Search

We report expression profiles of several genes of carbohydrate metabolism, cell wall invertase (CWI) in particular, to better understand sugar transport and its utilization in developing caryopses of grain sorghum [Sorghum bicolor (L.) Moench]. Gene expression analyses for CWI using RNA gel blot and real-time quantitative PCR approaches on developing caryopses, including the glumes (maternal tissue appended to the seeds), showed expression of SbIncw (ZmIncw2 ortholog) primarily in the basal sugar unloading zone of endosperm. The expression of ZmIncw1 ortholog was significantly less abundant and restricted to the glumes. The protein and enzyme activity data corroborated the temporal transcript expression profile that showed maximal CWI protein (INCW) expression preceding the starch-filling phase of endosperm development, i.e. 6-12d-after-pollination (DAP). Protein gel blot analysis using polyclonal maize INCW1 antibodies showed a single polypeptide of 72kDa. The highest level of enzyme activity was unique to the basal part of the endosperm, in particular the basal endosperm transfer cell (BETC) layer and the maternal pedicel region that were highly enriched for the INCW protein, as seen by immunolocalization. High hexose-to-sucrose ratio in 6-12 DAP seeds, and negligible starch deposition in glumes corroborated the CWI activity data. Additionally, we report transcription profiles of several other genes related to sugar-to-starch metabolism in developing sorghum endosperm. As in maize, the INCW-mediated apoplastic cleavage of sucrose in the BETC and pedicel during the early developmental stages of caryopses is essential for the normal development of filial tissues. The unique cell-specificity of the INCW protein to both proximal and distal ends of placental sac shown here for the first time is likely to greatly increase uptakes of both hexose sugars and water through turgor sensing into developing seed. This trait is unique to sorghum among cereals and may facilitate its survival in drought environment.

The restoration of male fertility in the sorghum IS1112 C (A3) male-sterile cytoplasm is through a two-gene gametophytic system involving complementary action of the restoring alleles Rf3 and Rf4. To develop markers suitable for mapping rf4, AFLP technology was applied to bulks of sterile and fertile individuals from a segregating BC(3)F(1) population. Three AFLP markers linked to rf4were identified and subsequently converted to STS/CAPS markers, two of which are co-dominant. Based on a population of 378 BC(1)F(1) individuals, two STS/CAPS markers, LW7 and LW8, mapped to within 5.31 and 3.18 cM, respectively, of rf4, while an STS marker, LW9, was positioned 0.79 cM on the flanking side of rf4. Markers LW8 and LW9 were used to screen sorghum BAC libraries to identify the genomic region encoding rf4. A series of BAC clones shown to represent a genomic region of linkage group E were identified by the rf4-linked markers. A contig of BAC clones flanking the LW9 marker represent seed clones on linkage group E, from which fine mapping of the rf4 locus and chromosome walking can be initiated.

A marked reduction of mitochondrial atp6 transcript-editing capability in sorghum anthers and pollen has been invoked as a factor in the loss of viability of male gametophytes in lines carrying the IS1112C male-sterile cytoplasm. We initiated a systematic examination of transcript editing of sorghum atp6 during microgametogenesis, from microspores through pollen, in two sets of male-fertile and near-isogenic, male-sterile lines. Transcript editing in cDNA populations from fertile lines indicated an average of about 90% editing among 15 sites examined. Parallel samples from sterile lines demonstrated a detectable reduction in editing efficiency, most evident in late pollen samples. However, editing efficiency in these cells averaged 80%, substantially higher than the 20% previously observed. Thus male-sterile lines retain substantial atp6-editing efficiency in an environment where the sterility and fertility restoration traits are stably expressed. Drastically reduced editing of atp6 transcripts in anthers and pollen of male-sterile sorghum is not demonstrable in a field environment; and thus it is not a factor in the loss of microgametophyte viability in this system.

RNA editing and cytoplasmic male sterility are two important phenomena associated with higher plant mitochondria. We recently have shown a potential function of RNA editing in CMS development. The frequency of atp6 RNA editing was specifically reduced in anthers of male-sterile Sorghum bicolor, which increased in frequency in partially restored progeny. Here we present data that show that the loss of RNA editing capability also occurs in a second nuclear background that allows the expression of male sterility. Loss of RNA editing thus appears to be associated with unique combinations of male-sterile cytoplasm and non-restoring nuclear backgrounds. In addition, the reduction of RNA editing affects both gametophytic and sporophytic anther cell-types but not other floral tissues. An analysis of F(2) plants exhibiting different levels of fertility indicates a co-segregation of fertility restoration and atp6 RNA editing. The atp6 transcript abundance is similar in seedlings and anthers of male-sterile, partially restored, and male-fertile lines and thus is not associated with loss of atp6 RNA editing in anthers. A model for RNA editing and male sterility based on the data available is presented. Functional correlations with other CMS systems are also discussed.

The sorghum line IS1112C carries a male sterility-inducing cytoplasm when introduced into nuclear backgrounds that do not include fertility restoration genes. An mtDNA chimeric configuration resulting from recombination/duplication with atp9 resulted in the formation of orf107, a chimeric open reading frame. Transcription of orf107 is driven by three promoters, and abundant whole-length transcripts are detected in male-sterile lines. Fertility restoration is exacted through a unique two-gene gametophytic system requiring complementary action of genes designated Rf3 and Rf4. In male-sterile lines carrying Rf3, or lines restored to fertility, an enhanced nucleolytic transcript processing activity is targeted within orf107, cleaving 75% of whole-length transcripts. Rf3 thus confers or regulates the nucleolytic processing activity. A correlation between the frequency of RNA editing at two sites in orf107 and transcript processing suggests that processing may be dependent on templates edited at these sites. In addition, editing of atp6 transcripts is specifically reduced in anthers/pollen of male-sterile lines. Partially restored F1s and segregating F2s exhibit atp6 editing frequencies consistent with the possibility that Rf4 may confer the restitution of normal editing frequency. Thus RNA editing may be involved in features of fertility restoration in this unusual system.

Defects in the human mitochondrial genetic system result in some diseases. These disorders are the result of rearrangements or point mutations in mitochondrial genes. In higher plants mutations and rearrangements in the mitochondrial DNA are believed to cause cytoplasmic male sterility (CMS), a mitochondrially inherited inability to produce viable pollen. In sorghum, formation of CMS is strongly correlated with anther-specific loss of mitochondrial atp6 RNA editing. Here we show that this loss of atp6 RNA editing mimics point mutations at codons that cause severe disorders in humans. We conclude that (i) loss of RNA editing in sorghum anthers probably causes CMS, (ii) similarities exist in the onset of mitochondrial dysfunction in plant and human tissues, and (iii) the evolutionary appearance of RNA editing provided a mechanism to compensate for otherwise lethal point mutations.

Defective nuclear-cytoplasmic interactions leading to aberrant microgametogenesis in sorghum carrying the IS1112C male-sterile cytoplasm occur very late in pollen maturation. Amelioration of this condition, the restoration of pollen viability, involves a novel two-gene gametophytic system, wherein genes designated Rf3 and Rf4 are required for viability of individual gametes. Rf3 is tightly linked to, or represents, a single gene that regulates a transcript processing activity that cleaves transcripts of orf107, a chimeric mitochondrial open reading frame specific to IS1112C. The mitochondrial gene urf209 is also subject to nucleus-specific enhanced transcript processing, 5' to the gene, conferred by a single dominant gene designated Mmt1. Examinations of transcript patterns in F2 and two backcross populations indicated cosegregation of the augmented orf107 and urf209 processing activities in IS1112C. Several sorghum lines that do not restore fertility or confer orf107 transcript processing do exhibit urf209 transcript processing, indicating that the activities are distinguishable. We conclude that the nuclear gene(s) conferring enhanced orf107 and urf209 processing activities are tightly linked in IS1112C. Alternatively, the similarity in apparent regulatory action of the genes may indicate allelic differences wherein the IS1112C Rf3 allele may differ from alleles of maintainer lines by the capability to regulate both orf107 and urf209 processing activities.

Nucleolytic processing of transcripts within mitochondrial orf107, associated with male sterility in sorghum, is regulated by the fertility restoration gene Rf3, conferring 75% cleavage of whole-length transcripts. Two transcript editing sites are 81% and 61% edited in rf3rf3 lines, while these sites are 41% and 10% edited in the remaining whole-length transcripts in an Rf3Rf3 line. RNA editing and processing efficiency in F1 progeny were similar to the Rf3Rf3 parent, and analyses of backcross progeny indicated that all rf3rf3 lines were characterized by high editing efficiency. We postulate that highly edited transcripts within the population are quickly processed in lines carrying Rf3, generating a residual population of poorly edited transcripts. Thus, action of Rf3 may have no direct affect on RNA editing, and may be dependent on a substrate of highly edited transcripts. These data indicate a potentially novel role of RNA editing in gene expression through an influence on the efficiency of transcript processing.

Transcriptional initiation and processing was examined for three sorghum mitochondrial DNA genes (atp6-1, atp6-2, urf209) and two open reading frames (orf265/130, orf107) to characterize sequences associated with initiation and other transcriptional strategies for this species. The 5' termini of ten transcripts were determined by primer extension, and mtRNA was capped with guanylyl transferase and annealed to anti-sense riboprobes to identify transcriptional initiation regions. Eight transcript termini were suitable substrates for guanylyl transferase, indicating the presence of one (atp6-1, atp6-2, urf209), two (orf265/130), or three (orf107) promoters for the five examples. The majority of the putative promoters were associated with single primer-extension termini, while two examples exhibited two transcript-initiation sites within the promoter. Four examples were characterized by initiated transcripts without subsequent processing, indicating that processing is not obligatory. Each of the putative promoter regions included significant A/T-rich 5' regions, consistent with previous examples, but four exceptions to a consensus core YRTA sequence were identified. The anomalies (AATA, CTTA) suggest plasticity in the primary structure of the core region of higher-plant mitochondrial DNA promoters.

Sorghum mitochondrial atp9 is polymorphic among male-sterile cytoplasms, but each cytoplasm is characterized by a major 650 nt transcript, regardless of fertility status. The gene is positioned 323 bp 3' to trnfM. Primer extension revealed multiple atp9 5' transcript termini, distributed from +1 to +28 3' to trnfM; the termini could be labeled with polynucleotide kinase, suggesting that they result from the maturation of trnfM. Guanylyltransferase experiments, however, showed that four of the termini were capable. The juxtaposition of a putative promoter 3' to trnfM results in a unique atp9 transcript population consisting of primary and processed transcripts.

A chimeric mitochondrial DNA (mtDNA) configuration of the cytoplasmic male-sterile (cms) sorghum line IS1112C includes a 321 bp open reading frame designated orf107, encoding a predicted polypeptide product of 11.85 kDa. The open reading frame, similar to several other genes associated with cms, consists of amino-terminal sequences derived from an obligate gene. Unlike other examples to date, however, the carboxy-terminal sequences are highly similar to the carboxy terminus of an open reading frame implicated in cms of rice, orf79. The amino-terminal 31 residues of orf107 are 84% similar to atp9, and the carboxy-terminal 49 residues are 57% identical and 80% similar to the carboxy terminus of orf79. Transcripts of orf107 are edited, with four C-to-U changes that alter amino acids. Sorghum lines partially or fully restored to fertility exhibit a high-efficiency internal-orf107 transcript processing activity, precluding abundant whole-length transcripts, while male-sterile lines exhibit only a trace of the activity. Previous data on the abundance of a 12kDa in organello-synthesized polypeptide in male-sterile versus male-fertile lines are correlated with differential orf107 transcript processing activity of these lines. Examinations of backcross and F2 lines suggest a gametophytic mode of restoration, and indicate that enhanced transcript processing activity is necessary, but not sufficient, to restore full fertility. These novel observations indicate that mitochondrial open reading frames associated with cms in different species can include highly similar motifs, and that fertility restoration could involve a mechanism by which synthesis of a cms-associated gene product may be precluded through internal transcript cleavage.

We describe fundamental characteristics of sorghum mitochondrial orf25, urf209, and a related chimeric configuration, orf265/130, which is restricted to the IS1112C source of cytoplasmic male sterility in sorghum. Transcripts of urf209 are edited at ten nucleotides, resulting in nine amino-acid changes predicted from genomic sequences. The cDNA-predicted polypeptide product is 23.6 kDa, while Western blot analyses identify a product of 20k Da. Transcription of urf209 is characterized by one or two transcripts, dependent on nuclear background, but this difference is not related to male fertility status. The orf265/130 chimeric region includes 288 bp 95% identical to sequences 5' to maize T-cytoplasm T-urf13 and atp6, which includes a common transcription initiation site, and terminates with a recombinational event involving urf209. The urf209 similarity extends 189 bp, followed by sequences duplicated 5' to sorghum atp6-2. Sequences immediately 3' to the atp6-2 similarity include a second in-frame start codon, defining orf130. Structural features 5' to orf130 are shared with motifs found 5' to several translated mitochondrial open reading frames. The orf265/orf130 configuration is uniquely transcribed, and transcripts of orf130 exhibit one silent RNA editing event. Transcription in somatic cells is not altered by male fertility status.

We have observed numerous examples of silent or rare non-silent editing sites in the amino-extension and part of the conserved core of mitochondrial atp6 transcripts of Sorghum. In this region of the 1.4-kb atp6-2 mRNA (position 300 to 550) two editing sites, which alter the amino-acid sequence and occur in all cDNAs analysed, were already known, while nine others were found which are silent or occur in a few mRNAs only. Many aspects of RNA editing in the mitochondria of higher plants are still unknown. This includes the influence of genomic background or silent RNA editing. We were interested in the influence of nuclear and mitochondrial backgrounds on RNA editing. Previous preliminary results indicated the possibility of line-specific editing at silent sites. However, a more comprehensive approach gave no consistent evidence for such editing. These results are discussed with respect to their potential impact on the evolution of mitochondrial genes.

The interaction of nuclear and mitochondrial genomes in the alteration of maize (Zea mays L.) mitochondrial orf221 transcript patterns was examined. Northern analyses involving specific maize nuclear genotypes associated with N, C or S cytoplasms revealed considerable orf221 transcript heterogeneity. F1 progenies were developed from maize inbred-cytoplasm combinations that differed for orf221 transcript patterns. Northern analyses revealed that the presence or level of abundance of certain orf221 transcripts was dependent on nuclear genotype. The maize inbred B37(C) exhibits orf221 transcripts of 3500, 3200, 2800, and 1300 nt whereas the F1 of B37(C) x Ky21(N) does not exhibit a 2800-nt transcript but does give transcripts of 2100 and 1250 nt in addition to 3500-, 3200- and 1300-nt transcripts. Northern analyses also suggested that the size or the presence of certain orf221 transcripts was related to the mitochondrial genome configuration. Maize inbred A619 exhibits a 2300-nt orf221 transcript when associated with N cytoplasm and a 2100-nt orf221 transcript when associated with C and S cytoplasms. As a result of deletion of the gene T-urf13, the A188(T7) mitochondrial mutant exhibits only a 3100-nt orf221 transcript and not the very complex T-urf13/orf221 transcript pattern associated with A188(T). The genetic stock A188(T7) x W64A(N)(2) gives a highly abundant 2100-nt orf221 transcript not detected in A188(T7). Deletion of T-urf13 has enabled the nuclear genotype of W64A(N) to alter orf221 transcript patterns in a manner not detected in T cytoplasm. This observation suggests that alteration of the mitochondrial genomic configuration adjacent to orf221 results in a different response to nuclear gene products from that observed when or2f21 is present in the T mitochondrial genome configuration.

Sorghum mitochondrial atp6 occurs as one copy in the line Tx398 and as two copies in IS1112C. In IS1112C a repeated sequence diverged within the atp6 open reading frames. The two open reading frames (1137 bp, atp6-1; 1002 bp, atp6-2) share an identical conserved region of 756 bp but are flanked 5' by divergent extensions of 246 (atp6-1) or 381 bp (atp6-2). Tx398 carried only atp6-2. The breakpoint of the repeated sequence of the conserved core region corresponds to the amino acid sequence Ser-Pro-Leu-Asp, which is the amino terminus of the proteolytically processed yeast ATP6. The 5' extensions of atp6-1 and atp6-2 were similar to those of rice and maize, respectively. Each open reading is transcribed, however nuclear background influenced transcriptional patterns of atp6-2 in IS1112C.

The mitochondrial gene T-urf13 in T cytoplasm maize is associated with sensitivity to disease toxins and with cytoplasmic male sterility. T-urf13 is co-transcribed with an open reading frame designated orf221. We have detected alterations in the transcription of the T-urf13/orf221 region that are affected by nuclear genotype. There are multiple mRNA transcripts generated from the T-urf13/orf221 region, one of which is a processed 1538-nucleotide (nt) transcript. This 1538-nt transcript is present in Wf9 (T), but was not found in mitochondrial RNAs (mtRNAs) from maize inbreds B14A (T) and 33-16 (T). For B14A (T) a 1500-nt transcript was detected and for 33-16 (T) a 1400-nt transcript was detected. In F1 progeny of the cross of Wf9 (T) x 33-16 (N), only the 1400-nt transcript was present. Genetic analyses revealed this processing event is nuclear controlled with dominant gene action and is independent of nuclear restorer gene Rf1-associated processing events. T-urf13/orf221 transcriptional patterns were shown to vary in both sterile and fertile states. Segregation analysis of a 1100-nt orf221-specific transcript indicated that the genetic basis of nuclear control for the presence of this transcript was relatively simple. Analysis of the A188 (T4) tissue culture mutant, which has reverted to male fertility but displays the same T-urf13/orf221 transcript pattern as A188 (T), indicated no DNA sequence differences between T4-orf221 and T-orf221. Presence of the nuclear gene Rf2 was not necessary for expression of the T4 cytoplasm-associated malefertile phenotype.

Three sequence blocks of 10-12 bp are conserved in sequence and order 5' to putative start codons of several higher-plant mitochondrial genes. At least 25 examples were found, primarily associated with coxII, atp6, and orf25, in monocotyledons and dicotyledons. The proximal block can be 9 bp from start codons, and the three blocks generally occur within 100 bp 5' of start codons. In three examples 5' termini of the blocks represent recombination breakpoints, resulting in conservation of the blocks in resultant configurations. The two proximal blocks can form a secondary structure motif. The occurrence of the blocks near start codons, and conserved sequence and order, is consistent with a possible role in translation initiation or regulation.

Genomic and cDNA sequences of the ATP synthase complex subunit 9 (atp9) genes from two sorghum lines were determined. Sequences of cDNAs revealed eight C to U transcript editing events resulting in six amino acid changes and a new stop codon which eliminated 12 carboxy-terminal residues, compared to the genomic sequence. Sorghum atp9 has a unique five-residue amino-extension relative to other higher plants. The resulting predicted 79-residue gene product has a molecular weight of 8.179 kDa. The predicted phe-val-phe carboxy-terminus is identical to that from cDNA sequences of wheat, Oenothera, and petunia. Partial editing of transcripts was detected in each sorghum line.

Sequencing of sorghum mitochondrial atp6 cDNA clones revealed 19 C-to-U transcript editing events within a 756 bp-conserved core gene; three were silent and 16 resulted in 15 amino acid changes. Only one edit, which was silent, was found in the 381 bp amino-extension to the core gene. Eleven of the 15 changed amino acids were identical with or else represented conservative changes compared to yeast atp6. Editing of a CAA codon to TAA truncates the carboxy-terminus to a position identical to that of yeast. The frequency of editing at sites which change amino acids was very high in contrast to partial editing at silent, third base, sites.

We asked whether the mitochondrial T-urf13 gene, associated with the male sterility phenotype of T cytoplasm in maize, can be expressed in Saccharomyces cerevisiae and whether this expression can mimic the effects observed in maize. We introduced the universal code equivalent of the T-urf13 gene into the S. cerevisiae nucleus by transformation and directed its translation product into mitochondria by means of a fusion with the targeting presequence from Neurospora crassa AT-Pase subunit 9. We show that expression of the universal code equivalent of the T-urf13 gene in the yeast nucleus does indeed mimic its effects in maize: respiratory growth of yeast is inhibited, respiration-deficient cytoplasmic mutants accumulate and NADH oxidation of isolated mitochondria is uncoupled. All these effects are observed only if the mitochondrial targeting peptide and methomyl or HmT toxin are present.

Plants resistant to Helminthosporium maydis race T were obtained following selection for H. maydis pathotoxin resistance in tissue cultures of susceptible, Texas male-sterile (T) cytoplasm maize. The selected lines transmitted H. maydis resistance to their sexual progeny as an extranuclear trait. Of 167 resistant, regenerated plants, 97 were male fertile and 70 were classified male sterile for reasons that included abnormal plant, tassel, anther or pollen development. No progeny were obtained from these male-sterile, resistant plants. Male fertility and resistance to the Phyllosticta maydis pathotoxin that specifically affects T cytoplasm maize were co-transmitted with H. maydis resistance to progeny of male-fertile, resistant plants. These three traits previously were associated only with the normal (N) male-fertile cytoplasm condition in maize. Three generations of progeny testing provided no indication that the cytoplasmic association of male sterility and toxin susceptibility had been broken by this selection and regeneration procedure. Restriction endonuclease analysis of mitochondrial DNA (mtDNA) revealed that three selected, resistant lines had distinct mtDNA organization that distinguished them from each other, from T and from N cytoplasm maize. Restriction patterns of the selected resistant lines were similar to those from T cytoplasm mtDNA; these patterns had not been observed in any previous analyses of various sources of T cytoplasm. The mtDNA analyses indicated that the male-fertile, toxin-resistant lines did not originate from selection of N mitochondrial genomes coexisting previously with T genomes in the T cytoplasm line used for selection.

Mitochondrial DNA (mtDNA) from cell suspension cultures of two intergeneric somatic hybrids of Pennisetum americanum (pearl millet) + Saccharum officinarum (sugarcane) was examined by restriction endonuclease digestion and hybridization with sorghum mtDNA cosmids. The mtDNA of one somatic hybrid was indistinguishable from that of pearl millet, while the second exhibited a combination of parental mtDNAs, suggesting mitochondrial fusion. Several novel, possibly recombinant, mtDNA restriction fragments were detected in this hybrid, which may have resulted from intergenmic recombination.

Agarose gel electrophoresis of mitochondrial DNA (mtDNA) from the IS1112C male-sterile cytoplasm of Sorghum bicolor (S. bicolor) revealed plasmid-like DNAs additional to the linear N1 and N2 molecules. Mitochondrial plasmids were separated from the principal mitochondrial genome and used in the construction of molecular clones. Clones with EcoRI inserts of 1.7 and 2.3 kb were recovered. Hybridization of these clones to Southern blots of unrestricted and EcoRI-digested IS1112C mitochondrial plasmids indicated the cloned inserts were complete or nearly-complete copies of minicircular DNA molecules. These clones were used to probe Southern blots of mitochondrial genomes from six cytoplasmic male-sterile (CMS) and five male-fertile (MF) lines of S. bicolor, as well as Southern blots of IS1112C chloroplast, kafir chloroplast, IS1112C nuclear, and kafir nuclear genomes. The 2.3 and 1.7 kb plasmids had a very limited distribution among the sorghum entries we examined. We found no evidence for integrated copies of these sequences in any of the mitochondrial, chloroplast, or nuclear genomes probed in this study. However, the 2.3 kb sorghum minicircle did hybridize to the 1.94 kb minicircle from maize mitochondria. Hybridization of the 1.7 and 2.3 kb clones to IS111L2C mitochondrial RNA reveal a transcript of 1.1 kb from the 1.7 kb minicircle and transcripts of 1.9 and 1.4 kb from the 2.3 kb molecule.

Cytochromes a + a(3), b, and c of mitochondria prepared from fertile and cytoplasmic male-sterile single- and double-cross maize (Zea mays L.) lines were examined by difference spectra at 25 C. The sterile lines contained the full complement of cytochromes, and absorption maxima were identical for each fertile-sterile combination. A small, but significant, increase of b cytochromes was detected in each sterile line, whereas the quantities of cytochromes a + a(3) and c were essentially unchanged. The yield of mitochondrial protein per gram of tissue was higher for the single-than for the double-cross, whereas the sterile cytoplasm did not affect ultimate mitochondrial yield.

Tissue culture-derived mutants of male-sterile and disease toxin-sensitive Texas (T)-cytoplasm maize that exhibit male fertility and toxin insensitivity carry numerous alterations in mitochondrial DNA. In these mutants, a 6.7-kilobase Xho I fragment characteristic of parental T cytoplasm has been rearranged. In the mutant T-4, the parental 6.7-kilobase Xho I fragment contains a guanine to adenine transition adjacent to a 5-base-pair insertion not found in T cytoplasm. The insertion, internal to a 345-base-pair open reading frame (T ORF13), generates a frameshift, resulting in a premature stop codon that terminates the open reading frame at base pair 222. In other mutants, the 345-base-pair ORF is part of a 3-kilobase deletion, which extends into a 5-kilobase repeat characteristic of mtDNA from T but not N male-fertile cytoplasm. Clones specific to T ORF13 hybridize to eight transcripts in T and T-4, yet only hybridize to three transcripts in T-7, a deletion mutant. Transcription of the T ORF13 region appears not to be altered in T-4, but the frameshift mutation in the T ORF13 reading frame indicates that a biologically inactive gene product could be associated with the mutational events. The results suggest that cytoplasmic male sterility and disease toxin sensitivity may be associated with presence of T ORF13 in T-cytoplasm maize.

Mitochondrial DNA from three somatic hybrid cell lines of Pennisetum americanum + Panicum maximum was compared with mitochondrial DNA of the parents. Gel electrophoresis of BamHI-restricted mitochondrial DNA indicated that extensive rearrangements had occurred in each of the three hybrid lines. The hybrid restriction patterns showed a combination of some bands from each parent plus novel fragments not present in either parent. Additional evidence for rearrangements was obtained by hybridization of eight DNA probes, carrying sequences of known coding regions, to Southern blots. Each of the somatic hybrids exhibited a partial combination of the parental mitochondrial genomes. These data suggest recombination or amplification of the mitochondrial genomes in the somatic hybrids.

Two large double-stranded RNA molecules, 15 and 16 kilobases, were detected in cytoplasmic male sterile (CMS) Phaseolus vulgaris by agarose gel electrophoresis. A number of smaller RNA molecules were observed in 'Sprite', a maintainer line, and recurrent backcrossing of CMS P. vulgarisx'Sprite' resulted in a combined electrophoretic pattern of the two large and numerous small RNA molecules. The large RNA molecules were seed and pollen-transmissible, but were not transmitted by grafting. The RNAs were present in revertant and restored lines derived from CMS-Sprite and therefore were not associated with the cytoplasmic male sterile trait.

The linear N1 and N2 plasmid-like DNAs were recovered from mitochondria of the IS1112C line of cytoplasmic male-sterile (CMS) Sorghum bicolor (S. bicolor). Molecular clones containing internal sequences of these plasmids were constructed. These clones were used to probe Southern blots of mitochondrial genomes from six CMS and five male-fertile (MF) lines of S. bicolor, as well as Southern blots of IS1112C chloroplast, IS1112C nuclear and kafir nuclear genomes. We found no evidence for integrated copies of N1 or N2 in any of the mitochondrial, chloroplast or nuclear genomes probed in this study. Our clones did detect an N1-homologous transcript of 3.1 kb and N2-homologous transcripts of 3.9 and 1.4 kb in IS1112C mitochondrial RNA prepared from lines with and without nuclear, fertility-restoring genes.N1 and N2 DNAs were degraded by exonuclease III but were resistant to lambda exonuclease, presumably due to the presence of 5' terminal proteins. We detected multimeric forms of N1 and N2 in Southern blots of unrestricted, IS1112C mitochondrial DNA (mtDNA). These forms apparently also had associated protein molecules.

Genomic variation in S1 and S2 homologous sequences, defined as the S regions, were examined in mitochondrial DNAs of 12 normal cytoplasm maize lines collected in the United States. Three genomic variants were detected among the 12 cytoplasms, eight of which were identical to the Wf9 model structure. Hybridization data with S1 and S2 DNAs and with two cosmids spanning these regions were consistent with the concept that S1 and S2 sequences are found in each normal cytoplasm. Three variations of the S1 region were established; the Wf9 structure, a second group consisting of F6, A188, and W182BN, and a third, Black Mexican. Genome structure was conserved through the S2 region in all lines examined. None of the cytoplasms included complete copies of S1; the 1400 bp repeat characteristic of S1 and S2 was absent in the S1 region of all lines. A 2.1 kb linear DNA was observed instead of a 2.3 kb DNA in F6, A188, and W182BN. Integrated copies of S1 and S2 sequences may be a constituitive characteristic of normal, male-fertile maize cytoplasms.

Maize mitochondrial and chloroplast DNA's were prepared from normal (fertile) lines or single crosses and from members of the T, C, and S groups of male-sterile cytoplasms. Restriction endonucleases HindIII, BamI, EcoRI, and SalI were used to restrict the DNA, and the resultant fragments were electrophoresed in agarose gels. The results show that the N (fertile), T, C, and S cytoplasms each contained distinct mitochondrial DNA (mtDNA). These distinctive patterns were unaffected by nuclear genotype. No evidence of paternal inheritance of mtDNA was observed. Chloroplast DNA (ctDNA) from the N, C, and T cytoplasms was indistinguishable by HindIII, SalI, or EcoRI endonuclease digestion. The S cytoplasm ctDNA, however, was slightly different from that of other cytoplasms, as indicated by a slight displacement of one band in HindIII digests. The molecular weight of maize ctDNA was estimated to be as high as 88 x 10(6). Estimates of the minimum molecular weight of maize mtDNA ranged from 116-131 x 10(6), but the patterns were to complex for an unambiguous determination. Based on HindIII data, a comparison of the molecular weight of mtDNA bands common to the N, T. C, and S cytoplasms suggests that C cytoplasm most closely resembles N cytoplasm. The T and S sources are more divergent from the C and N cytoplasms. These results indicate a possible gradation of relatedness among male-sterile cytoplasms. The marked variation in mtDNA, with apparently less variation in ctDNA, represents circumstantial, but compelling, evidence that mtDNA may be involved in the male sterility and disease susceptibility traits in maize.

Differences in fertility restoration and mitochondrial nucleic acids permitted division of 25 accessions of S-type male sterile cytoplasm (cms-S) of maize into five subgroups: B/D, CA, LBN, ME, and S(USDA). S cytoplasm itself (USDA cytoplasm) was surprisingly not representative of cms-S, since only two other accessions, TC and I, matched its mitochondrial DNA pattern. CA was the predominant subgroup, containing 18 of the 25 accessions. The B/D and ME subgroups were the most fertile and LBN the most sterile. The exceptional sterility of LBN cytoplasm makes it the most promising of the 25 cms-S accessions for the production of hybrid seed. The most efficient means of quantifying the fertility of the subgroups was analysis of pollen morphology in plants having cms-S cytoplasm and simultaneously being heterozygous for nuclear restorer-of-fertility (Rf) genes. This method took advantage of the gametophytic nature of cms-S restoration. The inbred NY821LERf was found to contain at least two restorer genes for cms-S. Fertility differences were correlated with mitochondrial nucleic acid variation in the LBN, ME, and S (USDA) subgroups.

Mitochondria were prepared from etiolated maize shoots (Zea mays L. var. McNair 508) by homogenization followed by differential centrifugation and equilibrium banding in discontinuous sucrose or Renografin-sucrose gradients. Mitochondria prepared by sucrose banding showed better physiological integrity than those prepared by renografin-sucrose banding, although both procedures yielded mitochondria that showed respiratory control and coupling of oxidation to phosphorylation of ADP. Mitochondria prepared by Renografin-sucrose banding were free of dectectable cytoplasmic ribosomal RNA, while sucrose banding resulted in a low level of contamination. Ribosomes isolated from mitochondria sedimented at about 78S, with subunits sedimenting at 60 and 44S. Using Escherichia coli ribosomal RNA as internal standards, the molecular weights of mitochondrial ribosomal RNAs were found to be 0.74 to 0.75 and 1.26 x 10(6) daltons by polyacrylamide gel electrophoresis, before or after denaturation in formaldehyde. Cytoplasmic ribosomal RNA molecular weights were 0.70 and 1.26 x 16(6) before denaturation, and 0.68 and 1.5 x 10(6) after denaturation, suggesting an unusual reaction of the heavy ribosomal RNA to formaldehyde.

Spontaneous reversion to fertility in S male-sterile cytoplasm of maize is correlated with the disappearance of the mitochondrial plasmid-like DNA's, S-1 and S-2, and changes in the mitochondrial chromosomal DNA. Hybridization data indicate that one of the plasmid-like DNA's, S-2, is prominently involved in the mitochondrial DNA rearrangements.

The chloroplast genome of the IS1112C cytoplasm of sorghum was mapped by the construction of a Bam-HI library in pUC8, and hybridization with BamHI, SalI, and PstI digests of chloroplast DNA (ctDNA) of sorghum and maize. The molecules are extensively colinear, with only one of 13 SalI fragments differing slightly from maize. Seven of 70 restriction sites differed in the two species. A total molecular size of ca. 138 kb was estimated for sorghum. The inverted repeat was not conserved between sorghum and maize, as revealed by a slightly larger BamHI 16S rDNA fragment in sorghum. Homology of a sequence adjacent to the γbcl gene and one end of the inverted repeat was detected. These homologies were also observed in maize, and suggest that the ctDNA genomes of sorghum and maize share small reiterations of sequences of the inverted repeat.

Mitochondrial DNAs were prepared from maize lines with normal cytoplasm and with the T, C, S, and EP sources of male-sterile cytoplasms. Agarose gel electrophoresis of these preparations revealed a main high-molecular-weight DNA band. In addition, the S cytoplasm was characterized by the presence of two faster migrating DNAs of molecular weight 3.42 to 3.48 x 10(6) and 4.01 to 4.10 x 10(6). Electron microscopy showed these unique DNAs to be of different length, but their molecular configuration was not clearly established. It is possible that these unique DNAs represent physical evidence of an episomal system previously postulated to function in the S male-sterile cytoplasm.

Chloroplast and mitochondrial DNAs from six races of annual teosinte (Guatemala, Huehuetenango, Balsas, Central Plateau, Chalco, and Nobogame), perennial teosinte, and maize were compared and grouped by restriction endonuclease fragment analyses. Three groups of chloroplast DNAs were detected: (i) perennial teosinte and Guatemala; (ii) Balsas and Huehuetenango; and (iii) all other teosintes. Four groups of mitochondrial DNAs were separated: (i) perennial teosinte; (ii) Guatemala; (iii) Nobogame; and (iv) all other teosintes. Separation of the teosinte and maize organelle DNAs into five groups (Guatemala; perennial teosinte; Balsas and Huehuetenango; Central Plateau and Chalco; Nobogame and maize) approximated the biosystematic relationships of the taxa. It was suggested that the evolutions of the chloroplast and mitochondrial DNAs may be independent of each other, that variation of organelle DNA within a species complex of an organism may be the common condition, and that the DNAs of the organelle and nuclear systems evolve in reasonable harmony.

Mitochondrial DNA from normal and cytoplasmic male-sterile maize was digested with restriction endonucleases RI from Escherichia coli or dIII from Hemophilus influenzae. Electrophoresis of resulting fragments revealed distinctions between the two cytoplasmic types. These distinctions suggest that factors responsible or cytoplasmic male sterility are located in the mitochondrial DNA, and that the mitochondrial genome is not inherited paternally.

Mitochondrial DNA (mtDNA) of the Black Mexican sweet line of maize isolated from tissue cultured cell suspension cultures and young seedlings was examined. Restriction fragments generated by two endonucleases were comparatively analyzed by visualization of ethidium bromide stained gels as well as by membrane hybridization with nick-translated DNA probes of plasmid-like S1 and S2 DNA. Although no major molecular alterations were seen in tissue cultured cells, the samples were clearly not identical. The variation was mainly in the stoichiometry of several restriction fragments. Hybridization analyses with S1 and S2 probes show no evidence of molecular rearrangement in this part of the genome in tissue cultured cells. Minor variations in restriction patterns could reflect alterations in frequency of circular mtDNA molecules, perhaps related to nuclear alterations during the extended period of culture.

Five accessions of members of the C group of male sterile maize cytoplasms (BB, C, ES, PR, and RB) in two nuclear backgrounds (A619 and A632) were examined to elucidate the nature of mitochondrial genome diversity within a related group of cytoplasms. Cosmid and plasmid clones carrying single copy and recombinationally active sequences from N and S cytoplasms of maize were used as probes. Although restriction patterns are quite similar, each of the five could be discriminated by evidence of sequence duplication and recombination, deletion of recombinationally active sequences of N, normal cytoplasm, population of mini-circular DNAs, and by restriction patterns. Each member of the group carried a 1,913 bp minicircular mtDNA, while all entries but RB carried a 1,445 bp minicircular mtDNA. Members of the C group clearly are not molecularly identical; evolution of the group included principal genome reorganization involving sequence duplication/deletion events, apparently independent of the cms trait.

The mitochondrial genome of Black Mexican Sweet maize consists of the principal genome, a 2.3 kb minilinear DNA, a 1,913 bp (1.9 kb) and a 1,445 bp (1.4 kb) minicircular DNA. The three extrachromosomal DNAs exhibit characteristics of autonomous replication in cell suspension culture. The complete sequence of the 1.4 kb minicircle was determined. It has 61 bp of near perfect sequence homology to the 1.9 kb minicircle. Both minicircular DNAs are transcriptionally active; the longest open reading frame of the 1.4 kb minicircle was 231 bp. A putative origin of replication was identified as a high A + T sequence. These minicircles were present in some but not all of 20 maize lines surveyed. None of the lines examined carried the 1.4 kb minicircle without the 1.9 kb minicircle. Nuclear DNA of one line of the seven examined carried homology to both DNAs.

Polyspecific antibody to a 17 amino acid synthetic peptide from the maize T-cytoplasm urf13-T mitochondrial open reading frame immunoprecipitated a 13 kD polypeptide from (35)S-methionine incorporations of T cytoplasm maize. Male-fertile, toxin-insensitive mutants in which the urf13-T sequence is deleted do not synthesize the 13 kD polypeptide. A mutant designated T-4, which carries a 5 bp insertion and a premature stop codon, synthesizes a truncated polypeptide, corresponding to an open reading frame of 8.3 kD. Thus the 13 kD polypeptide is trunctated or absent in mutants expressing male fertility and toxin insensitivity in T-cytoplasm maize.

The mitochondrial genome of Black Mexican Sweet (BMS) maize includes the principal genome and two transcriptionally active minicircular DNAs of 1913 and 1445 bp. A cell suspension of this line was used to study the biology and replication of the minicircular DNAs. Synthesis of the DNAs was measured by (32)P incorporation; all mitochondrial DNAs (mtDNAs) were synthesized rapidly during logarithmic growth phases, whereas no synthesis could be detected in stationary phase. When stationary phase cultures were placed in fresh medium and incorporation was measured over time, the 1.9-kb minicircle renewed (32)P incorporation prior to incorporation into the 1.4-kb minicircle, the principal mitochondrial genome, or the nuclear genome. Interestingly, plastid DNA renewed incorporation at the same time as the 1.9-kb minicircle. The early replication of the 1.9-kb minicircle relative to the other DNAs increased the copy number of this DNA relative to the other mitochondrial DNAs. The copy number of the minicircular DNAs also varied between leaf cells and the cell suspension. This indicates that components of the mitochondrial genome exhibit differential replication. The ability to follow replication of individual mitochondrial components makes this system valuable for studies of DNA replication.

The mitochondrial genome of the S-type male-sterile cytoplasm of maize appears to be a collection of linear DNA molecules that are maintained by two linear episomal DNA species, S1 (6397 bp) and S2 (5453 bp), recombining with larger circular molecules. The M825 inbred nuclear genotype promotes a high frequency of spontaneous fertile mutants. Where the mutation is maternally inherited, analysis of the mitochondrial DNA reveals that the S1 and S2 episomes and all associated linear molecules have been lost. Integrated copies of the S1 and S2 episomes are retained, though part of one of the two terminal inverted repeats of the integrated S2 sequence has been deleted in all revertants studied. This deletion appears to correlate with the absence of the S2-URF-1 transcript.

Extensive recombination events characterize higher-plant mitochondrial DNAs. Numerous recombination events resulted in the appearance of an unusual mitochondrial open reading frame, urf13-T, which encodes a 13 kDa polypeptide in the male-sterile T cytoplasm of maize. Maize lines with T cytoplasm are unusually susceptible to two fungal pathogens which produce host-selective toxins. Mutants derived from tissue culture expressing male fertility and toxin-insensitivity are characterized by truncation or deletion of urf13-T. These events result from a frameshift associated with a tandem 5 base pair repeat, placing a premature stop codon in frame, or from a recombination event, apparently limited to tissue culture, resulting in the deletion of urf13-T. Neither class of mutants produces the 13 kDa gene product. Repeated sequences that participate in recombination in sorghum appear to be randomly distributed among male-fertile or male-sterile cytoplasms. Processes involved in the evolution of mitochondrial DNAs in higher plants therefore include the generation and deletion of configurations through recombination.

Endonuclease restriction fragment patterns of Pennisetum americanum L. mitochondrial DNAs (mtDNAs) from a cytoplasmic male-sterile (CMS-A1), fertile revertants and a normal fertile cytoplasm were variable, while chloroplast DNA from those lines lacked variation. Comparisons between mtDNAs of CMS-A1 (parental) and fertile revertant lines revealed the presence of a unique 4.7 kbp PstI fragment in the sterile line that was not detected in any of the revertant lines. A 9.7 kbp PstI fragment was found in all of the revertants, but not in the CMS-A1. Neither of those fragments was found in the normal cytoplasm mtDNA. Hybridization studies revealed two sets of multiple homologies: 1) the 4.7 kbp fragment had homology with a 10.9 kbp and a 13.6 kbp fragment; and 2) the 9.7 kbp fragment was homologous with the 13.6 kbp fragment. The presence of those two repeated mitochondrial sequences on the altered fragments suggests that they may be involved in the recombinational associated events with reversion from CMS to fertility in P. americanum.

Restoration of pollen fertility to cytoplasmic male sterile (CMS) Phaseolus vulgaris by a nuclear restorer gene provides a system for studying nuclear-cytoplasmic interactions. Introduction of a nuclear restorer gene to this CMS line of P. vulgaris (CMS-Sprite) results in a mitochondrial genome rearrangement similar to that observed upon spontaneous cytoplasmic reversion to fertility. Three spontaneous heritable cytoplasmic revertants were derived from CMS-Sprite. Five fully fertile restored lines were also produced by using restorer line R-351 (BC(3)F(3) populations). Comparison of the mitochondrial DNA restriction patterns of CMS-Sprite, the three fertile revertants, and the five restored lines revealed loss of a 6.0-kilobase (kb) Pst I fragment in all restored and revertant lines. Southern hybridizations with a 1.3-kb BamHI clone, internal to the 6.0-kb Pst I fragment, as a probe revealed two configurations of 6.0-kb homologous sequences in the sterile cytoplasm; one of the configurations was lost upon reversion or restoration. Mitochondrial DNA rearrangement has thus been observed upon restoration by a nuclear restorer gene in this CMS system.

The molecular weights of maize cytoplasmic and mitochondrial rRNAs were determined by gel acrylamide electrophoresis under non-denaturing and denaturing conditions. The molecular weights of mitochondrial rRNAs (0.76-10-6 and 1.25-10-6) exceeded those of cytoplasmic rRNAs (0.67-10-6 and 1.19-10-6) when electrophoresed in 8 M urea at 60 degrees C. Electrophoresis in 1.1 M formaldehyde resulted in similar values except for heavy mitochondrial rRNA, which exhibited a higher molecular weight than observed in 8 M urea. The observed values for cytoplasmic rRNAs, especially the heavy component, represent a decrease from estimates obtained under non-denaturing conditions. This is the first report of the electrophoretic examination of higher plant rRNAs under denaturing conditions.