Pitfalls in determination of petiteness
Our approach eliminated several drawbacks in the characterization of the petite forming ability. First, it is important to spot the symptoms of mtDNA elimination, such as the formation of petites or microcolonies introduced by Bulder (1964b). Some yeasts reported as resistant to the commonly used concentrations of EB were already classified as petite-negative (Šubík et al., 1974). However, cultivation with an elevated concentration of EB (200 μg mL−1) allows to determine petiteness in any characterized species. Obvious EB doses used were 25 μg mL−1 (Fox et al., 1991; Dunn et al., 2006) or 5–50 μg mL−1 (Piškur et al., 1998; Møller et al., 2001; Schneider-Berlin et al., 2005). However, Maleszka (1994) reported a concentration of 2.5 mM for C. parapsilosis (equals 1000 μg mL−1), which is five times higher than the maximal concentration used in our work.
Besides the potential to grow infinitely, it is important to determine the inability of petites to grow on nonfermentable substrates. The complete elimination of mtDNA could be determined simply by DAPI staining (Fig. 1d). Because of the nuclear petite mutants that are also pleiotropically deficient in cytochrome oxidase (Tzagoloff & Dieckmann, 1990), the DAPI approach is much more reliable than the traditional monitoring of cytochromes a+a3 (Bulder 1964a). Staining is also more convenient for rapid screening procedure than Southern blot hybridization (Fox et al., 1991) or purification of mtDNA by CsCl bisbenzimid gradient (Møller et al., 2001). Our screening procedure also allows us to determine the petite phenotype if the original strains (such as Kazachstania transvaalensis) do not grow, or grow extremely slowly, on the media with a nonfermentable carbon source (such as the Hanseniaspora clade). In spite of this inability, ρ0 mutants can be distinguished according to smaller colony size and the DAPI staining analysis.
Although the oldest data are difficult to trace due to the missing CBS numbers (Bulder, 1964a; de Deken, 1966), the identity of many species can be deduced according to their former designation in the CBS database. In general, the results from different sources correspond with our observations (Bulder, 1964a; de Deken, 1966; Piškur et al., 1998; Middelhoven & Kurtzman, 2003; Merico et al., 2007). A few minor variations may result from employment of different strains, minor contaminations, different cultivation conditions (especially temperature) or be due to the variable sensitivity to the intercalating agents.
Temperature sensitive petite phenotype
Initial experiments were carried out at 28°C, which is optimal for the growth of Saccharomyces. However, later on we tested species that were hardly capable of growing on YPD at this temperature (among them Kazachstania piceae, Kazachstania barnettii and others). To exclude the synergistic effect of the temperature on the petiteness we examined the ability to tolerate the loss of mtDNA after the cultivation with EB at lower temperatures (13, 18, 23°C). Indeed, Kazachstania africana, Kazachstania kunashirensis, N. delphensis, C. castellii, Vanderwaltozyma yarrowii and Zygotorulaspora florentinus, which were unequivocally petite negative at 28°C, generated petite colonies at temperatures of 23°C or lower. Species Kazachstania africana and C. castellii were capable of forming visible sectored colonies containing much more than thousands of cells at 28°C, but their proliferating ability is limited as they could not grow after the placement on the fresh YPD plates. With the exception of Zygotorulaspora florentinus (formerly Zygosaccharomyces florentinus), petite colonies from the aforementioned species grew at higher temperatures (28°C), implicating the role of temperature during the process of petite formation.
Transition of the ability to generate petites
We characterized petiteness of all species assigned to the Saccharomyces/Kluyveromyces complex (Fig. 2). Interestingly, the genera from postduplication lineages (Saccharomyces, Kazachstania, Naumovia, Nakaseomyces) are invariably petite-positive. Watson et al. (1980) reported petite-positivity of thermophilic enteric yeasts Tetrapisispora pintolopesii and Candida slooffii. These species were recently reclassified to the genus Kazachstania by Kurtzman et al. (2005), indicating that this clade consists entirely of species that tolerate the loss of mtDNA.
The lineages branching out just prior to and after the whole genome duplication are a mixture of petite-negative and petite-positive species (Fig. 2). Apparently, a clear ability to generate petites has been fixed in the lineage, which underwent the whole genome duplication but not in all Tetrapisispora species. However, the petite positive trait emerges sporadically also in the species from the preduplication lineages (Zygotorulaspora florentinus, Torulaspora globosa, Torulaspora pretoriensis, H. osmophila).
In pre- and postduplication yeasts the partially incapacitated ability to generate petites can sporadically be found, and they exhibit mixed ‘moot’ phenotype. Herein, besides respiring grande colonies and some petite colonies, the majority of the plated cells (50–99.9%) generate microcolonies. Mixed phenotype most often occurs in the Tetrapisispora clade (Fig. 2).
To ascertain whether a particular yeast species is a mixture of petite-negative and petite-positive variants, we examined petite phenotype in single cell cultures of Kazachstania lodderae, N. delphensis, Tetrapisispora nanseiensis and Kazachstania africana. Ten different cultures arising from distinct single colonies exhibited unchanged mixed petite phenotype, with the similar ratio of true petites and microcolonies indicating that this is a typical feature of the entire yeast population. Evidently, the characteristic of this phenotype is that the main part of the population does not tolerate the loss of mtDNA, but a significant portion, 1–50%, does. Therefore, it could be considered as a transition step from petite-negativity to petite-positivity, where cells are still not perfectly tuned for the life without the mitochondrial functions.
We assume that a similar effect was already spotted by Bulder (1964a), who also noticed that ‘petite colonies occurred rather infrequently’ in Schizosaccharomyces pombe, Brettanomyces lambicus (currently refers to Dekkera anomala) and Saccharomyces florentinus (currently likely refers to Zygotorulaspora florentinus).
Petite-positivity can be easily converted to petite-negativity through single mutation and vice versa (reviewed in Chen & Clark-Walker, 1999; Contamine & Picard, 2000; Dunn et al., 2006). Even though a single mutation can switch over petite phenotype, mutations that allow the transition of just a minor part of the yeast population (likewise in yeasts with moot phenotype) have not been reported yet.
The partial ability and unusual high frequency of petite formation distinguishes this class of yeasts from ‘petite susceptible’ species such as Schizosaccharomyces pombe and Kluyveromyces lactis. These yeasts are capable to convert to ρ0 variants due to single nuclear mutations. However, the frequency is extremely low (one to a few colonies per experiment) and requires long-term exposure (14–17 days) to EB (Haffter & Fox, 1992; Chen & Clark-Walker, 1995, 1999). Mutations converting Kluyveromyces lactis to petite-positive have been already identified in three largest subunits of the mitochondrial F1-ATPase, suggesting the essential nature of the functional ATPase in mitochondria, especially in the cells that have lost mtDNA. Apparently, two key features are important for the ability to survive without mtDNA. It is the capacity to provide enough energy from glycolysis, when the key energy source is disabled (Merico et al., 2007), and the capability of supporting mitochondrial biogenesis (Chen & Clark-Walker, 1999; Clark-Walker, 2003; Smith & Thorsness, 2005).
Functional protein import machinery is essential for the mitochondrial biogenesis and consequently for the viability of cells. Two energy supplies, membrane potential and ATP, are needed for mediating protein translocation across and insertion into the inner membrane (for reviews Mokranjac & Neupert, 2005; de Marcos-Lousa et al., 2006). Consequently, survival and growth of yeasts depends on the generation of a voltage gradient, across the mitochondrial inner membrane (St-Pierre et al., 2000; Clark-Walker, 2003; Schnaufer et al., 2005). However, in cells lacking mtDNA it can be generated only as a result of the electrogenic nature of an ADP/ATP translocator, an inner membrane protein exchanging the cytosolic ATP4− for mitochondrial ADP3− (for reviews see Pebay-Peyroula & Brandolin, 2004; Nury et al., 2006). Keeping a constant level of ADP inside the mitochondria by ATP hydrolysis is essential for the maintenance of membrane potential. Therefore, the activity of the F1-ATPase subunit is indispensable and can be tuned up by the mutations in F1 subunits (Chen & Clark-Walker, 1995; Clark-Walker, 2003; Smith & Thorsness, 2005) or by increasing the overall mitochondrial ATPase activity (Kominsky & Thorsness, 2000).
Unfortunately, the majority of the examined species are wild-type diploids and most of the ‘moot species’ are very likely homothalic (Butler et al., 2004; Piškur et al., 2006). This is the major setback for genetic experiments that can distinguish whether ‘moot’ phenotype is due to the mutation. Alternatively, it could be just a selection of a part of the population with tuned up maintenance of mitochondrial membrane potential (with elevated mitochondrial ATPase activity or increased capacity to transport cytosolic ATP into the mitochondria).
Apparently, the whole genome duplication is directly and indirectly connected to the petiteness. However, the whole genome duplication is not involved in the petite positive phenotype in some of the Brettanomyces/Dekkera species (Bulder, 1964a; Šubík et al., 1974; Hoeben et al., 1993; Woolfit et al., 2007). It is very likely that petite-positive species are interspread also among other unrelated yeasts, such as Candida albicans (Gyurko et al., 2000), but they are not known in filamentous fungi, algae or plants. On the other hand, tolerance to the loss of mtDNA has been reported in some protozoa (Schnaufer et al., 2002) and animal cell lines (Inoue et al., 1997). The remarkable feature is that more evolved cells from higher eukaryotes (mammals, avians) are petite-positive, if they grow in the cell culture, even though the organisms as a whole do not tolerate the elimination of mtDNA (Desjardins et al., 1986; King & Attardi, 1989). The distribution of petite-positivity appears to be thought-provoking paradox in evolution, as it is hard to understand why the majority of lower eukaryotes cannot tolerate mtDNA elimination, while human cells do. The mosaic distribution pattern of petite-positivity is mysterious because it is difficult to understand the advantage of petite positivity or negativity. Perhaps it could have been just an outcome of ‘tinkering’ in evolution (Jacob, 1977) associated with the fermentative life-style and therefore it emerges randomly but infrequently.