The Fo-ATPase proteolipid, also referred to as subunit 9 or the dicyclohexylcarbodiimide-binding protein, is encoded by a mitochondrial gene in maize that we have des- ignated atp 9. The clone containing atp 9 was selected for inves- tigation from a mitochondrial DNA library because of its abundant transcript in total maize mitochondrial RNA prepa- rations. Sequence analysis of the clone revealed an open read- ing frame that was readily identified by its nucleotide homolo- gy with the ATPase subunit 9 gene of yeast. As deduced from the nucleotide sequence, the maize ATPase subunit 9 protein contains 74 amino acids with a molecular weight of 7368. Sub- stantial amino acid sequence homology is conserved among maize, yeast, bovine, and Neurospora mitochondrial ATPase subunit 9 proteins, regardless of whether the gene is nuclearly encoded (bovine and Neurospora) or mitochondrially encoded (yeast and maize). RNA transfer blot analysis indicated that the gene sequence is actively transcribed, producing an initial transcript that is large and extensively processed.

URF13, a mitochondrial membrane protein in maize (Zea mays L.) carrying the cms-T cytoplasm, is responsible for Texas cytoplasmic male sterility and susceptibility of cms-T maize to the fungal pathogens Cochliobolus heterostrophus race T and Mycosphaerella zeamaydis. URF13 is a ligand-gated, pore-forming receptor for the pathotoxins produced by these fungal pathogens. URF13, which contains three transmembrane α-helices, forms oligomers in cms-T maize mitochondria and when expressed in Eschericia coli cells. To study the tertiary and quaternary structure of URF13 oligomers and their interactions with the pathotoxins, we employed site-directed mutagenesis and chemical cross-linking. We determined that URF13 oligomers contain a central core of helices II. More recently, we introduced Cys residues into consecutive positions 78–82 and tested for diamide-induced cross-linking of the introduced Cys residues to Cys-27, the only Cys residue in native URF13. Our results suggest that amino acids 78–82 extend out of the membrane and into the center of the URF13 oligomers.

A sterile strain of corn that has been used for making hybrid seed corn carries the Texas male-sterile cytoplasm and does not make pollen. In a report in this issue of Science (see Cui et al., page 1334), cloning of a gene that can restore fertility to these plants is announced. In his Perspective, Levings discussed why the identification of this gene Rf2 as an aldehyde dehydrogenase is surprising and how it may work to reverse the male sterile phenotype.

The URF13 protein, which is encoded by the maize mitochondrial T-urf13 gene, is thought to be responsible for pathotoxin and methomyl sensitivity and male sterility. We have investigated whether T-urf13 confers toxin sensitivity and male sterility when expressed in another plant species. The coding sequence of T-urf13 was fused to a mitochondrial targeting presequence, placed under the control of the cauliflower mosaic virus 35S promoter, and introduced into tobacco by Agrobacterium tumefaciens-mediated transformation. Plants expressing high levels of URF13 were methomyl sensitive. Subcellular analysis indicated that URF13 is mainly associated with the mitochondria. Adding methomyl to isolated mitochondria stimulated NADH-linked respiration and uncoupled oxidative phosphorylation, indicating that URF13 was imported into the mitochondria, and conferred toxin sensitivity. Most control plants, which expressed the T-urf13c construct lacking the mitochondrial presequence, were methomyl sensitive and contained URF13 in a membrane fraction. Subcellular fractionation by sucrose gradient centrifugation showed that URF13 sedimented at several positions, suggesting the protein is associated with various organelles, including mitochondria. No methomyl effect was observed in isolated mitochondria, however, indicating that URF13 was not imported and did not confer toxin sensitivity to the mitochondria. Thus, URF13 confers toxin sensitivity to transgenic tobacco with or without import into the mitochondria. There was no correlation between the expression of URF13 and male sterility, suggesting either that URF13 does not cause male sterility in transgenic tobacco or that URF13 is not expressed in sufficient amounts in the appropriate anther cells.

The mitochondrial Rieske iron-sulfur protein is an obligatory component of the respiratory electron transport chain that is encoded by a single-copy gene in mammals and fungi. In contrast, this protein is encoded by a small gene family in dicotyledonous tobacco and monocotyledonous maize. We cloned four cDNAs from tobacco that encode the mitochondrial Rieske iron-sulfur protein. These clones, along with a previously isolated cDNA, represent five independent members of the gene family that can be divided into three subfamilies. All of these genes were derived from the two progenitor species and were expressed in amphidiploid tobacco. The proteins encoded by these five genes are probably functional because they all contain the universally conserved hexyl peptides necessary for the 2Fe-2S cluster formation. The expression of the Rieske protein gene family is differentially regulated; a 6- to 11-fold higher level of steady state transcripts was found in flowers than in leaves, stems, and roots. Members of at least two subfamilies were preferentially expressed in flowers, indicating that they share a common cis-regulatory element(s), which can respond to a flower-specific signal(s). Although approximately 10 times more transcripts occurred in flowers than in leaves, flower and leaf mitochondria contained a similar amount of the Rieske protein. Flowers, however, contained seven times more Rieske proteins than leaves. These results indicated an increase in mitochondrion number in flowers. High-energy demands during anther development might bring about an increase in mitochondrion numbers in flowers and the flower-enhanced expression of the Rieske protein gene family. Our results suggested that nuclear genes encoding mitochondrial respiratory proteins could sense and respond to changes in energy metabolism and/or changes in mitochondrion numbers.

URF13 is a mitochondrially encoded, integral membrane protein found only in maize carrying the cms-T cytoplasm. URF13 is associated with cytoplasmic male sterility, Texas type, and causes susceptibility to the fungal pathogens Bipolaris maydis race T and Phyllosticta maydis. URF13 is predicted to contain three transmembrane alpha-helices and is a receptor for the pathotoxins (T-toxins) produced by B. maydis race T and P. maydis. Binding of T-toxin to URF13 leads to membrane permeability. Cross-linking of URF13 oligomers with N,N'-dicyclohexylcarbodiimide (DCCD) protects Escherichia coli cells expressing URF13 and cms-T mitochondria from the permeability caused by T-toxin or methomyl. Using mutated forms of URF13 expressed in E. coli cells, we determined the molecular mechanism of DCCD protection. We separately changed Lys-37 in helix II to isoleucine (K37I-URF13) and Lys-32 in the helix I/helix II loop region to alanine (K32A-URF13). DCCD treatment of K37I-URF13-expressing cells did not protect the cells from permeability caused by T-toxin or methomyl. DCCD cross-linking was greatly reduced in K37I-URF13 and in D39V-URF13-expressing cells, but it was unaffected in K32A-URF13-expressing cells. Binding of methomyl or T-toxin decreases DCCD cross-linking of URF13 oligomers expressed in either E. coli or cms-T mitochondria. We conclude that Asp-39 in helix II is cross-linked by DCCD to Lys-37 in helix II of an adjacent URF13 molecule and that this cross-linking protects against toxin-mediated permeabilization. Our results also indicate that helices II form a central core in URF13 oligomers.

The URF13 protein, which is encoded by the mitochondrial gene T-urf13, is responsible for cytoplasmic male sterility and pathotoxin sensitivity in the Texas male-sterile cytoplasm (cms-T) of maize. Mitochondrial sensitivity to two host-specific fungal toxins (T toxins) is mediated by the interaction of URF13 and T toxins to form pores in the inner mitochondrial membrane. A carbamate insecticide, methomyl, mimics the effects of T toxins on isolated cms-T mitochondria. URF13 was expressed in Spodoptera frugiperda (fall armyworm) cells (Sf9) in culture and in Trichoplusia ni (cabbage looper) larvae with a baculovirus vector. In insect cells, URF13 forms oligomeric structures in the membrane and confers T toxin or methomyl sensitivity. Adding T toxin or methomyl to Sf9 cells producing URF13 causes permeabilization of plasma membranes. In addition, URF13 is toxic to insect cells grown in culture without T toxins or methomyl; even a T-toxin-insensitive mutant form of URF13 is lethal to cell cultures. Baculoviruses expressing URF13 are lethal to T. ni larvae, at times postinjection comparable to those obtained by injecting a baculovirus expressing an insect neurotoxin. This result suggests that URF13 could be useful as a biological control agent for insect pests. Our data indicate that URF13 has two independent mechanisms for toxicity, one that is mediated by T toxin and methomyl and one that is independent of these toxins. Similarly, male sterility and toxin sensitivity in cms-T maize may be due to independent mechanisms.

Cytoplasmic male sterility (CMS) is a maternally inherited trait that suppresses the production of viable pollen grains. The Texas, or T, cytoplasm (cms-7) of maize, which was first described in the Golden June line in Texas (Rogers and Edwardson, 1952), carries the CMS trait. The Texas cytoplasm was an important discovery to geneticists and plant breeders because it eliminated the costly detasseling procedure used in maize hybrid seed production. Its commercial value has prompted many applied and basic studies of cms-T aimed at understanding CMS and its relationship with mitochondrial and nuclear genes. Although CMS has been observed in more than 150 plant species (Laser and Lersten, 1972), the mechanism by which it interrupts normal pollen development is not well understood. In this review, the basis of CMS in cms-T maize is considered. In cms-T; male sterility is characterized by the failure of anther exertion and pollen abortion. Female fertility is not affected by CMS, so male-sterile plants can set seed if viable pollen is provided. Comparisons of plants carrying the T and normal cytoplasms reveal very slight differences in severa1 other morphological characters (see below; Duvick, 1965). Initially, the three male-sterile cytoplasms of maize-cms-T, cms-C, and cms-S-were distinguished by specific nuclear genes, termed restorers of fertility (Rf), that suppress the male-sterile effect of the various cytoplasms and allow viable pollen production. For example, two genes, Rf7 and Rf2, acting jointly, restore pollen fertility to cms-7: Rf7 and Rf2do not restore pollen fertility to cms-C or cms-S; instead, different restorer genes are necessary to restore pollen fertility to these cytoplasms. Other characteristics also distinguish male-sterile and normal maize cytoplasms; they include mitochondrial DNA (mtDNA) restriction fragment length polymorphisms, variations in mitochondrial RNA (mtRNA), and differences in mitochondrial translational products (Pring and Levings, 1978; Leaver and Gray, 1982; Newton, 1988). Early on, these distinctions indicated that the various CMS types are based on different mechanisms and hereditary factors. The T cytoplasm is best known for the part it played in the U.S. epidemic of Southern corn leaf blight of 1969 and 1970 (Williams and Levings, 1992). In the two decades before the epidemic, cms-T had replaced detasseling as the chief method of pollen control in hybrid corn production (Wych, 1988), and by 1970, 85010 of the hybrid corn grown in the United States carried the T cytoplasm. After it was determined that cms-T was specifically susceptible to Bipolaris maydis race T, the organism responsible for the blight, its use by the hybrid seed corn industry was largely terminated (Ullstrup, 1972). Phyllosricra maydis, another fungal pathogen, is also specifically virulent on cms-7: Normally, B. maydis race T is a serious pathogen only on maize containing the T cytoplasm. Other male-sterile and normal maize cytoplasms support only limited colonization by the pathogen, and the lesions on the leaves remain small and isolated. By contrast, B. maydis lace T can quickly and completely colonize cms-Tmaize plants, causing extensive plant damage and sometimes death. Susceptibility of cms-Tto B. maydis race T is caused by mitochondrial sensitivity to a host-specific pathotoxin, designated BmT toxin, produced by the pathogen. P maydis also produces a pathotoxin that is structurally similar to the BmT toxin of B. maydis, to which cms-Tmitochondria are sensitive. Accordingly, susceptibility of cms-Tmaize to these fungal pathogens is due to mitochondrial sensitivity to the pathotoxins, whereas disease-resistant maize types have mitochondria that are insensitive to the pathotoxins.

URF13, an inner mitochondrial membrane protein of the maize Texas male-sterile cytoplasm (cms-T), has one orientation in the inner membrane of maize mitochondria but two topological orientations in the plasma membrane when expressed in Escherichia coli. Antibodies specific for the carboxyl terminus of URF13 and for an amino-terminal tag fused to URF13 in E. coli were used to determine the location of each end of the protein following protease treatments of right-side-out and inside-out vesicles derived from cms-T mitochondria and the E. coli plasma membrane. Cross-linking studies indicate that a portion of the URF13 population in mitochondria and E. coli exists in membranes in an oligomeric state and, in combination with proteolysis studies, show that individual subunits within a given multimer have the same orientation. A three-membrane-spanning helical model for URF13 topology is presented.

cDNA clones coding for the nuclear-encoded mitochondrial Rieske iron-sulfur protein (RISP) have been isolated from maize and tobacco. Complementation analysis of hybrid proteins consisting of different domains of plant and yeast RISPs showed that the carboxyl two-thirds of the plant protein is functionally equivalent to that of the yeast protein. The amino terminus of the RISP, however, seems to be species specific because this region is not interchangeable between plant and yeast proteins. Complementation analysis of hybrid proteins also identified a structurally conserved domain probably essential for the function of bc1 complex RISPs. A specific domain from the plant RISP was found to cause temperature-sensitive respiratory growth in yeast. We have demonstrated that yeast can serve as a model system to study the structural and functional relationships of plant gene products that are enzymatic components of the mitochondrial respiratory chain.

Aberrant recombinations involving the mitochondrial atp9, atp6 and coxII genes have created unique chimeric sequences in the C male-sterile cytoplasm (cms-C) of maize. An apparent consequence of the rearrangements is the interchanging of transcriptional and/or translational regulatory signals for these genes, and alterations in the reading frames encoding the atp6 and coxII genes in the C cytoplasm. Particularly unusual is the organization of the atp6 gene in cms-C mitochondria, designated atp6-C. The atp6-C sequence is a triple gene fusion product comprised of DNAs derived from atp9, atp6 and an open reading frame of unknown origin. Although there is no direct evidence indicating that these chimeric genes are responsible for the cytoplasmic male sterility (cms) trait, their novel arrangements and the strong correlation between these genes and the C type of male sterility suggest such a role.