CYTOPLASMIC MALE STERlLlTY AND MALE STERILE MUTANTS
Plant mitochondrial dysfunction can occur for at least two reasons: spontaneously arising mutation/genomic rearrangements and nuclear-cytoplasmic incompatibilities. Nuclear- cytoplasmic incompatibilities generally arise when the cytoplasm from one species is combined with the nucleus of a different species. This is effected most readily either by interspecifc protoplast fusion or by wide hybridization followed by subsequent recurrent backcrossing to the nuclear donor (pollen parent in the original cross). In most cases of mitochondrial mutation or nuclear-cytoplasmic incompatibility the observed phenotype is that of cms. The cms phenotype is characterized by the inability of the plant to produce and/ or shed viable pollen. The phenomenon demonstrates maternal inheritance and is generally not accompanied by changes in female fertility.
The apparent association of plant mitochondrial genome alterations with aberrant pollen development remains an enigma. It is not clear why this particular plant developmental stage would be most vulnerable to alterations in mitochondrial function when, for most of the identified mitochondrial mutations, expression of the sterility-associated mitochondrial product is detected at all plant developmental stages. The processes of microsporogenesis and microgametogenesis are affected differently in many cases of cms. For example, in at least two cases, only the individual developing microspores (gametophytic stage) are affected by the mitochondrial lesion (Lee et al., 1980; Johns et al., 1992), whereas in many other cases the abnormality is most pronounced as premature or aberrant breakdown of the tapetum (sporophytic tissue surrounding the developing microspores). In certain cases, male sterility is the result of abnormal development or feminization of the male reproductive structures (Kofer et al., 1991; Goiirret et al., 1992). Consequently, the precise points in the pollen development process most profoundly affected by mitochondrial dysfunction are not known. It is generally assumed, however, that the development of the male gametophyte in plants is accompanied by a pronounced increase in respiratory activity, because mitochondrial numbers increase markedly following meiosis. A commensurate increase in mitochondrial DNA replication accompanies these processes (reviewed by Dickinson, 1987).
The mitochondrial mutations associated with cms in several plant species have been identified and characterized. In each case the alteration is distinct, although particular regions of the mitochondrial genome, such as the 3' end of the atpA gene, appear to be more highly susceptible to rearrangement. cms-associated alterations may involve DNA deletion (e.g. Chetit et al., 1992), insertion of sequences of unknown origin (e.g. Laver et al., 1991; Johns et al., 1992), multiple intragenic recombination events generating chimeric gene arrangements (e.g. Younig and Hanson, 1987; Levings, 1993), often producing polycistronic messages and, in one case, unusual viruslike particles (Grill and Garger, 1981). This diversity of mutations initially made it difficult to identify those features
linking each system to a similar aberrant phenotype. One feature that appears to link many of the sterility-associated mitochondrial sequences is that they generate novel open reading frames. Perhaps the most important feature of the predicted gene products of these open reading frames is the presence of a hydrophobic stretch of amino acids, often at the amino terminus of the predicted polypeptide. This observation suggests that, although each sterility-associated sequence appears distinct, the mechanism for inducing mitochondrial dysfunction may involve a common phenomenon, namely the anchoring to, or complete insertion within (Levings, 1993), the inner mitochondrial membrane of an aberrant mitochondrial peptide. It is conceivable that such unusual peptide insertions may have relatively minor consequences to mitochondrial function during most stages in plant development, but may cause phenotypic changes when respiratory demand is sharply increased. This hypothesis, although often suggested, has not yet been tested for any cms system. Alternatively, cms mutations may invoke tissue-specific or stage-specific changes in gene expression, as has been suggested in the case of cms petunia by Conley and Hanson (1994).