Project Topics
Get Complete Project Material File(s) Now! »
Cytoplasmic male sterilities are the result of a cytonuclear incompatibility
As it was mentioned in section 1.1.2, CMS is of special interest in agronomy. The genetics behind this sterility has been widely studied since the discovery of the maize CMS-T, and later on in other crop species. Even though the sterilizing determinant is challenging to identify, some studies succeeded. In 1927, R. Chittenden and C. Pellew showed that CMS is the result of combination of nuclear genes together with cytoplasmic factors [6]. In fact, the sterilizing factor is mostly comprised in the cytoplasmic fraction, whereas the nucleus can counter-act its e ect. This section aims at explaining in detail these interactions: where they come from, and their consequences.
The plant cell is composed of three genomes that must cooperate and have co-evolved
At least three genomes are present in the plant cell, that have di erent inheritance patterns. Other genomes can be present, such as those of endosymbionts, parasites or viruses, but they are not taken into consideration here. The plant genomes have di erent inheritance patterns: most species inherit mitochondria and plastids maternally [16]. There are few exceptions: some species, for example those of the Oenothera genera, perform maternal and bi-parental plastid inheritance [17]. In Medicago sativa plastids are inherited paternally [18], whereas in cucumber mitochondria are inherited paternally [19]. Nevertheless such bi-parental or paternal inheritance is scarce. The di erential inheritance of the three genomes of plant cells is represented in gure 1.3.
Genomic con ict in cytoplasmic male sterilities
The arising of a sterilizing gene in cytoplasmic genomes has to be explained in the light of the genomic con ict theory from L. Cosmides and J. Tooby (1981) [47]. In most cases, the cytoplasmic genomes, i.e. the plastome and the chondriome are inherited maternally whereas the nuclear chromosomes are inherited following Mendelian’s segregation (see gure 1.3). Therefore, the plant genome can be divided into fractions, the tness of all genes in a fraction being maximized in the same way. Each fraction is selected in a way that the genes comprised in it maximally propagate, possibly in a deleterious way for the other fractions’ genes [47]. Cytoplasmic genes are not disadvantaged by the male sterility (as long as the plant can be pollinated by a male-fertile plant), whereas nuclear genes are [48]. In response to the presence of the sterilizing cytoplasm, the nuclear genome can evolve to counter-act the e ect of the sterilizing gene [49]. The nuclear genes counteracting the sterilizing gene e ect are called restorers of fertility, and they are historically labeled Fr or Rf depending on the species. In CMS systems, the nucleus can be of two types: restorer (of fertility) if it carries restorer genes counter-acting the sterilizing factor, or maintainer (of sterility) if it allows the sterilizing factor e ect.
Females can spread and persist in natural populations for two reasons. Firstly, it is widely accepted that female plants in gynodioecious populations have better female fertility (for example more seeds, or higher germinating seeds) than their hermaphroditic counterparts [50, 51]. This \female advantage » might be explained by the arrest of male gametophyte development at early developmental stage that thus saves resources that are re-allocated for female gametophyte production [52]. In some species however, the CMS individuals do not produce more or better seeds. In these species, the persistence of females can be explained in the light of the \restoration cost » phenomenon: in the absence of the sterilizing gene, the restorer (of fertility) alleles are counter-selected in a population [53]. This cost is di erent in the CMS systems considered, regarding the speci city of the restorer allele. For example, some restorers have deleterious pleiotropic e ects, and are therefore counter-selected when they are not in presence of the sterilizing cytoplasm [54]. In other CMS systems, the restorer evolved from nuclear genes that regulate cytoplasmic gene expression, called \housekeeping » genes. Hemizygous plants for these restorers carrying the sterilizing cytoplasm are restored, and the non-mutated restorer allele can perform its housekeeping role simultaneously. Now, xation of such restorer has a cost, as the housekeeping duty of the non mutated restorer allele is not achieved [54].
To sum up, cytoplasmic male sterility results from an incompatibility between two cell fractions that have di erent interests, the cytoplasmic and the nuclear compartments. The cytoplasm can be either sterilizing or non sterilizing, whereas the nucleus can carry maintainer (of sterility) or restorer (of fertility) alleles, allowing or counter-acting the e ects of the sterilizing gene, respectively. The di erent possible combinations of cy-toplasms and nuclei in one CMS system are presented in gure 1.5. It is noteworthy that individuals carrying the sterilizing cytoplasm can be hermaphroditic if they have restorer alleles, so that the e ect of the sterilizing protein is hidden in these individuals. Actually, the high frequency at which CMS appears when breeding di erent species or genera [55] indicates that sterilizing cytoplasms might be common, but their e ect is hidden because of the xation of restorer alleles. Indeed, if the female advantage is too low, or if the restoration is not costly, the restorer allele is eventually xed [56]. The intra- or inter-speci c hybridization in this case leads to co-adaptation disruption (see section 1.2.1), the restorer alleles are not present anymore to counter-act the sterilizing factor e ect and male-sterile plants arise. These cytoplasmic male sterilities are called cryptic [56]. Therefore, a \naive » nucleus in contact with a sterilizing cytoplasm results in female plants and gynodioecious populations.
Identi cation and validation of cytoplasmic male sterility associated genes
In the previous section, I detailed the common features of CMS-related genes. For-mally identifying them is often tricky. Approaches that have been employed are reviewed in M. Hanson and S. Bentolila (2004) [52]. Two approaches allow unequivocal identi ca-tion of a mitochondrial sterilizing gene. The rst strategy consists in using comparative genomics: by comparing sterilizing mitochondria with closely related fertile revertants, one can identify a DNA region that would be speci c for the sterilizing genome. Un-fortunately, this method is often limiting as it is di cult to nd mitochondrial genomes that recently diverged and just di er at the CMS locus. Maize CMS-T urf13 [79, 80] and CMS-S orf355/orf77 [81] as well as common bean pvs [73] sterilizing genes have been identi ed in this way. Another strategy is to search for the segregation of a DNA sequence inducing the male sterility phenotype.
As the cytoplasmic compartment is maternally inherited in most species, this approach is made possible by the technique of protoplast fusion.
This technique breaks the monoparental inheritance and allows recombinations. For ex-ample, recombined mitochondrial genomes of rapeseed cybrids were used in addition to comparison of fertile revertants to identify orf138 as the sterilizing gene in Ogura CMS [70, 82, 83].
In most cases, genetic proofs such as recombinant or revertant fertile plants are not available. In this instance, several lines of evidence from di erent approaches are neces-sary to validate the sterilizing gene. Comparing mitochondrial genomes and transcript pro les, and researching for unusual recombinants at the same time provide good hints on a candidate. The comparison of mitochondrial genomes is fastidious, as plant chondriomes are very large [84] and new orfs that could potentially induce CMS are common [85, 86]. As it was detailed in the previous section, CMS-related genes are often co-transcribed with essential mitochondrial genes. Therefore, looking at transcription pro les of these essential genes can highlight CMS-related candidate genes. Yet, these approaches are often not su cient, and they need to be supported by other lines of evidence. I already mentioned that CMS was determined by cytonuclear epistasis. The nucleus can carry restorer alleles, whose identity and mode of action will be presented in detail in section 1.3.2.2. Yet, it is interesting to mention that they often silence the sterilizing factor by acting on its expression. Therefore, investigating the candidate expression in sterile and restored lines could give clues to formally identify the CMS factor. Indeed, most of the fertility restorer genes alter the sterilizing gene expression [87]. However, a restorer can modify the expression pro le of genes other than the sterilizing one [88], so this method cannot prove alone the identity of the sterilizing gene neither. Examples of multiple clues that had to be advanced to formally identify CMS-related genes are detailed hereafter. The factor responsible for CMS-WA was identi ed via the study of mitochondrial transcription pro les by RNA blotting: the mRNA pro le detected with an rpl5 probe was di erent between the CMS-WA line and a fertile line. It was then shown that this mRNA was a ected in a restored nuclear background, and its sequence revealed an ORF of unknown origin, named wa352. This ORF led to male sterility when it was transgenically expressed in the nucleus and its product addressed to mitochondria. Altogether these results were considered su cient to validate the role of this ORF in WA-CMS [89]. Another example is the identi cation of the orf108 sterilizing gene associated with mori CMS in Brassica juncea. Longer atp1 transcripts in the CMS line were detected. The atp1 region was sequenced and the 5’ and 3’ mRNA ends were identi ed in fertile, CMS and restored lines. A new orf was detected, which was expressed in the CMS and restored lines but had a modi ed expression in the restored background [69]. Transgenic A. thaliana plants expressing orf108 were sterile, supporting the orf108 role in sterility [90]. Moreover, mutations in orf108 blocked the M. arvensis restoration of fertility, suggesting that the restorer could not interact with its cytoplasmic partner anymore and thus further supporting orf108 ’s role in mori CMS [91]. Finally, one more example is the BT-type CMS, for which an additional copy of the atp6 region containing a predicted ORF named orf79 was identi ed. It was proven that this ORF was transcribed and translated in the CMS line, and that the two restorers of fertility act on its expression either at the transcriptional or the translational level. Also, transgenic plants expressing the orf79 phenocopied the CMS. These pieces of evidence validated the role of orf79 in BT-type CMS [92].
Gametophytic and sporophytic cytoplasmic male sterilities and their restoration
As I mentioned in section 1.1.3, male gametophyte development is tightly regulated, and it requires the cooperation of gametophytic and sporophytic tissues. In gametophytic CMS the defects occur in the male gametophytes. In sporophytic CMS, the tapetal cells are the primary targets of the sterility factor. The genetic behavior of restoration can allow to distinguish between these two cases, since a plant that is heterozygous for the restorer allele Rf, called Rf /rf, does not segregate Rf and rf similarly in gametophytic and sporophytic CMS. In sporophytic CMS, the presence of one Rf allele is su cient to rescue the male fertility of all pollen grains whatever their allele at the restorer locus. The Rf /rf plant thus presents 100 % of viable pollen. The Rf and rf alleles segregate according to Mendel’s laws of genetic inheritance, and so one fourth of this Rf /rf plant’s o spring will be male sterile. Now, in gametophytic CMS the presence of one Rf allele rescues the fertility only in the gametophytes carrying it. Therefore, heterozygous Rf /rf plants present half dead pollen. The rf allele is never transmitted via the pollen, but it is transmitted via the egg cell according to Mendel’s laws, and so a segregation bias is expected with half of the progenies that are homozygous for the restorer allele Rf /Rf, and the other half that are heterozygous Rf /rf. No male sterile progeny is observed. The inheritance patterns of alleles at the restorer locus in the two di erent CMS systems is represented in gure 1.7. As the two di erent types of CMS are widespread, and as the two tissues on which a defect occurs are very di erent, it is not known whether the same functional mechanisms lead to pollen abortion in gametophytic and sporophytic CMS.
Energy production through oxidative phosphorylation
The eukaryotic cell is highly compartmentalized, and each compartment ful lls a spe-ci c function. Mitochondria are often de ned as the \powerhouse » of the cell as they are the main producers of Adenosine Tri Phosphate (ATP), which is the energy currency for the cell, together with chloroplasts. Yet, it is noteworthy to mention that ATP is also produced in the cytosol via glycolysis, the metabolic pathway by which glucose is converted into pyruvate while producing two molecules of ATP. Nevertheless, in non-photosynthetic cells, the ATP production necessary for all cell functions relies mainly on mitochondria. Mitochondria have a complex internal structure. They are compartmen-talized by two membranes, the outer and inner membranes. The space between the two membranes is called the inter-membrane space. Inside the inner membrane is found the matrix, that contains the mitochondrial DNA and has a high protein content. The two membranes have very di erent properties. The outer membrane controls the exchanges of molecules between the mitochondria and the cytosol. It is permeable to ions and small molecules through protein-based pores and carries the TOM complex that translocates mitochondrial proteins synthesized in the cytosol. It also carries the proteins involved in mitochondria dynamics thus controlling their shape and size [137]. The inner mem-brane is very impermeable and supports the transmembrane potential which drives ATP synthesis by the ATP synthase. It presents many invaginations that form the cristae, that increase the exchange surface. Moreover, it carries the oxidative phosphorylation (OXPHOS) system complexes, and its properties allow the maintainance of a transmem-brane potential. Most of the proteins encoded by the mitochondrial genome are part, or involved in the assembly, of OXPHOS system complexes.
In mitochondria, aerobic respiration is the main pathway to produce ATP [138]. The respiration process leads to the reduction of oxygen to water, the release of carbon dioxide and the phosphorylation of ADP to ATP. The main steps of the respiration process are the Krebs cycle, taking place in the mitochondrial matrix, and the oxidative phosphorylation. The Krebs cycle involves acetyl-coA oxidation. Acetyl-coA is notably produced from decarboxylation of the pyruvate produced during glycolysis. Glycolysis ocurs in the cytosol, but also in mitochondria, where it would directly provides pyruvate that will be used as a respiratory substrate [139]. The Krebs cycle reactions produces, inter alia, NADH and FADH2 molecules that will be oxidized by the respiratory complexes that compose the electron transport chain. The plant OXPHOS system is composed of ve complexes supplemented with proteins allowing bypasses pathways.
The complex I, or NADH-Ubiquinone oxidoreductase is composed of a mixture of nuclear and mitochondrial-encoded proteins so that genomes coordination must be tightly regulated in order to form a complex with the correct stoichiometry [140]. Complex I is the main entrance point of electrons into the respiratory chain: it transfers them to the ubiquinone using NADH as an electron source [141]. Concomitantly, it translocates four protons from the matrix to the inter-membrane space. Complex II, or succinate dehydrogenase, is the second entrance point of electrons, as it transfers electrons from succinate to ubiquinone without simultaneous protons translocation. Complex III, or cytochrome c reductase, transfers electron from the reduced ubiquinone (called ubiquinol) to cytochrome c. Simultaneously, four protons are translocated from the matrix to the inter-membrane space. Complex IV, or cytochrome c oxidase, is the terminal complex of the respiratory chain. It transfers the electrons from the cytochrome c to oxygen, simultaneously reduces the latter to water, and translocates two more protons from matrix to the inter-membrane space. The electron transport via respiratory chain results in a proton gradient across the inner membrane, which is the motor for complex V, or ATP synthase, to produce ATP. The OXPHOS system complexes can form supercomplexes [141], that might provide kinetic advantage and thus prevent reactive oxygen species (ROS) production [142].
ROS production in mitochondria and redox biology
At high concentrations, ROS are deleterious to the cell. Their major targets are nucleobases (DNA), disulfure bounds and methionine in proteins, and unsaturated fatty acids. At high concentration they are thought to induce cell death, when the rate of oxidative damage to cellular components exceeds their repair or replacement. Numerous lines of evidence indicate that mitochondria are involved in cell death in animals [145, 146]. In plants, the involvement of mitochondria and chloroplasts in triggering programmed cell death (PCD) is highlighted by increasing pieces of evidence. In mitochondria, pro-death signals involve ROS production, that trigger signals leading to the opening of mitochondrial permeability transition pore (MPTP), dissipation of the mitochondrial membrane potential and a burst of ROS production from the electron transport chain, dysfunction and nally disintegration of the organelle. Proteins normally contained in the inter membrane space and that are cell death signals, such as cytochrome c, are released in the cytosol [147]. Thus, continuous oxidation leads to cell death, but it is important to note that this e ect is cell type dependent [147]. Nevertheless, even though ROS are deleterious to the cell when they are produced in too high amount to be e ciently scavenged, they are also signaling molecules involved in many biological processes [148, 149]. They are early signals in many stress responses, and they are necessary for the cell to survive [150]. ROS signaling controls many cell processes, such as gene expression and metabolism [151]. It also mediates ower senescence, root architecture formation, polar cell growth in pollen tube and root hair cells [152], cell growth [153], and certainly many other processes that we have not identi ed so far. As mitochondria produce superoxide and hydrogen peroxide, but also are central actors in their control through e cient redox couples, they play a central role in such ROS signaling. It is not known whether ROS are directly signaling molecules on their own, or if they initiate a redox signaling cascade from their metabolism. Both can be possible, as oxidation of transcription factors have been reported, and a small shift in the glutathione redox potential is associated with changes in gene expression and plant development [154].
In addition, hydrogen peroxide also participates in many defense mechanisms against various stresses, via induction of defense and resistance genes [155]. ROS production are thought to protect cells from stress by triggering a redox signalling cascade that leads to the appropriate response in the appropriate compartment [154]. Mitochondria are involved in this stress response [138]. Perturbations of the mitochondrial electron transport chain can result in altered pathogen resistance, likely by their e ect on the redox status. For example, the Nicotiana sylvestris CMSII line has a de ciency in complex I and a resulting increase in mitochondrial ROS production, and this line has higher resistance to tobacco mosaic virus than the wild type [156]. Also, A. thaliana mutant in complex II subunit SDH1-1 show less mitochondrial ROS production and a higher susceptibility to several pathogens [156]. Finally, the slo2 mutant for which a reduction in complexes I, III and IV as well as an increased mitochondrial ROS production has been reported, is more susceptible to Botrytis cinerea [156]. A tight and accurate regulation of the cellular redox status is therefore needed to mediate pathogen attack response. Finally, the AOX, which is involved in mitochondrial regulation of ROS production, has also been reported to participate in plant defense [157].
Table of contents :
1 Introduction
1.1 Cytoplasmic male sterilities in the context of owering plant reproduction
1.1.1 Reproduction systems in owering plants
1.1.2 Gynodioecy and cytoplasmic male sterility
1.1.3 Male gametophyte development
1.2 Cytoplasmic male sterilities are the result of a cytonuclear incompatibility
1.2.1 The plant cell is composed of three genomes that must cooperate and have co-evolved
1.2.2 Genomic conict in cytoplasmic male sterilities
1.3 Cytoplasmic male sterility associated genes and nuclear restorers
1.3.1 Cytoplasmic genes associated to male sterility
1.3.2 The restoration of fertility
1.4 The plant mitochondria functions and their role in cytoplasmic male sterilities
1.4.1 Energy production through oxidative phosphorylation
1.4.2 Redox homeostasis in the regulation of cellular processes and response to stress: a central role for mitochondria
1.4.3 Mitochondria and the cell cycle
1.4.4 Main hypotheses to explain cytoplasmic male sterilities
1.5 The Sha cryptic cytoplasmic male sterility of A. thaliana
1.5.1 A gametophytic cytoplasmic male sterility system in A. thaliana .
1.5.2 The mitochondrial orf117Sha is strongly suspected to cause cytoplasmic male sterility
1.5.3 Objectives of the study
2 Results
2.1 Analysis of the orf117Sha gene expression in dierent nuclear backgrounds
2.1.1 orf117Sha mRNA accumulation in sterile and fertile lines, and in vegetative and reproductive tissues
2.1.2 orf117Sha transcript processing in the sterile and restored lines .
2.1.3 ORF117SHA protein accumulation in sterile and restored lines
2.2 Production of transgenic plants to phenocopy the CMS
2.2.1 Expression prole of the promoters in pollen development
2.2.2 Production and characterization of transgenic plants
2.3 Analysis of cytological events during abortion
2.3.1 Identication of the developmental stage at which the abortion starts
2.3.2 Verication of the male germ lineage fate
2.3.3 Mitochondria morphology during pollen development
2.4 Mitochondrial functions in sterile and fertile lines
2.4.1 In gel analysis of respiratory complexes
2.4.2 ATP production in sterile and fertile lines
2.4.3 Glutathione pool redox state in sterile and fertile lines
2.4.4 AOX response in sterile and fertile lines
3 Discussion
3.1 Is the orf117Sha the sterilizing gene?
3.1.1 Expression of the orf117Sha
3.1.2 Functional evidence for the role of orf117Sha in CMS
3.1.3 Concluding remarks on the role of the orf117Sha in male sterility
3.2 How does a mitochondrial gene cause pollen abortion in CMS?
3.2.1 Chronology of pollen death
3.2.2 Physiological modications in sterilizing mitochondria
3.2.3 Possible mechanisms causing the Sha cytoplasmic male sterility .
3.3 General conclusions and perspectives
4 Material and methods
4.1 Material
4.1.1 Plant material
4.1.2 Bacteria
4.1.3 Vectors
4.1.4 Antibodies
4.1.5 Oligonucleotides
4.2 Methods
4.2.1 Biological material culture
4.2.2 Nucleic acids analyses
4.2.3 Cloning and vector constructions
4.2.4 Biochemistry: proteins analysis
4.2.5 Cytological approaches
Bibliography
