes was amplified in microspores at stages preceding the accumulation of these transcripts. Plastid transcripts of the photosynthetic psbA and rbcL genes could not be detected after the two mitoses, whereas precursors of the 16S rRNA are detected at low levels.

Cytoplasmic maternal inheritance and CMS2 are two biological traits showing the importance of organelles during microsporogenesis. In the case of maize, genetic and cytological evidence has been given for the lack of paternal inheritance of plastid and mitochondrial DNA (2, 4, and references therein). Different authors have tried to explain the mechanisms implied in maternal inheritance (5) and different hypotheses have been brought forward. One of these hypotheses suggests (7) destruction of the paternal plastid genome during microsporogenesis in barley to account for maternal inheritance. Knowledge about the persistence and integrity of organelle DNA during pollen formation would therefore help to evaluate this hypothesis. Recently, pollen DNA from Medicago sativa or Antirrhinum majus, known to transmit organelle genes biparentally and maternally, respectively, was probed with a plastid-specific DNA fragment (5). It was shown that the presence or absence of plastid DNA correlated with the type of inheritance. It would be interesting to know whether this result might be generalized and valid in other plant species. In the case of CMS, blocks occur in critical steps in the development of the male flower. CMS has been extensively studied in maize and it is now accepted that the CMS phenotype is correlated with recombination events in the mito-

MATERIALS AND METHODS Growth Conditions for Maize

Seeds of two fertile lines of maize (Zea mays), MO 17 and RP 704, were provided by Rh6ne Poulenc Agrochimie. Plants were grown in a greenhouse under defined conditions: light, 50,000 lux; temperature, 22°C in the day (15 h) and 20°C in the night (9 h); RH approximately 60%. Plants produced pollen after 2 months. Preparation of Microspores The timing of differentiation of the microspores follows a gradient along the tassel. To minimize heterogeneity before the preparation of microspores, tassels were separated in three groups (17): (a) young tassels that are still embedded in the

This work was supported by Rh6ne Poulenc Agrochimie. M.F.N. is financially supported by a grant from l'Institut de la Recherche

Agronomique. 2 Abbreviations: CMS, cytoplasmic male sterility; PM, premature; M, mature; S, starch-containing microspores; V, vacuolated microspores; bp, base pairs; kb, kilobase pairs. 396

ORGANELLE GENE EXPRESSION DURING MICROSPOROGENESIS

leaves and contain microspores with a vacuole (V) and (b) mid-tassels that are clearly visible and contain microspores accumulating starch (S1, S2). The third group contained mature tassels with dehiscent anthers. These contained the premature microspores full of starch (PM) and completely mature microspores (M). Spikelets collected from tassels of the same group were quickly ground in water with a Polytron. The homogenate was fractionated by successive filtration through 120- and 50-,um mesh size metallic sieves. For young tassels, a 50-,um mesh size sieve was needed to collect the Vstage microspores. For mid-tassels a 30% Percoll cushion was used to separate the V-stage microspores from the S microspores that pelleted through the cushion. The mature tassels were shaken to collect mature pollen grains released from dehiscent anthers. The remaining flowers, still closed, contained PM-stage microspores that were nearly mature compared with the S2-stage microspores. After filtration, microspores were collected and sedimented. These preparations were free of other tissues and debris. Viability of M microspores was checked by incubation in germination medium according to Cook and Walden (3). Material was either used immediately or pelleted, frozen in liquid nitrogen, and stored at -700C.

Nucleic Acids Preparation from Microspores

Microspores were ground in a liquid nitrogen-cooled Spex 6700 Freezer Mill (Bioblock, Strasbourg, France). The extraction of nucleic acids from the resulting powder was performed in lysis buffer (50 mm Tris-HCl, pH 7.5, 150 mm NaCl, 2% SDS, 2% sodium sarcosinate, and 10 mm EDTA) and deproteinized using phenol/chloroform according to ref. 18. Purification of DNA was achieved by CsCl centrifugation. Difficulties were encountered in estimating the amount of DNA to be used for hybridization due to the presence of contaminating polysaccharides in CsCl-purified microspore DNA. To overcome this problem, the initial CsCl-ethidium bromide centrifugation was followed by a centrifugation in CsCl-bisbenzimide, which separated DNA from the starch (16). In the case of DNA extracted from M pollen, polysaccharides were still present even after this treatment. The DNA obtained was found to be of sufficient purity for hybridization with different probes. To prepare RNA, total nucleic acid extracted from microspores was precipitated with cold ethanol, pelleted, and resuspended in water. High mol wt RNA was prepared by differential precipitation from LiCl as described (18) and used for further analysis. Chloroplast DNA Preparation Chloroplasts were purified on a continuous 20 to 60% sucrose gradient and DNA was isolated from green leaves of maize plants following conventional procedures and followed by centrifugation in CsCl gradients. Southern, Northern, and DNA Gel Blot Experiments Total DNA was restricted using BamH 1 and EcoR 1 (Boehringer) restriction endonucleases, electrophoretically sepa-

397

rated and transferred to nylon membrane using standard procedures. "DNA gel blot" experiments were done in the same manner as Southern experiments but the DNA was not restricted with any enzyme prior to loading. Ten or 15 Ag of total cellular RNA were size fractionated by electrophoresis in a 1.2% agarose/5 mM hydroxymethyl mercury gel, stained with ethidium bromide, and transferred to nylon membrane by capillary blotting in 10 x SSC. Chloroplast probes were from spinach. Nuclear and mitochondrial probes were from maize. The following DNA fragments were used as probes: the 750 bp XbaI-PstI intragenic fragment of the psbA gene (26); the 550 bp SaIl-EcoRV fragment containing the rpsl9 gene (27); the 1200 bp EcoRI-PstI intragenic fragment of the rbcL gene (30); the 1200 bp EcoRI-BamHI intragenic fragment of the rrn 16S gene (1); the 1200 bp EcoRI-HindIII intragenic fragment of the atp6 gene (8); the 2200 bp XbaI fragment containing the atp9 gene (9); the intragenic 680 bp HindIII-EcoRI fragment of the cob gene(6); the 1000 bp EcoRI intragenic fragment of the rbcS gene. Labeling of inserts was done by the random hexamer priming method to a specific activity ranging from 1 to 5 x 109 cpm/,ug. Prehybridization and hybridization conditions were at 42°C in 50% formamide, 1 M NaCl, 1% SDS, and 0.2 mg/mL salmon DNA. Samples (108 cpm) of each probe was used for hybridization. Washes were performed 10 min at room temperature in 2 x SSC, 1 h in 2 x SSC, 1% SDS at 65C, and 1 h at room temperature in 0.1 x SSC. RNA molecular size markers (0.3-7.4 kb) from Boehringer were used in the northern experiment with the rrn 16S probe; the 5.3 and the 1.6 kb RNA fragments of the marker are revealed with the labeled probe. Experiments were made from two to four times,

independently. RESULTS Expression of Chloroplast and Mitochondrial Genes in Microspores Homogeneous populations of microspores were used from the following stages defined in "Materials and Methods": V, SI, S2, PM, and M. To obtain experimental data on plastid genes during microsporogenesis, northern-type experiments were carried out using total RNA extracted from microspores at these stages and from leaves with cpDNA probes containing intragenic fragments of psbA, rbcL, and rrn 16S genes (Fig. 1). As can be seen after ethidium bromide staining, the amount of RNA for samples of mature stages (PM and M) is lower than for the young stages, although the same amount of RNA, as estimated by UV absorbance, has been loaded on the gel. We observe that many additional bands are present in young microspores (V and S 1-2), which tend to disappear in older stages. These bands seem to reflect the in vivo situation instead of a degradation during the extraction process because an identical electrophoretic pattern was obtained with RNA extracted from fresh whole anthers immediately dropped into liquid nitrogen after harvesting (not shown). The origin of these bands remains unknown but seems to be specific for young microspores. A very small amount of psbA transcript, but no rbcL transcript, was detected in vacuolated microspores (V stage) at the beginning of microsporogenesis.

MONEGER ET AL.

398 Figure 1. Hybridization of the chloroplast psbA, rbcL, and rrn 16S gene probes to RNA isolated from microspores at different stages of development (V, Si-2, PM, M) and from leaves (L). Ten micrograms of RNA were size separated on hydroxymercury gel. Ethidium bromide fluorescence of RNAs is shown on the left side. Probes were labeled as indicated in "Materials and Methods." Exposure time was of 24 h (psbA), 96 h (rbcL), 10 h (rrn 16S with microspore RNA), or 6 h (rrn 16S with leaf RNA). Size of transcripts are indicated in kb.

V

SI2 PM M

416 5

iL

V SI.. PM M

qw

rbc

psbA

related to a disappearance of the plastid genome. The psbA, rbcL, and rpsl 9 genes have been located in specific restricted fragments of maize chloroplast DNA by other authors (13). The psbA gene is in the 4.6-kb BamHI-8 fragment, the rbcL gene in the 4.35-kb BamHI-9 fragment, and the rpsl9 gene in the 5.3-kb BamHI-6 and 4.6-kb BamHI-8 fragments or in the 2.2-kb EcoRI-m fragment. In our experiments, these characteristic fragments were detected in leaf DNA as well as in the microspore DNA (Fig. 3). Plastid DNA is also present in M pollen as shown by the identification of the BamHI restriction fragment bearing the psbA gene. The three chloroplast gene probes reveal additional bands that are detected after a longer exposure in microspore DNA but not in leaf DNA: a 12- and a 5.6- kb BamHI fragment containing sequences homologous to the rbcL gene, two EcoRI fragments (of 4 and 1.8 kb, respectively) and one 10.5-kb BamHI fragment revealed by the rpsl9 probe, and a 6.7-kb BamHI fragment revealed by the psbA probe. By their sizes, these cannot be explained as incomplete digestion bands. They could be of mitochondrial origin because it is well established

S.

rbcL rps 19

psbA L

L

S2 M

U2

L 82 L 82

L 52

M L

I

_

_

*

S12 PM M

418S

Analysis of Plastid DNA Contained in Microspores Southern experiments using chloroplast gene probes were carried out to see whether the absence of transcripts could be S. M L

V

L *25S

At later stages, no transcripts from the two photosynthetic genes could be detected even after prolonged exposure time. The results obtained with the rrn 16S probe reveal the presence ofa small amount of an RNA fragment with a higher size than the mature 16S rRNA. The size of the fragment corresponds approximately to that of the precursor 16S rRNA (about 1.9 kb). This putative pre-16S RNA is present at all stages of microspore development but possibly in a decreasing amount. The presence of the same precursor 16S rRNA species was also detected in mature pollen in another independent experiment (not shown). Similar experiments were performed using mitochondrial probes prepared from cob, atp6, and atp9 genes (Fig. 2). The presence of abundant transcripts from these genes is revealed in S2 microspores. In contrast, no transcripts could be detected at a significant level in M pollen. In leaves, the atp9 and cob transcripts are clearly detected, whereas the atp6 transcripts were only detected after a longer exposure time of the autoradiograph.

S2 M L

Plant Physiol. Vol. 99, 1992

4.6

L

5

.6.8

.s64

4

L-. *

'p6

.... f::

:~~~~~~~~~~~~~~~~~~~~'

cob

atp9

2:

atp6

Figure 2. Hybridization of the mitochondrial cob, atp6, and atp9 probes to RNA isolated from S2 microspores, from pollen at dehiscence (stage M) and from leaves (L). Fifteen micrograms of RNA were size separated on an agarose gel containing hydroxymercury. Ethidium bromide fluorescence of RNA is shown on the left side . Autoradiographs were obtained after 8 h (atp6 and cob probes) or 1 h (atp9) of exposure. Transcripts indicated with a star are more clearly detected in leaf RNA after a longer time of exposure of the autoradiograph but are not detected in M pollen RNA.

B3amHll

BamHl BamHl F-'-oRl

Figure 3. Hybridization of organelle gene probes to restricted DNA fragments from leaves (L), microspores (S2), and mature pollen (M). Total DNA was restricted by BamHl or EcoRI. Sizes of restriction fragments are given in kb. Probes used are indicated at the bottom of each autoradiogram. The size of additional bands observed on overexposed autoradiograms is also indicated. The amount of DNA loaded on the gel is: L and S2, 0.3 Ag; M, 1 /tg.

ORGANELLE GENE EXPRESSION DURING MICROSPOROGENESIS

that some cpDNA have been transferred to the mitochondrial genome (15, 23-25). Analysis of Mitochondrial DNA Contained in Microspores An amplification of the mtDNA in microspores relative to leaves is observed by Southern-type hybridizations performed with mitochondrial probes (Fig. 3). Total DNA from leaves or from microspores at the stage of starch accumulation (S2) was digested with BamHI and hybridized with mitochondrial atp6 and cob probes. Ethidium bromide fluorescence confirmed that complete digestion was achieved (not shown). The two genes are much more abundant in S2 microspores than in leaves. Identical restriction fragments are revealed with atp6 and cob probes in leaf and microspore DNA. The 6.8kb atp6 and 13-kb cob fragments correspond to the copy of the gene located on the "master circle" molecule according to Fauron and Havlik (11). The other fragments likely correspond to copies of the gene located on submolecules of the mtDNA. Two fragments of 1.95 and 1.8 kb bearing the cob gene appear specifically in microspore DNA. They are not detected in leaf DNA even after a longer time of exposure.

They might arise from amplification of submolecules of the mitochondrial genome or from recombination events occurring during microsporogenesis. A second series of experiments were undertaken to estimate the relative amount of mitochondrial genes during development of microspore. The same samples of unrestricted DNA extracted from microspores at different stages of development were purified by electrophoresis and after blotting were hybridized with a nuclear probe (rbcS) and two different mitochondrial genes (cob and atp6). The results (Fig. 4) clearly show that mitochondrial genes are present in all the stages of development of the microspores, even in the M pollen grain where no expression of the three genes examined was detected (Fig. 2). The amount of mitochondrial DNA reached a maximum at a time of microspore development between stages V and S.

V

Si S2 M ....

L

.

cob Figure 4. Hybnidization of nuclear (rbcS) and mitochondrial (cob and atp6) gene probes to total DNA isolated from microspores at several stages of development or from leaves. DNA was purified by electrophoresis before blotting (see "DNA gel blot" experiments in "Materials and Methods"). The same samples of DNA were used for the three hybridization experiments. Ten micrograms of DNA were deposited in each slot. Blots were hybridized with random-primed labeled probes (rbcS, 1 t9 X 16 iCpM; cob, 5 x 1 m6 cpm; atp6, 15 x 1e6CpM). X-ray films were exposed for 48 h (rbcS and cob) or 96 h (atp6). Stages of microspore development are given at the top of the figure.

399

DISCUSSION

Our results indicate an increase of the steady-state level of mitochondrial transcripts in microspores at mid-term of development. The accumulation of mitochondrial transcripts might result from a gene dosage effect because it corresponds to the amplification of specific mitochondrial genes. We cannot exclude that, alternatively, activation of the transcription of the mitochondrial genes or stabilization of the mitochondrial transcripts occurs in mid-term microspores. Very little is known about mitochondrial gene expression during pollen development in comparison with nuclear gene expression (19). It has been found that the accumulation of cytosolic transcripts occurs during two separate phases of microspore development. First, the "early" genes become activated after meiosis, and second, the "late" genes become active after microspore mitosis. The mRNAs of the latter set of genes increase in abundance up to maturity (19). The observed pattern of mitochondrial gene expression during microsporogenesis appears to be distinct from either the early or late nuclear gene expression pattern. Mitochondrial transcript abundance is highest in the middle stages of pollen development but decreases significantly toward maturity. We can speculate that the increase of mitochondrial transcripts is related to an increase in mitochondrial proteins and possibly to a higher respiration rate. An electron microscopic observation of microspores during development has allowed Lee and Warmke (14) to detect an increase in the number of mitochondria as well as a decrease in size of individual mitochondria at postmeiotic stages. This observation has not been correlated with gene expression. Our results show an increase in mitochondrial gene copy number relative to the copy number of a nuclear gene and we suggest that this increase is related to the larger number of mitochondria observed (14). Mitochondrial division, DNA replication, and gene transcription might occur concomitantly during the

mid-phase of microspore development. Amplification of the mitochondrial genome might have important consequences in the context of CMS. Following a suggestion of Warmke and Lee (28), we speculate that genome amplification is at the origin of the abortion of mitochondria in anthers of the male sterile lines of maize. The products of chimeric genes occurring in cytoplasmic sterile lines (10, 22, 29) might result in the abortion of anther cells in overexpression of mitochondrial genes accompanying amplification of part of the mitochondrial genome, which does not take place in other tissues of the plant. The psbA and the rbcL genes have been used as representative plastid-encoded genes. Very low numbers of transcripts of the psbA gene have been detected in juvenile microspores only. These transcripts might be remnants of premeiotic transcription. Our results indicate that the two photosynthetic genes are not transcribed at an appreciable level during microsporogenesis, or that a rapid degradation of the RNA products occurs. This result might be extended to the expression of other plastid-encoded photosynthetic genes in microspores. A comparable situation occurs in chromoplasts that result from the redifferentiation of chloroplasts in tissues, such as those of the tomato fruit during ripening. In the latter, the transcript levels for photosynthetic genes decrease to low

MONEGER ET AL.

400

or nondetectable levels in parallel with the differentiation of chloroplasts into chromoplasts (12, 21). We assume that the rRNA species detected in microspores corresponds to the unprocessed form of 16S rRNA. As free RNA molecules appear to be degraded in plastids, any existing pre- 16S rRNA molecules are probably found in complexes with ribosomal proteins. The detection of pre- 1 6S rRNA might suggest that the synthesis of plastid ribosomes, and hence translation, in microspores is inhibited, or proceeds at an extremely low level. A similar situation might occur in tomato chromoplasts where transcripts of the plastid gene coding for the ribosomal protein S4 has been undetectable (12) and where the level of 16S rRNA decreases during chromoplast differentiation (20). Results concerning the plastid genome in microspores show that no rearrangement or deletion ofthe few chloroplast genes that have been tested occurs in microspores during their development or at maturity. Our results are in contrast with those obtained by Corriveau et al. (5). These authors could not detect plastid DNA in M pollen of A. majus, a species, like maize, known to transmit plastids maternally. We suggest that plastid DNA is present in the vegetative cells ofthe pollen only and not in sperm cells. The absence of plastid in sperm cells would explain the strict maternal inheritance previously observed (2, 4).

ACKNOWLEDGMENTS

We thank Prof. C.S. Levings III and Prof. C. Leaver for their authorization to use mitochondrial probes originated from their laboratory. We are grateful to F. Vedel and M. Lebrun for the gift of clones, and to H. Pesey for her assistance. LITERATURE CITED 1. Briat JF, Dron M, Loiseaux S, Mache R (1982) Structure and transcription of the spinach chloroplast operon rDNA leader region. Nucleic Acids Res 10: 6865-6878 2. Conde M-F, Pring D-R, Levings C-S III (1979) Maternal inheritance of organelle DNA's in Zea mays-Zea perennis reciprocal cross. J Hered 70: 2-4 3. Cook FS, Walden DB (1965) The male gametophyte of Zea mays L. II. In vitro germination. Can J Bot 43: 779-786 4. Corriveau JL, Coleman AW (1988) Rapid screening method to detect potential biparental inheritance of plastid DNA and results for over 200 angiosperm species. Am J Bot 75: 1433-1458 5. Corriveau JL, Goff LJ, Coleman AW (1990) Plastid DNA is not detectable in the male gametes and pollen tubes of an angiosperm (Antirrhinum majus) that is maternal for plastid inheritance. Curr Genet 17: 439-444 6. Dawson AJ, Jones VP, Leaver CJ (1984) The apocytochrome b gene in maize mitochondria does not contain introns and is preceded by a potential ribosome binding site. EMBO J 3: 2107-2113 7. Day A, Ellis THN (1984) Chloroplast DNA deletions associated with wheat plants regenerated from pollen: possible basis for maternal inheritance of chloroplasts. Cell 39: 359-368 8. Dewey RE, Levings CS III, Timothy DH (1985) Nucleotide sequence of ATPase subunit 6 gene of maize mitochondria. Plant Physiol 79: 914-919 9. Dewey RE, Schuster AM, Levings CS III, Timothy DH (1985) Nucleotide sequence of Fo-ATPase proteolipid (subunit 9) gene of maize mitochondria. Proc Natl Acad Sci USA 82: 1015-1019

Plant Physiol. Vol. 99, 1992

10. Dewey RE, Levings CS III, Timothy DH (1986) Novel recombinations in the maize mitochondrial genome produce a unique transcriptional unit in the Texas male-sterile cytoplasm. Cell 44: 439-449 11. Fauron CMR, Havlik M (1988) The BamHl, Xhol, SmaI restriction enzyme maps of the normal maize mitochondrial genome genotype B37. Nucleic Acids Res 16: 10395-10396 12. Kobayashi H, Ngernprasirtsiri J, Akazawa T (1990) Transcriptional regulation and DNA methylation in plastids during transitional conversion ofchloroplasts to chromoplasts. EMBO J 9: 307-313 13. Larrinua IM, Muskavitch KMT, Gubbins EJ, Bogorad L (1983) A detailed restriction endonuclease site map of the Zea mays plastid genome. Plant Mol Biol 2: 129-140 14. Lee SLJ, Warmke HE (1979) Organelle size and number in fertile and T-cytoplasmic male-sterile corn. Am J Bot 66: 141-148 15. Lonsdale DM, Hodge TP, Howe CJ, Stern DB (1983) Maize mitochondrial DNA contains a sequence homologous to the ribulose-1,5-bisphosphate carboxylase large subunit gene of chloroplast DNA. Cell 34: 1007-1014 16. Macherel D, Kobayashi H, Akazawa T (1985) Amyloplast nucleoids in sycamore cells and presence in amyloplast DNA of homologous sequences to chloroplast genes. Biochem Biophys Res Commun 133: 140-146 17. Mandaron P, Niogret MF, Mache R, Moneger F (1990) In vitro protein synthesis in isolated microspores of Zea mays at several stages of development. Theor Appl Genet 79: 134-138 18. Maniatis T, Fritsch EF, Sambrook J (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 19. Mascarenhas JP (1990) Gene activity during pollen development. Annu Rev Plant Physiol Plant Mol Biol 41: 317-338 20. Piechulla B, Chonoles Imlay KR, Gruissem W (1985) Plastid gene expression during fruit ripening in tomato. Plant Mol Biol 5: 373-384 21. Piechulla B, Pichersky E, Cashmore AR, Gruissem W (1986) Expression of nuclear and plastid genes for photosynthesisspecific genes proteins during tomato fruit development and ripening. Plant Mol Biol 7: 367-376 22. Rottmann WH, Brears T, Hodge TP, Lonsdale DM (1987) A mitochondrial gene is lost via homologous recombination during reversion of CMS T maize to fertility. EMBO J 6: 1541-1546 23. Sederoff RR, Ronald P, Bedinger P, Rivin C, Walbot V, Bland M, Levings CS III (1986) Maize mitochondrial plasmid SI sequences share homology with chloroplast gene psbA. Genetics 113: 469-472 24. Stern DB, Lonsdale DM (1982) Mitochondrial and chloroplast genomes of maize have a 12-kilobase DNA sequence in common. Nature 299: 698-702 25. Stern DB, Palmer JD (1984) Extensive and widespread homologies between mitochondrial DNA and chloroplast DNA in plants. Proc Natl Acad Sci USA 81: 1946-1950 26. Thomas F, Zeng GQ, Mache R, Briat JF (1988) Transcription study of the genes encoded in the region of the junction between the large single copy and the inverted repeat A of spinach chloroplast DNA. Plant Mol Biol 10: 447-457 27. Thomas F, Massenet 0, Dorne AM, Briat JF, Mache R (1988) Expression of the rp123, rpl2 and rpsl9 genes in spinach chloroplasts. Nucleic Acids Res 16: 2461-2472 28. Warmke HE, Lee SLJ (1978) Pollen abortion in T cytoplasmic male-sterile corn (Zea mays): a suggested mechanism. Science 200: 561-563 29. Young EG, Hanson MR (1987) A fused mitochondrial gene associated with cytoplasmic male sterility is developmentally regulated. Cell 50: 41-49 30. Zurawski G, Perrot B, Bottomley W, Whitfeld PR (1981) The structure of the gene for the large subunit of ribulose-1,5bisphosphate-carboxylase from spinach chloroplast DNA. Nucleic Acids Res 9: 3251-3269