 |
|||
| Volume 1, Issue 1, 2007 | |||
| Towards Cytoplasmic Male Sterility in Cultivated Tomato | |||
|
Pravda K. Stoeva-Popova, Winthrop University, USA,
stoevap@winthrop.edu |
|||
|
Abstract 1. The current state of studies on cytoplasmic male sterilityCytoplasmic male sterility (CMS) is a phenomenon observed in more than 150 plant species, and has been studied in a number of species including maize, sorghum, petunia, sunflower, radish, rice beans, Brassica, tobacco, and wheat (2, 3, 10, 11, 14, 15, 20, 21, 26, 29, 32, 37, 44, 45, 46, 49, 50, 51, 59, 60, 61, 62, 65, 68, 76, 77, 79, 80, 83, 84, 87, 92, 93, 102, 105). It can arise spontaneously, can be induced by mutagens, or be the result of interspecific, intraspecific and intergeneric crosses (28, 39). CMS is maternally inherited and based on changes in the mitochondrial (mt) chromosomal DNA structure and gene expression as influenced by nuclear genes (43). Alloplasmic male sterility results from interspecific and intergeneric crosses due to incompatibility between the nuclear genome of the recurrent pollinating parent and the mitochondrial genome of the pistillate parent (56, 95). The combination of alien cytoplasmic and nuclear genomes leads to mutations (rearrangement, changes) and disturbances in the mitochondrial genome, and/or reveals the expression of abnormal mitochondrial genes, the detrimental effect of which is not suppressed by nuclear genes present in the original maternal species, thus resulting in CMS (28, 27). Studies at the molecular level revealed the presence of CMS-associated chimeric genes resulting from mtDNA rearrangements in sterile cytoplasms. The CMS-associated loci demonstrate common features that are composed of copies or portions of coding regions of known mt genes and/or of unidentified sequences. However, the genetic structures of the CMS loci studied to date differ (26, 92), they often have ATP synthase subunit sequences and genes located within or near the CMS-associated loci as seen in the cytoplasms of petunia, maize S and T, Brassica pol, nap, tour and Ogura, rice, sorghum, tobacco, sunflower PET-1 and Arabidopsis. The coding regions of the cytochrome oxidase subunit gene have also been implicated in CMS-associated loci in wheat, rice, and petunia, while the coding regions of the nad subunit gene were found in petunia, Brassica nap and tour sterile cytoplasms. The unidentified sequences found in CMS-associated loci have no sequence homology to other known plant genes (reviewed in 27). CMS plants usually appear normal, vigorous, and undistinguishable from the fertile analogue (28). The main effect of CMS is revealed in the development of anthers and pollen, and leads to pollen abortion. In microscopic investigations of different CMS systems, these developmental aberrations occur at various stages from before meiosis through pollen maturation (39). Other changes may affect floral morphology and color, and recent studies indicate CMS disturbs the expression of floral homeotic genes (66, 53, 9, 27) Since the tapetum provides nutrition and materials for the formation of the complex pollen wall (81), its development and function plays a critical role in microsporogenesis, and aberrations in these can lead to male sterility. The abnormal development of the tapetum is a prominent feature of the CMS phenotype in plants. (12, 13, 39, 97, 26, 98, 106, 107). Tapetal cells undergo developmentally regulated lysis, the timing of which is critical for pollen development and when it occurs too soon causes pollen abortion as in genetically engineered plants (63, 38). Premature degeneration of the tapetum at the early to mid uninucleate microspore stage leads to the development of nonviable pollen as shown in tobacco arc-barnase transformants (78). Izhar and Frankel (34) showed that the changes in tapetum behavior are associated with changes in pH and callase activity. The faulty timing of enzyme activity and the subsequent effect on callose dissolution was suggested to be a primary factor in PMCs (pollen mother cells) and microspore abortion in Petunia. There are nuclear genes that can restore fertility, termed nuclear restorer (Rf) or fertility restorer (Fr) genes, which are specific for each studied CMS system. The action of the fertility restorer genes is usually correlated with changes in the transcript profile or protein accumulation of the mt CMS inducing gene. An exception to this is the first cloned and characterized fertility restorer gene, Rf, in maize. It codes for mitochondrial aldehyde dehydrogenase but does not affect the accumulation of the abnormal CMS related mt protein in cms-T maize (18, 54). Its role is defined as a “compensatory restorer”, which makes up for the malfunctions in metabolism caused by CMS-associated gene products (54). Other cloned restorer genes are the Rf in Petunia (8), Ogura Rfo/Rfk1 in Brassica and Raphanus (16, 19, 47) and Rf1 (PPR8-1 gene) in Boro rice (42, 48). The three genes are similar in that they encode proteins with multiple copies of a pentatricopeptide (PPR) motif that control the expression of the corresponding CMS-associated mt transcript. The specific mechanisms by which these genes prevent the expression of mitochondrial CMS-related mutations are currently unknown (reviewed in 27). However, Wise and Pring (101) hypothesized that with its abundance and potential to bind to RNA, the representatives of the PPR multigene family were most likely involved in RNA transcript processing. The genome-wide bioinformatics, genomic expression and functional study of Arabidopsis PPR proteins family supports the hypothesis; it uncovered that they play an important role in biogenesis of plant organelles. They have RNA binding ability and are for the most part targeted to specific transcripts. Due to their structure their presumed roles are as adaptors, binding together with other proteins to the transcript. It is considered that PPR proteins are probably involved in the posttranscriptional processing of organellar transcripts (57). In Phaseolus vulgaris, introduction of the nuclear fertility restorer gene, designated Fr, produced a permanent, non-segregating condition of male fertility with the shifting of the CMS-associated open reading frame pvs-orf239 to substoichiometric levels (58, 59, 36). An apparently identical genomic shifting phenomenon is observed upon spontaneous reversion to fertility (35, 36), with frequencies likewise dependent on nuclear background (60). 2. CMS in genus LycopersiconIn the tomato genus, Lycopersicon, CMS does not occur naturally, however, there are a few examples of induced CMS phenomenon (40, 91, 75, reviewed in 39). In these systems, CMS arose from interspecific crosses (4, 5, 94). Andersen (4, 5) reported the emergence and increase of pollen abortion respectively in F1 and backcrosses (BC) of the crosses between L. esculentum, L. cheesmanii var. minor, L. minutum, and L. hirsutum f. glabratum used as pistillate parents and L. pennellii as the recurrent pollinating parent. Pleiotropic effects of the CMS phenotype included the reduction of anther length and size, and the lengthening of the filaments. The reduction in pollen viability varied among the different crosses and was most pronounced (100%) in the BC1 (L. cheesmanii var. minor x L. pennellii) x L. pennellii. The reduction of anther size was negatively correlated to the percent of aborted pollen. Similar results were observed by Valkova-Atchkova (94) in crosses involving L. peruvianum as pistillate parent and L. pennellii and L. hirsutum f. typicum as pollinating parents. A progressive decrease in pollen viability was established after each backcross to the respective recurrent parent. In the cross L. peruvianum x L. pennellii, pollen abortion increased from 40.7% in BC1 to 99.2% in BC3. For the cross L. peruvianum x L. hirsutum f. typicum pollen abortion was 61.5% in BC1 and 97.9 % in BC3. The sterile plants used as female parents manifested normal fertility. Further introgression of the nuclear genome of the recurrent parents confirmed the stability of the CMS phenotype over many generations. Extensive somatic hybridisation between the cultivated tomato and a number of species from the subgenus Eriopersicon, genus Solanum and Nicotiana resulted in novel recombination patterns of nuclear and mitochondrial genomes (reviewed in 87). In only one study (64) a CMS phenotype appeared as a result of the fusion of cytoplasmic inactivated tomato protoplasts with nuclear inactivated Solanum (Solanum acuale and Solanum nigrum) protoplasts. Among regenerated fusion products female fertile plants with normal tomato characteristics were observed, but they completely lacked or had malformed anthers, had shrunken pollen, and pollen that did not germinate. The restriction analysis of mtDNA revealed that the mitochondrial genome of the CMS somatic hybrids did not combine all elements of the parental species and included new recombinant fragments. Presently, a CMS system does not exist for the cultivated tomato, Lycoperison esculentum. We have further studied the characteristics of the CMS line developed by Valkova-Atchkova (94), termed CMS-pennellii, that was produced from the late backcrosses of the cross L. peruvianum (pistillate parent) and L. pennellii (recurrent pollinating parent) (96, 70, 71, 72). Since its development in 1980, the maternal inheritance of the male sterility in CMS-pennellii has been stable over many subsequent backcross generations with L. pennellii as the pollinator. 2.1. Morphological and microscopic studies of CMS-pennellii. To study the morphological characteristics of the CMS-pennellii line we compared it to the fertile analogue L. pennellii. There were no differences in the leaves (number, size and form of leaf segments), the type of growth of the stem, and the arrangement of leaves and inflorescences. There was also no difference in the type of inflorescence between CMS-pennellii and L. pennellii. A current study of the correlation between the bud size, anther size, and stage of PMCs development indicates that CMS has a distinct differentiated morphometric influence on the development of the corolla, the anther and the filament. The flowers of CMS-pennellii have a smaller size corolla than the fertile analogue, which is not completely opened and is a pale yellow color, compared to the bright yellow color of L. pennellii. The mean bud size before anthesis in the CMS line is 6.04 mm, which is significantly smaller compared to L. pennellii - 8.25 mm. Reduction in the anthers size is established in the sterile line, which is observed at all stages of their development. At anthesis the mean length of the anther in the CMS-line is 1.64 mm, while for the fertile analogue it is 5.77 mm. Although CMS has a reduction effect on the size of the corolla and the anther, the reverse is detected for the size of the filaments. They continuously increase in length in CMS-pennellii reaching 1.92 mm at anthesis, which is twice as long as L. pennellii’s (72, 75, Stoeva, Petrova, Vulkova, and Dimaculangan unpublished data). The anthers are not coalesced laterally and do not form a staminal cone. They do not form an apical pore or stomial slit and do not shed pollen. The CMS-pennellii phenotype shows high similarity to the phenotypes produced from introgressive crosses with L. pennellii as a recurrent pollinator in the studies of Andersen (4, 5). This indicates that a common set of molecular mechanisms may be responsible for the CMS phenomenon in Lycopersicon, when the recurrent parent is L. pennellii. It also points to the fact that as in other well-studied CMS systems, the pleiotropic effect of CMS extends to a number of key flower characteristics (67, 95, reviewed in 27). The CMS line is completely female fertile and readily produces seeds after pollination with L. pennellii. The seed set and the size of the formed fruits slightly surpasses the size and seed set of the fertile analogue, L. pennellii (72, 75). Microscopic studies demonstrated that the development of the PMCs is normal. They undergo all stages of meiosis with a low percent of abnormalities and tetrads are formed. The degeneration of the microspores takes place after tetrad disintegration, resulting in shrunken pollen that does not stain with 1% acetocarmine (72, 75). The CMS-related structural changes in anther and pollen development in this system have not been well studied. Initial microscopic analysis revealed that the structure of the locules of the fertile analogue L. pennellii and the male sterile line were of identical type and had similar tissue developmental patterns (75). L. pennellii had four locules, while CMS-pennellii had two to four locules. Both plant types contained the outer epidermal layer of cells, two layers of outer tapetum, sporogenous tissue that gave rise to PMCs surrounded by callose, and an inner tapetal layer. The inner tapetal layer was highly vacuolated and built from larger cells. During prophase I the tapetal cells had two nuclei in both L. pennellii and CMS-pennellii. At the tetrad stage both L. pennellii and CMS-pennellii had well formed tetrads with microspores in callose. Each microspore had a well-distinguished central nucleus. The callose surrounding the tetrads was not uniform and the tapetal tissue was intact and contained large vacuoles. 2.2. Hybrids of CMS-pennellii with L. esculentum Although CMS-pennellii shows unilateral incompatibility like the recurrent parent L. pennellii, we were able to incorporate L. esculentum germplasm through a bridge hybrid technique developed by Stoeva (89, 90). Unique complex hybrids with sterile cytoplasm originating from L. peruvianum and the nuclear genomes from L. pennellii and L. esculentum were developed. The introgression of L esculentum germplasm yielded few plants with 100% restored pollen stainability which indicated that the cultivated tomato carries fertility restoration gene(s) for this system (72). The study of the F1, F2, F3 progenies of the complex hybrids CMS-pennellii x F1 L. esculentum x L. pennellii demonstrated the same pattern of segregation in all progenies into 3 major groups according to pollen stainability: (sterile: semi- sterile: fertile) with a ratio, which was not statistically different from 2:1:1 (72). 2.3. Studies of the mitochondrial genome of CMS-pennellii and its hybrids One of the specific features of plant mitochondrial genomes is their bigger size in comparison with mammals and fungi ranging from 200 kb in some Brassica species to 2500 kb in Cucurbita species. This phenomenon is due to the existence of the high percent of repeated sequences, chimeric and pseudo genes (103). These regions have a great capacity for recombination and rearrangements. In the CMS systems characterized to date, recombination events in the mt genomes created novel, chimeric open reading frames whose expression cause the male sterility. Since a similar phenomenon is most likely the cause of CMS in Lycopersicon, current efforts are directed toward identifying the CMS-associated loci in the CMS-pennellii mt genome. Presently, several approaches have been used to study the mitochondrial genome of CMS-pennellii and its complex hybrids with L. esculentum. 2.3.1. Restriction fragment length polymorphism (RFLP) studies To identify changes in the cytoplasm of the CMS line resulting from the interaction between the nuclear genome of L. pennellii and the cytoplasm of L. peruvianum we studied first the structural organization of essential mt genes. Heterologous probes from the following protein coding genes: atpA, atp6, atp9, nad3, coxII, coxIII, cob and genes coding for the mt 18S and 5S rRNAs were used in combination with different restriction enzymes. The hybridization profiles generated from blots of bulk DNA from the CMS–pennellii line and L. peruvianum, digested with different restriction enzymes, were analyzed. RFLP was found with atpA (with a 1.5 kb probe from Pisum sativum) and nad3 with a 0.8 kb probe from Petunia hybrida (104). In the HindIII digestions, the atpA probe hybridized respectively to 5.4 kb fragment in the CMS line and 5.3 kb in L. peruvianum. Double digestions with EcoRI/HindIII revealed two fragments in L. peruvianum – 3.6 and 5.0 kb, and only one fragment in the sterile line – 3.6 kb. The hybridization with the nad3 probe showed that this gene is proximal to atpA and lies on the same HindIII and EcoRI/HindIII fragments in CMS–pennellii and L. peruvianum (72). Polymorphic patterns were detected after hybridization with a 264 bp PCR probe from the coding region of atp9 from L. esculentum (41). The EcoRI/HindIII DNA digestions uncovered that the probe hybridizes to different bands in the CMS profile – 780 bp and 2.2 kb, compared to the 600 bp and 2.5 kb in L. peruvianum (75). The detected polymorphisms in the regions of mitochondrial genes atpA, nad3 and atp9 can be considered as an evidence of rearrangements generated from the interaction between the cytoplasm of L. peruvianum and the nucleus of L. pennellii in CMS–pennellii line. The repeated sequences are characteristics for the mitochondrial genomes of higher plants. They are sources of recombination and generate subgenomic mt DNA molecules (55). Small et al. (86) proposed that mtDNA evolution could result from changes in the proportions of substoichiometric molecules. Such reorganizations in the mt genome could be the result of nuclear directed changes as is the case in CMS and restoration of fertility in beans (60, 59) and leaf variegation in Arabidopsis (1). The significance of the low intensity fragments was studied by Yesodi et al. (104) in relation of the evolution of S-Pcf, the CMS-associated chimeric locus in Petunia hybrida. The authors detected fragments with low intensity, homologous to S-Pcf locus, situated on subgenomic molecules. Such CMS related sequences were identified in beans, sunflower, maize, rapeseed (6, 7, 36, 85, 88). Our hybridization experiments with specific probes on individual plants from the CMS line, L. peruvianum, and complex hybrid plants with restored fertility showed the existence of polymorphic fragments with low intensity in the regions of atpA and nad3 genes, which is an indication of the presence of recombinant molecules at substoichiometric level (75). The detection of low intensity fragments implies that these regions have a high recombination capacity. 2.3.2. PCR with specific mitochondrial gene primers As our second approach in investigating the differences arising from nuclear-cytoplasm interactions in the mt genome of CMS-pennellii and restored fertility plants, in comparison with L. peruvianum, we used several combinations of primers, generated from the sequences of cloned Solanaceae mitochondrial genes, to amplify intergenic mt DNA regions. Polymorphic patterns were produced by using of primer combinations atp9/coxI and atp9/rps12. (atp9L – gaaaatcgtgatggaaaaagc, coxIR – aattgctaaatcaactgctcct, Rps12R – ttccagaggcatcttccattca). The PCR product with primers atp9/coxI is 800 bp in length in the profile of CMS–pennellii line and in Rf plants, and 750 bp in the profile of L. peruvianum respectively. The primer combination atp9/rps12 generated different amplification products for the CMS line, Rf plants and the donor of cytoplasm L. peruvianum. Two fragments are detected in the CMS line – 340 bp and 200 bp. In plants with restored fertility a 200 bp product is amplified. The pattern of L. peruvianum contains a 340 bp amplicon. The primer combination atp9/rps12 is suitable as a PCR marker for early selection of plants with different phenotypes: cytoplasm male sterile plants or plants with restored fertility in the hybrid progenies (75). 2.3.3. Mitochondrial-specific randomly amplified polymorphic DNA analysis A third way to detect structural differences between the mt genomes of CMS-pennellii and the cytoplasmic donor species, Lycopersicon peruvianum is based our mt-specific randomly amplified polymorphic DNA (mtRAPD) technique (23). A comparison using 9 different primers revealed that there is a 68% genetic similarity (or 32% change) between the two mt genomes (69, Olson, Griffin, Stoeva, and Dimaculangan, unpublished). We also determined the genetic similarity between CMS and a fertility-restored hybrid to be 89%. In comparison, the percent similarities between three Lycopersicon species, L. esculentum, L. pennellii, and L. peruvianum were 42 to 48% in a 6-primer analysis. These data indicate the L. peruvianum mt genome has indeed undergone rapid evolution during the generation of the CMS line. In addition, we generated several polymorphic RAPD banding patterns between L. peruvianum and CMS (24, Olson, Griffin, Stoeva, and Dimaculangan, unpublished), and we are currently determining which bands most likely correspond to a CMS–associated locus. 2.4. Studies of the nuclear genome - isozymes A comparison of isozyme activity between CMS-pennellii, and the fertile analogue L. pennellii, was carried out for malate dehydrogenase (MDH), malic enzyme (ME), esterase (EST), peroxydase (PRX), superoxide dismutase (SOD) and glutamate oxalacetate transaminase (GOT). Qualitative changes were established for three of the studied isozymes in CMS-pennellii compared to L. pennellii. New fractions appeared in the elecrophoretic spectrum of CMS line for EST and ME, while two fractions of MDH were not present. The new fractions in the electrophoretic spectrum of EST and ME can be used as isozyme markers for the CMS line (25, 72). These changes in isozyme patterns point at altered expression of L. pennellii nuclear genes in L. peruvianum alien cytoplasm and are an indication for a crosstalk between the two genomes. According to Poyton (74) communication from the mitochondrion to the nucleus involves metabolic signals and one or more signal transduction pathways that function across the inner mitochondrial membrane. 3. Perspectives for future studies:For many cultivated plants the CMS phenomenon allows for the development and use of a simple and reliable method for hybrid seed production. CMS provides the opportunity for large scale production of hybrid seeds, enabling the exploitation of hybrid vigor in F1 generation. The creation of CMS sources in cultivated tomato has been (52) and is still an unsolved problem. What are the prospects to create CMS in cultivated tomato? In order to create a CMS system for the cultivated tomato one must produce a maintainer of sterility line and maintainer of fertility line(s), with the phenotype of the cultivated tomato. 3.1. Development of maintainer of fertility lines. The maintainer of fertility line would have the phenotype of L. esculentum and carry restorer of fertility genes. At present the knowledge of our CMS system presumes that restorer of fertility gene(s) are part of the genome of the cultivated tomato (72, 75). The first step in the development of this line was made by overcoming the unilateral incompatibility of the CMS-pennellii line (inherited from the recurrent parent L. pennellii) when used as pistillate parent in crosses with L. esculentum (self-compatible species). We overcame the unilateral incompatibility by the using pollen from the bridge hybrid F1 L. esculentum X L. pennellii (89, 90, 72). The pollination of 578 flowers of CMS-pennellii with F1 L. esculentum X L. pennellii yielded 15 fruits with 1-5 seeds per fruit. Thirty four complex hybrids were grown and studied. Five plants had pollen stainability in the range of 70-80%, while five plants had pollen stainability above 80 %. The grounds for accepting 70% pollen stainability as the lower range value for male fertile plants is the 70-75% pollen stainability of the interspecific hybrid F1 L .esculentum X L. pennellii. F2 and F3 generations of this complex hybrid were produced by sib-mating on male sterile plants (with pollen stainability 0-1%) with male fertile plants (80-100% pollen stainability). For further introgression of L. esculentum, plants from the F2 and F3 were backcrossed to F1 L. esculentum X L. pennellii. Plants from this second backcross generation were directly pollinated with the cultivated tomato (72). The one-step restoration of fertility in the complex hybrid F1 (CMS-pennellii x F1 L .esculentum X L. pennellii) was the indication that restorer of fertility genes are found in the genome of L. esculentum. The mode of segregation for pollen stainability in the studied F1, F2 and F3 generations infers the role at least one major dominant Rf gene from L. esculentum (72). Presently male fertile plants carrying the CMS cytoplasm and L. esculentum phenotype are available. Eventually most L. esculentum genotypes are expected to carry an Rf gene(s) although introgressive crosses with more L. esculentum genotypes should be carried out in support of this hypothesis. Once a mitochondrial CMS-associated region is identified, an important goal will be to demonstrate the effect of the introgression of L. esculentum nuclear genes on the expression of this detrimental mitochondrial locus. It is expected as in many studied CMS systems the presence of restorer of fertility gene(s) will affect the expression of the mitochondrial CMS-associated locus by reducing the abundance of transcript and/ or of the encoded protein. Such data will support the data produced from the genetic studies (72) and will provide in molecular data evidence that the cultivated tomato is a carrier of Rf genes. 3.2. Maintainer of sterility line CMS in our system arose from the interaction between the nuclear genome of L. pennellii Atico and the mitochondrial genome of L. peruvianum. The creation of maintainer of sterility lines with the tomato phenotype and maintainer of sterility (rf) genes from the wild species L. pennellii Atico, is an essential step in the development of CMS system in cultivated tomato. As a model species for genetic and molecular studies, the tomato has well-developed, high-density molecular maps, genomic and cDNA libraries http://www.sgn.cornell.edu/, and a rapidly developing expressed sequence tags (ESTs) database (180,000) (http://www.tigr.org/tigr-scripts/tgi/T_index.cgi?species=tomato). Eshed and Zamir (22) produced tomato near isogenic lines (NILs) with L. pennellii Atico. Each of the NILs harbours specific individual chromosome fragment from L. pennellii into homogeneous L. esculentum background. The set of 50 lines represents a permanent mapping population which covers the whole genome of L. pennellii and is publicly available (http://tgrc.ucdavis.edu/pennellii-ILs.pdf). Most of these lines are highly fertile. 25 sublines with smaller chromosome segments are available, which are additional tools for greater resolution mapping studies and could be implemented in advanced stages of the study. To search which segment(s) of L. pennellii chromosomes harbour maintainer of sterility genes we should produce F1 hybrids between the CMS-pennellii line and each of the NILs. This will be the milestone to overcome, since CMS-pennellii shows unilateral incompatibility and by rule does not cross with L. esculentum as a pollinating parent. In our experience with this line we have been able to produce few seeds (Stoeva unpublished) after pollinating 100 flowers of CMS plants grown in environmental chamber with pollen of L. esculentum. The production of the F1 hybrid will need a broad scale pollination of CMS with individual NILs. Our hypothesis is that a F1 hybrid between CMS-pennellii and the line harbouring a chromosomal segment with maintainer of sterility gene will exhibit (support) the CMS phenotype, in comparison to F1 hybrids with other NILs which will be male fertile. The next step will be to perform repeated backcrosses of the (CMS) F1 hybrids with the same NIL (presumed carrier of rf genes) as a recurrent pollinator. Available EST-SSRs and genomic SSRs (http://www.tigr.org/tigr-scripts/tgi/T_index.cgi?species=tomato) as well as PCR based DNA markers will be used to map the specific region(s) carrying the maintainer of sterility gene(s). The developed DNA markers will be used to apply the PCR-based approach for marker assisted selection of maintainer of sterility lines for respective tomato lines, suitable for hybrid seed production. 3.3. Ultrastructural microscopic studies The established CMS system has not been well characterized by microscopic techniques. By using light, transmission electron and confocal microscopy we could determine the ultrastructural abnormalities in the CMS-pennellii line in comparison to the fertile analogue. Our preliminary studies (Stoeva, Petrova, Vulkova, and Dimaculangan unpublished data) correlating bud size, anther size, and stages of microsporogenesis among various plants in the tomato CMS system will facilitate the studies and be a good background to make valid comparisons between various components of the system. The tapetum is critical for pollen development, because it is the source of nutrients for developing pollen and the enzymes necessary for microspore liberation from the tetrads, and it provides pollen wall components. During the late tetrad stage, the tapetum begins production of the b-1,3 glucanase enzyme that is thought to be involved in the dissolution of the callose wall surrounding the tetrads, and the tapetum derived sporopollenin deposition consolidates the exine of the mature pollen grain (81). Fibrils, vesicles and lipid droplets are released from the intact tapetum cell and transferred to the developing pollen grain (73). Therefore of special interest will be to study the development of the tapetum, pollen mother cells, and microspores and track the temporal changes in correlation with CMS phenomenon. The development of the exine and the intine of the microspore could be studied to assess its relationship to microspore degeneration in CMS versus fertile plants. Since CMS is associated with mt DNA mutations it will be of particular significance to focus on the structural organization of the mitochondria in the tapetum and the microspores. It will be important to compare the numbers of mitochondria in the CMS line, fertile analogue and hybrids with different pollen fertility. The information from ultra structural studies will enrich the knowledge about the system, allow us to compare the system to other studied ones, and for in situ hybridization studies. 3.4. Identification of the CMS-associated mitochondrial loci Since there is a propensity for plant mt genomes to undergo a high amount of recombination, the finding of recombinant mitochondrial genes is only the first step in identifying the CMS-associated mitochondrial loci. We are using the following additional criteria when assessing if a mtDNA arrangement is responsible for CMS: 1) it should contain or be near ATP genes and/or cox /or nad genes [coding regions and/or promoter regions] and have sequences of unidentified origin; 2) it should be present in CMS lines but absent in the donor of the cytoplasm; 3) it should be unchanged in the restored hybrids if the restorer gene acts to modify the expression as most commonly found, or it could appear to be missing in restored lines as in Fr restored hybrids in Phaseolus vulgarus (31); and 4) it should be expressed at the mRNA and protein levels in the CMS lines and have altered expression in the restored hybrids. Additional evidence that a specific locus contains a CMS gene could come from functionally testing it as part of a nuclear transgene expressed in normal plants and assessing its ability to induce male sterility. This strategy worked for ORF239 cloned from Phaseolus vulgaris sterility when it was expressed in tobacco plants (30). However, the lack of CMS activity would not be conclusive, since similar transgenic constructs with recognized CMS-associated genes from petunia and maize did not produce male sterility (99, 17). 4. Concluding remarksThe review of the present state of the studies on the established CMS systems in plants, our current knowledge of the CMS system in Lycopersicon and the discussed perspectives outline several strategic priorities for our future research. The most important task is to identify, isolate, and characterize the molecular and functional level of the mitochondrial CMS-associated locus in CMS-pennellii. The discovery of the CMS-associated locus is the necessary grounds for the studies aiming at the identification and isolation of the Rf genes. Another important step and a prerequisite to our efforts in generating maintainer of sterility lines, is to develop mtDNA specific PCR based markers, which distinguish between male sterile and restored fertility plants in segregating populations. These important and attainable steps we consider our short-term goals. In the long run from the standpoint of our present understanding of CMS in Lycopersicon, the most challenging task towards CMS in the cultivated tomato will be the development of maintainer of sterility lines. Furthermore, we consider that of fundamental interest will be to research the hypothesis that a common mechanisms exists in alloplasmic CMS systems created when L. pennellii is the donor of the nuclear genome. For example, a comparison of the mtDNA lesions between CMS-pennellii and CMS systems in which L. esculentum (or another Lycopersicon species) are the donors of the cytoplasm and L. pennellii is the donor of the nucleus (4, 5) could provide an insight to the interactions between the nuclear and mitochondrial genomes in plants.
References
|