A Heritable Structural Alteration of the Yeast Mitochondrion

Corresponding author: Daniel Lockshon, Box 357350, University of Washington, Seattle, WA 98195., lockshon{at}u.washington.edu (E-mail)

Communicating editor: J. RINE

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Prions have revived interest in hereditary change that is due to change in cellular structure. How pervasive is structural inheritance and what are its mechanisms? Described here is the initial characterization of [Leu^P], a heritable structural change of the mitochondrion of Saccharomyces cerevisiae that often but not always accompanies the loss of all or part of the mitochondrial genome. Three phenotypes are reported in [Leu^P] vs. [Leu⁺] strains: twofold slower growth, threefold slower growth in the absence of leucine, and a marked delocalization of nuclear-encoded protein destined for the mitochondrion. Introduction of mitochondria from a [Leu⁺] strain by cytoduction can convert a [Leu^P] strain to [Leu⁺] and vice versa. Evidence against the Mendelian inheritance of the trait is presented. The incomplete dominance of [Leu^P] and [Leu⁺] and the failure of HSP104 deletion to have any effect suggest that the trait is not specified by a prion but instead represents a new class of heritable structural change.

HERITABLE change that results from modification of nonnuclear cellular structure has been documented in a number of biological systems (for reviews, see NANNEY 1968 ; FRANKEL 1989 ; GRIMES and AUFDERHEIDE 1991 ; WICKNER and CHERNOFF 1999 ). This was emphatically demonstrated in Paramecium by the 180° inversion of several rows of cilia. BEISSON and SONNEBORN 1965 showed this change in structure of the cellular cortex to be a heritable change that was completely independent of any type of nuclear change. Other less widely noted studies have established cortical inheritance to be a general property of the ciliated protozoa [see, for example, the work of TARTAR 1961 on Stentor and NG and FRANKEL 1977 on Tetrahymena].

Heritable structural change has received renewed interest with the discovery of prions in mammals (PRUSINER 1982 ) and in fungi (WICKNER 1994 ; COUSTOU et al. 1997 ). [PSI⁺], one of two extensively characterized yeast prions [discovered by COX 1965 and reviewed by LIEBMAN and DERKATCH 1999 ], is specified by one of two conformational states of Sup35p. In its nonprion conformation, [psi^-], Sup35p plays an essential role in translation termination (ZHOURAVLEVA et al. 1995 ). [PSI⁺] is heritable and dominant to [psi^-] because the [PSI⁺] conformers of Sup35p appear to recruit nonprion Sup35p subunits to the prion conformation, producing large multimeric aggregates (PATINO et al. 1996 ; PAUSHKIN et al. 1996 ). In its aggregated form Sup35p is presumed to be inactive in translation termination. Thus, as first proposed by WICKNER 1994 , one heritable state, [PSI⁺], exists when Sup35p adopts the prion conformation and a second heritable state, [psi^-], exists when Sup35p adopts its other conformation. Two structural states of Sup35p are thus manifested as two alternate translational phenotypes, which, in turn, specify alternate sets of diverse colony morphology and growth phenotypes (TRUE and LINDQUIST 2000 ).

Prions, the first examples of structural inheritance understood in molecular terms, have stimulated the search for additional types of heritable change that are not specified by genetic or nuclear change. The mitochondrion is a likely place to look. That this organelle has never been seen to form de novo suggests that information intrinsic to its very structure may be required to specify its reproduction. While the mitochondrion itself is essential for viability (YAFFE 1999 ), the 80-kb mitochondrial genome of Saccharomyces cerevisiae, termed , can be completely (NAGLEY and LINNANE 1970 ) or partially (MOUNOLOU et al. 1966 ) lost to yield strains termed ^° and ^-, respectively. Because some of the proteins required for oxidative phosphorylation are encoded by the mitochondrial genome, the resulting "petite" strains are unable to grow on nonfermentable carbon sources (such as glycerol) yet have not been shown to be defective in any other metabolic process. I report here a non-Mendelian trait that often accompanies the loss of mitochondrial DNA (mtDNA). This trait, termed [Leu^P], is characterized by slow growth and an additional requirement for leucine. Data presented here demonstrate that [Leu^P] is the result of a structural alteration of the mitochondrion.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Materials, media, and strains:
YPD, YPG, and HC growth media have been described (ADAMS et al. 1997 ). Strains 2247-21-2a (from Lee Hartwell), 132 (from Bob Sclafani), and 10507 (from Peter Pryciak), the three ⁺ parent strains from which all the ^° and ^- strains described in this work were derived, are all congenic with A364A. YM608 was from Mark Johnston. The 11 mit strains used to genetically define the ^- strains described here were from Alexander Tzagoloff. YDL121 was constructed from 2247-21-2a by first creating a ura3 allele and then transforming to Ura⁺ with pCox4-GFP (plasmid from Ron Butow). BLU was made by integrating URA3LEU2 at leu2 in strain 132. hsp104 derivatives were produced by transforming 16kar/0 and DL124 to Ura⁺ with ClaI/HindIII-digested pYSU2 (CHERNOFF et al. 1995 ). ^° and ^- derivatives of ⁺ strains were obtained by growth overnight in YPD containing between 0 and 10 µg/ml ethidium bromide (EB) and plating onto YPD. DL035 and DL156 arose spontaneously from 10507. 16kar/0, DL110, and 16K31 were each derived directly from 10507 by growth in 3, 0.3, and 1 µg EB/ml YPD, respectively. DL106 and BLU/0 were obtained by growing BLU in 0.3 and 10 µg/ml EB, respectively. 3Z and 27Z were obtained after growth of YDL121 in 0 and 3 µg/ml EB, respectively. DL125 and DL123 were isolated as ^° and ⁺ cytoductants, respectively, from the cross BLU x 16kar/0. DL124 was obtained by growth of DL123 in 3 µg/ml EB. Homozygous diploid derivatives of strains 3Z and 27Z were obtained by first transforming each to Ura⁺ with the HO endonuclease-expressing plasmid pCY204 (from Dan Gottschling). Candidate spontaneously diploidized transformants were grown briefly on 5-fluoroorotic acid plates; nonmating, Ura^- derivatives whose cells were twofold larger than those of the parental strains were presumed to be homozygous diploids. A fusion between Cox4p and green fluorescent protein (GFP), which is localized to the mitochondrial inner membrane, was expressed from the plasmid pCox4-GFP. A fusion between Cit1 and GFP, which is localized to the mitochondrial matrix, was expressed from the plasmid pRS416/CS1-GFP (from Ron Butow; ZELENAYA-TROITSKAYA et al. 1998 ). Abf2p was overexpressed from the URA3 plasmid pGAL68-AFB (from Ron Butow; ZELENAYA-TROITSKAYA et al. 1998 ).

Methods:
Yeast was stained with 4',6-diamidino-2-phenylindole (DAPI) dihydrochloride by first fixing in methanol and then resuspending in 1 µg DAPI/ml water. GFP fusion proteins were visualized in living yeast cells, grown from single cells to microcolonies in situ as described by KOHLWEIN 2000 , using a Zeiss fluorescent microscope equipped with DeltaVision stage movement motors. Images of the strains were collected identically: 20 exposures of each field were taken at 0.4-µm focal intervals. The exposures were first deconvoluted to eliminate signal that is out of focus and then superimposed.

Cytoductive mating was done by combining freshly grown strains on YPD for 3 hr, after which an aliquot of mating mixture, resuspended in water, was spotted onto YPD. Only trefoil-shaped zygotes (bearing medial buds) were chosen for isolation by micromanipulation. Buds forming over the next several hours were isolated until each zygote yielded at most six buds. Normal-growing derivatives of strain 3Z and 16kar/0 were obtained by plating a freshly grown stationary culture grown in HC liquid medium onto HC plates lacking leucine to give 9 ± 3 x 10⁵ colonies/100-mm plate. Determination of mtDNA sequences retained in the ^- strains was done by genetic complementation as follows: a ^- strain was mixed with one of each of five or six relevant mit strains (whose names begin with M and aM in Table 1) as droplets of freshly grown cultures and allowed to mate overnight on a YPD plate. Cells from each mating were resuspended in water and spotted onto YPG plates, where growth of the diploid demonstrated complementation while failure to grow demonstrated noncomplementation.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Loss of mtDNA yields strains with either of two patterns of growth:
While studying the effect of a nuclear gene on yeast mtDNA (LOCKSHON et al. 1995 ), a single ⁺ parent strain was found to give rise to two types of ^° strains. For example, the ^° strain DL156 and its ⁺ progenitor (strain 10507) grew equally well on complete synthetic media whereas DL035, another independent ^° derivative that also arose spontaneously from the same ⁺ parent, grew somewhat more slowly (Fig 1A). All slow-growing ^° derivatives grew even more slowly in the absence of leucine, yet all normal-growing ^° derivatives had no leucine requirement for optimal growth (Fig 1B).

View larger version (181K): In this window In a new window Download PPT slide   Figure 1. Growth properties of a + strain and its ° and - derivatives. The + strain 10507 and its ° (DL156 and DL035) and - (16K31 and DL110) derivatives were streaked onto two plates from a single batch of HC minus leucine plates to which a standard amount of leucine was (A) or was not (B) added. Photos were taken after 3 days of growth.

S. cerevisiae is also capable of losing only part of its mitochondrial genome to yield so-called ^- respiratory-deficient strains (DUJON 1981 ). Since the mass of mtDNA is the same in ^- strains and their ⁺ parents (NAGLEY and LINNANE 1972 ), ^- and ^° derivatives can be easily distinguished by DAPI staining. As is the case with ^° derivatives of 10507, two types of ^- derivatives were also readily isolated: those such as DL110 that grow slowly and have a partial leucine requirement and those such as 16K31 that show no growth defect and no leucine requirement (Fig 1A and Fig B). The fraction of spontaneous ^° derivatives of 10507 that demonstrated this pair of growth phenotypes was high: 2160 colonies grown from an untreated culture gave 26 ^° and 3 ^- derivatives of which 12 and 2, respectively, were slow growing and leucine dependent. Overexpression of Abf2p, which is known to cause the loss of mtDNA (ZELENAYA-TROITSKAYA et al. 1998 ), also yielded both normal-growing and leucine-dependent ^° derivatives of strain 10507, as did ethidium bromide treatment. Thus, slow-growing, leucine-dependent derivatives of strain 10507 arose independently of the manner in which mtDNA was lost.

Two other ⁺ strains were also examined for their ability to yield slow-growing, partially leucine-dependent ^° and ^- derivatives. Loss of mtDNA directly from the ⁺ strain BLU yielded only normal-growing ^° derivatives, such as BLU/0. Slow-growing, leucine-dependent ^- derivatives such as DL106 could also be readily isolated directly from BLU, yet normal-growing ^- derivatives of BLU were not found. The ⁺ strain YDL121 readily yielded ^° and ^- derivatives, which exhibited both normal growth and slow growth that was leucine dependent. The growth properties of two such ^° strains, 3Z and 27Z, were examined in more detail. In liquid synthetic medium 3Z grew 2.1 ± 0.2-fold slower than 27Z in the presence of leucine and 3.1 ± 0.3-fold slower than 27Z in its absence. Addition of up to 8-fold more leucine to complete synthetic medium did not enhance the growth rate of 3Z, suggesting a more global cellular defect. Removal of no single amino acid (or uracil or adenine) other than leucine (aside from His and Trp, two auxotrophies) affected the growth rate of 3Z. However, growth in minimal medium caused some reduction in growth rate of 3Z relative to that seen in complete medium, whether this comparison was done in the presence or absence of leucine. The ⁺ parent (YDL121) and the normal-growing ^° strain (27Z) grew at the same rate in minimal vs. complete medium (data not shown).

In summary, three ⁺ strains yielded ^° and ^- derivatives that either grew as well as their ⁺ parents did or grew slowly and showed an additional partial requirement for leucine. This pair of growth defects is hereafter referred to as the Leu^P phenotype (partial). What is the nature of the heritable change that led to Leu^P? Since a single ⁺ parent yielded derivatives completely lacking mtDNA that were either normal growing or leucine dependent (cf. DL156 and DL035), the heritable change could not be due to alteration of the mitochondrial genome. On the other hand, because removal of mtDNA might be expected to influence organellar structure and because part of leucine biosynthesis occurs in the yeast mitochondrion (RYAN et al. 1973 ), Leu^P could be caused by an alteration in structure of the mitochondrion itself. Alternatively, Leu^P could be due to a heritable structural change in some nonmitochondrial cellular component. Nuclear mutation must also be considered. Data addressing these hypotheses are presented below.

Analysis of meiotic products of a cross between a ⁺ strain and a ^° Leu^P strain:
Two ⁺ strains (10507 and YDL121) gave a far greater fraction of ^° derivatives that were Leu^P than would be expected by nuclear mutation. Moreover, it is difficult to understand how the biosynthetic (or utilization) pathway of only leucine could be affected by such mutations. Nevertheless, this possibility was directly addressed by examining the segregation of Leu^P in meiotic products of a diploid made by crossing a Leu^{P °} strain (3Z) with a ⁺ "tester" strain (the dependence of sporulation on respiration required the tester strain to be ⁺). Twenty-six of 31 tetrads from this diploid yielded four respiratory-competent spores. None of the colonies from any of the spores were Leu^P. Since ⁺ strains that are Leu^P have not been found, the failure of any of the ⁺ meiotic products to exhibit Leu^P could be due to the masking of its expression by the ⁺ genome. It was therefore necessary to remove the mtDNA from spore colonies to attempt to reveal the Leu^P phenotype that would have been due to a hypothesized nuclear mutation in 3Z. However, the inability to distinguish whether removal of mtDNA from the spore colonies revealed Leu^P or caused it to arise de novo required the use of a different experimental strategy.

An additional diploid was therefore made by mating the normal-growing ^° strain 27Z (derived from the same ⁺ parent as was the Leu^{P °} strain 3Z) to the ⁺ tester strain. If the trait specifying Leu^P persisted through meiosis, as would be expected of a nuclear mutation, a higher incidence of Leu^P should have been seen in ^° derivatives of spores derived from the 3Z-containing diploid than in ^° derivatives of spores derived from the 27Z-containing diploid. The two types of diploids were sporulated and the 16 ⁺ strains representing 2 tetrads from each diploid were treated with EB to produce ^° derivatives. Of five randomly chosen ^° derivatives of each of the 16 strains, the frequency of Leu^P is similar for the two types of diploids (Table 2). In other words, the presence of Leu^P in a mating partner had little if any influence on the frequency of Leu^P in ^° derivatives of meiotic products. In summary, no evidence could be found for the persistence of Leu^P through meiosis, as would be expected of a trait specified by a nuclear mutation.

View this table: In this window In a new window   Table 2. Expression of the LeuP in ° derivatives of meiotic products of two types of diploids

Cytoductive transfer of Leu^P:
If the heritable change that causes Leu^P is not a nuclear change, then Leu^P must instead be specified by nonnuclear cellular material. Specification of the trait by the mitochondrion, moreover, predicts that transfer of the trait should accompany mitochondrial transfer. Nonnuclear components of yeast can be readily transferred ("cytoduced") between mating partners if one of them is kar1-1 since haploid nuclei seldom fuse within zygotes thus formed. Buds arising from such a zygote are therefore mainly of one haplotype or the other (CONDE and FINK 1976 ). A Leu^{P °} strain (DL125) was crossed to a normal-growing ^- kar1 strain (16K31). Zygotes and their buds were isolated and colonies grown from each bud were examined for nuclear markers. All 12 bud-derived strains that were found to have the same nucleus as the Leu^{P °} mating partner (DL125) were streaked onto plates lacking leucine (Fig 2). While 4 of these 12 cytoductants retained the Leu^P phenotype of the Leu^{P °} mating partner (DL125), the other 8 switched to normal growth (compare to BLU/0, an isogenic normal-growing ^° strain). The act of mating itself was not the cause of the conversion in growth properties: in control ^° x ^° cytoduction experiments (16kar/0 x BLU/0, four independent experiments, 82 zygotes), the 99 buds of one haplotype retained Leu^P and the 205 buds of the other haplotype retained the normal growth phenotype. Therefore, cytoductive transfer of some nonnuclear component from 16K31 often converted DL125 from a Leu^P strain to one that grows normally.

View larger version (200K): In this window In a new window Download PPT slide   Figure 2. Growth of strains derived from a cytoductive cross. After mating a normal-growing, - kar1 strain (16K31) with a LeuP ° strain (DL125), 27 zygotes were isolated. From these zygotes, 87 viable buds were isolated, 75 of which were of 16K31 and 12 of which were of DL125 nucleotype. All 75 strains of 16K31 nucleotype contained mtDNA and grew identically to 16K31 (data not shown). Cells from the colonies that grew from the 12 buds of the DL125 nucleotype were picked directly off the dissection plate and streaked onto a plate lacking leucine. These were then photographed after 3 days of growth. The number indicates the zygote from which the bud was dissected and the following letter indicates the order in which the bud was formed. Liquid cultures, each grown directly from cells of a dissection plate colony, were DAPI stained and the percentage of cells containing mtDNA is shown.

Which nonnuclear component, when cytoduced, caused the conversion of DL125 from Leu^P to normal growth? For each of the 12 strains, the percentage of cells that contained mtDNA, as determined by DAPI staining, is also presented (Fig 2). Cytoductants that contain mtDNA and must therefore have received mitochondria from the normal-growing mating partner became normal-growing themselves. Those cytoductants that received no mtDNA—and hence possibly no mitochondrial material whatsoever—from the mating partner, retained Leu^P. This experiment demonstrates not only that a Leu^P strain can be converted by cytoduction to one that is normal growing, but also that this conversion strictly correlates with the transfer of mtDNA from the normal-growing mating partner.

In the converse experiment, I sought to determine whether a normal-growing strain could be converted by cytoduction to one that is Leu^P. A normal-growing kar1 ^° strain (27Z) was mated to a Leu^{P -} strain (DL106), zygotes and buds were isolated, and bud colonies were nucleotyped and DAPI stained as described above. In two independent experiments, 99 zygotes yielded a total of 275 viable buds of which 225 had the nucleotype of 27Z, the normal-growing parental strain. Two of the colonies that grew directly from the 27Z-nucleotype buds grew far more slowly than did the other 223 colonies of this nucleotype. None of the normal-growing colonies had mtDNA by DAPI staining. However, one of the slow-growing colonies, derived from bud 33-4C, stably maintained mtDNA and stably displayed the Leu^P phenotype (data not shown).

The other slow-growing colony, derived from bud 21-6B, was grown a minimal amount in YPD liquid media for DAPI staining and replating. About half of the cells of this culture contained mtDNA (data not shown). Plating of this liquid culture of 21-6B onto YPD gave mostly small colonies (Fig 3A). Three of the small colonies (i, ii, and iii) and one large one (iv) from Fig 3A were then restreaked onto plates lacking or containing leucine. The growth of a large fraction of colonies that grew from the cells of colony i were Leu^P yet all colonies that grew from cells of iv had no leucine requirement (Fig 3B). Colonies ii and iii behaved the same as did colony i (data not shown). Cells grown from colonies i–iv of Fig 3A were assayed for mtDNA as well: between 25 and 50% of the cells grown from colonies i, ii, and iii had mtDNA; cells grown from colony iv were devoid of mtDNA. In summary, it was possible, at a low frequency, to convert a normal-growing ^° strain into one that is Leu^{P -} by cytoduction. As most clearly demonstrated by bud 21-6B, the retention of the Leu^P phenotype strongly correlates with retention of mtDNA and, therefore, presumably of mitochondria, from the ^- mating partner. The ability to cytoduce the Leu^P phenotype into a normal-growing strain (Fig 3) and to cytoduce normal growth into a Leu^P strain (Fig 2) supports the hypothesis that a mitochondrial structural change is the basis of Leu^P.

View larger version (120K): In this window In a new window Download PPT slide   Figure 3. Growth of the strain derived from bud 21-6B. A LeuP - strain (DL106) was mated to a normal-growing ° kar1 strain (27Z). (A) The liquid culture grown directly from the colony derived from bud 21-6B was streaked onto a YPD plate, which was grown for 3 days and photographed. (B) Colonies i and iv in A were each restreaked onto synthetic plates lacking or containing leucine, grown for 3 days, and photographed. In the i, +leucine panel several slightly larger colonies were interspersed among the others after only 2 days of growth, but by 3 days the smaller colonies reached the size of the larger ones.

Is Leu^P specified by a prion?
The only heretofore known heritable structural changes in yeast are specified by prions. What aspects of the heritable trait that specifies Leu^P are shared with prions? Since prions are fully dominant (WICKNER and CHERNOFF 1999 ), the interaction of the traits that specify Leu^P and normal growth was investigated. The growth of six diploid strains was examined in the absence (Fig 4A) and presence (Fig 4B) of leucine. These strains were made by crossing either normal growing (Fig 4, DL124; c, e, and g) or Leu^P (Fig 4, 16kar/0; d, f, and h) ^° derivatives of one ⁺ strain with a second ⁺ strain (Fig 4, BLU; c and d) or its normal-growing (Fig 4, BLU/0; e and f) or Leu^P (Fig 4, DL125; g and h) ^° derivatives. Note the large difference in growth rates between the normal-growing (Fig 4, DL124) and Leu^P (16kar/0) haploid control strains (Fig 4, a vs. b). When each type of ^° strain was crossed to a ⁺ strain (BLU) to form ⁺ diploids, no detectable difference in growth rate was seen (Fig 4C vs. d).

View larger version (96K): In this window In a new window Download PPT slide   Figure 4. Dominance of LeuP. Six diploids, formed by mating either DL124 (which grows normally) or 16kar/0 (which is LeuP) with three strains [BLU (+), BLU/0 (a normal-growing ° strain), or DL125 (which is LeuP)], were isolated by growing mating mixtures on double-dropout plates. After >30 generations of growth as diploids, they were streaked onto plates lacking (A) or containing (B) leucine, along with DL124 and 16kar/0, the two haploid controls. After 3 days of growth the plates were photographed.

When the normal-growing (DL124) and Leu^P (16kar/0) ^° strains were mated instead to a normal-growing ^° strain (BLU/0), a difference in growth between these two ^° diploids was apparent (Fig 4E vs. f), yet this difference was smaller than the difference between the haploids (Fig 4, a vs. b). Such incomplete dominance of the Leu^P phenotype with respect to the normal growth phenotype was seen in all cases when comparing the growth of diploids formed by mating a normal-growing ^° to normal-growing vs. Leu^{P °} strains (BLU/0 x 3Z vs. BLU/0 x 27Z; DL124 x DL125 vs. DL124 x BLU/0; data not shown). Sectors g vs. h (Fig 4) further demonstrate the incomplete dominance of Leu^P: crossing a Leu^{P °} strain (DL125) to a normal-growing (DL124) vs. a Leu^P (16kar/0) ^° strain produced diploids with a smaller difference in growth compared to that seen between the haploids (Fig 4, compare a vs. b with g vs. h). This trend was borne out in all comparisons between diploids made by mating a Leu^{P °} strain with normal-growing vs. Leu^P isogenic ^° strains (DL125 x 3Z vs. DL125 x 27Z; 16kar/0 x DL125 vs. 16kar/0 x BLU/0; data not shown). As a control, the magnitude of the difference in growth rate between Leu^P x Leu^P vs. normal-growing homozygous diploids (3Z x 3Z vs. 27Z x 27Z), as judged by colony size on plates lacking leucine, was the same as the difference between their haploid counterparts 3Z vs. 27Z (data not shown). In summary, Leu^P exhibited incomplete dominance when a comparison was made of the growth of diploids formed by every combination of the normal-growing and Leu^P strains described here. Furthermore, incomplete dominance was stable: after >100 generations of growth in the absence of leucine, the difference in colony size between the diploids shown in Fig 4, sectors e and f, was just as great as those seen in Fig 5 (data not shown).

View larger version (91K): In this window In a new window Download PPT slide   Figure 5. The degree to which Cox4p-GFP is localized to mitochondria in a + strain and its ° derivatives. Six strains were grown from single cells on coverslips and fluorescent microscopy was performed on the 30-cell living microcolonies. Shown are (a) the + parent strain YDL121, (b) the normal-growing ° derivative 27Z, (c) the LeuP ° derivative 3Z, and (d, e, and f) three independent normal-growing derivatives of 3Z.

Spontaneous conversion of yeast from the nonprion to prion state occurs at a low frequency but the opposite spontaneous transition has not been seen. I have not observed spontaneous conversion of ^° or ^- strains from normal growth to Leu^P, although the slow growth of Leu^P strains may prevent this conversion from being observed. On the other hand, conversion of Leu^P strains to normal growth was observed to occasionally occur and was therefore examined in detail. Plating of 3 x 10⁷ cells from a nonselective HC liquid culture of 3Z onto plates lacking leucine gave 12 fast-growing colonies whose growth properties on HC ± leucine and YPD plates were indistinguishable from those of 27Z. Since the assumption that these derivatives arose in the last generation of liquid growth is not necessarily valid, the observed frequency with which they arise (4 x 10^-7) is an upper limit of the actual frequency of their occurrence. Plating of 9.8 x 10⁶ cells of a second Leu^P strain, 16kar/0, onto HC without leucine gave 25 normal-growing colonies. The upper limit of the actual conversion frequency in this second genetic background is therefore 2.6 x 10^-6.

The maintenance of all known yeast prions is dependent on the chaperonin Hsp104p (CHERNOFF et al. 1995 ; DERKATCH et al. 1997 ; MORIYAMA et al. 2000 ; SONDHEIMER and LINDQUIST 2000 ; OSHEROVICH and WEISSMAN 2001 ; WEGRZYN et al. 2001 ). Deletion of HSP104 had no detectable effect on the growth properties of 16kar/0 and DL124, Leu^P and normal growing (respectively) derivatives of strain 10507, in either the presence or the absence of leucine (data not shown). Thus, two features of Leu^P, its incomplete dominance over normal growth and its failure to be influenced by hsp104, are inconsistent with its specification by a prion. Though the ability of Leu^P to convert to normal growth is consistent with the prion model (if the prionic state were to specify normal growth), it is compatible with other models as well.

Visualization of mitochondria in the two types of ^° strains:
Direct cytological examination of mitochondria was made possible by the expression Cox4p-GFP, a fusion protein destined for the mitochondrial inner membrane, in the ⁺ strain (YDL121), its normal-growing (27Z) and Leu^P (3Z) ^° derivatives. Fig 5, a, b, and c, shows fluorescent images of 30-cell living microcolonies of the three strains. In the ⁺ parent (Fig 5A), note the tight localization of GFP to the mitochondria, which smoothly arc around the cortex of each cell. In the normal-growing ^° derivative (Fig 5B), although the pattern of the mitochondria is not as smooth as that seen in the ⁺ parent, GFP is still localized to the organelle; the cytoplasm is in most cases invisible. In contrast, the Leu^{P °} derivative (Fig 5C) shows a substantial amount of GFP that has failed to be localized to the mitochondrion. In multiple independent experiments no difference in shape between the mitochondria of normal-growing vs. Leu^P strains of a given nucleotype was discernible. The difference between the tight localization of GFP to the normal-growing ^° strain's mitochondria compared to the delocalized GFP in the Leu^P strain, however, was highly reproducible. The overall fluorescent intensity seen in microcolonies of normal-growing vs. Leu^P strains was never seen to differ appreciably. Results similar to those seen in Fig 5, a, b, and c, were obtained by expressing Cit1p-GFP, a fusion protein destined instead for the mitochondrial matrix, in these three strains (data not shown).

The ability of strain 3Z to switch to normal growth at a low frequency provided an opportunity to further examine the correlation between the growth and cytological phenotypes. Three of the 12 normal-growing derivatives of 3Z described above were examined by fluorescent microscopy (Fig 5D, Fig E, and Fig F). All three converted to the high degree of GFP localization seen in 27Z (see Fig 5B). The Leu^P growth phenotype is therefore well correlated with the failure of substantial amounts of two known mitochondrial proteins to fully localize to mitochondria. The identification of this cytological phenotype as an additional property of Leu^P strains further supports the hypothesis that Leu^P is caused by a structural alteration of the mitochondrion.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast strains lacking all or part of the mitochondrial genome have long been known to be incapable of growth on nonfermentable carbon sources. I describe here a second heritable change that often, but not always, accompanies the partial or complete loss of mtDNA. Affected strains grow slowly relative to the ⁺ parent and require leucine for optimal growth. These phenotypes appear to be the result of a heritable alteration in mitochondrial structure because:

1. Half of the spontaneous ^° derivatives from one ⁺ strain (strain 10507) exhibited Leu^P, an incidence far higher than could be explained by nuclear mutation.

2. The frequency of Leu^{P °} derivatives, relative to normal-growing ^° derivatives, was no greater in meiotic products from a diploid made using a Leu^{P °} mating partner than from a diploid made using a normal-growing ^° mating partner. Again, a nonnuclear basis of the trait specifying Leu^P is indicated.

3. A Leu^P strain was converted to normal growth when the mitochondrion of a normal-growing mating partner was transferred by cytoduction. Conversion from the Leu^P to the normal growth phenotype never occurred when there was no evidence for transfer of the mitochondrion from the normal-growing mating partner (Fig 2). Conversely, a normal-growing strain was converted by cytoduction to a Leu^P strain, but only when there was positive evidence for transfer of the mitochondrion from the Leu^P mating partner (Fig 3).

4. Fluorescence microscopy demonstrated a phenotypic difference between mitochondria of normal-growing and Leu^{P °} strains (Fig 5).

I conclude that ^° and ^- strains that grow normally compared to the ⁺ parent contain mitochondria that are structurally sound, whereas Leu^P strains contain mitochondria that are partially defective in function due to an as yet unidentified structural alteration of the organelle. I propose that the heritable trait of the Leu^{P °} and ^- strains be termed [Leu^P] (square brackets indicating cytoplasmic inheritance) and that of the normal-growing ^° and ^- strains be termed [Leu⁺].

The role played by the mitochondrial genome in the structure of the organelle is only starting to be understood. Deletion of any one of five proteins disrupts the shape of the yeast mitochondrion and each of the five proteins is required for mtDNA maintenance [Mgm1p (JONES and FANGMAN 1992 ; SHEPARD and YAFFE 1999 ; WONG et al. 2000 ), Mdm10p, Mdm12p (BERGER et al. 1997 ), Mmm1p (HOBBS et al. 2001 ), and Fzo1p (RAPAPORT et al. 1998 )]. It is therefore reasonable to expect that the partial or complete loss of mtDNA could cause a mitochondrial structural change, which at some frequency is stably propagated. The nature of the structural difference between the mitochondrion of [Leu^P] and [Leu⁺] strains is beyond the scope of this initial characterization.

The establishment of [Leu^P] as a nongenetic heritable change places it in a category previously occupied only by prions and the cortical inheritance phenomena extensively characterized in the ciliates (GRIMES and AUFDERHEIDE 1991 ) and more recently in a trypanosome (MOREIRA-LEITE et al. 2001 ). The elimination of the possibility that nucleic acid is the physical basis of [Leu^P], however, has a caveat: perhaps a yeast mitochondrial plasmid, the maintenance of which is independent of , can exist in sequence states that specify, alternately, [Leu^P] and [Leu⁺]. Such plasmids were first found in Neurospora (COLLINS et al. 1981 ) and later in other filamentous fungi (see BERTRAND 2000 for review). Since ^° yeast mitochondria lack gene expression machinery, such a postulated yeast mitochondrial plasmid would have to evoke the Leu^P and normal-growth phenotypes through the nucleus, a possibility that, although unlikely, cannot be ruled out.

The mitochondrion is intimately involved in leucine biosynthesis:

1. Since the Fe/S cluster required for function of Leu1p, a cytosolic enzyme, is synthesized in the mitochondrion and exported to the cytosol by Atm1p, atm1 confers leucine auxotrophy (KISPAL et al. 1999 ). The requirement of leucine for optimal growth of [Leu^P] strains could therefore be due to a failure of adequate amounts of Fe/S to reach the cytosol.

2. Leucine is required for optimal growth when a subunit of pyruvate dehydrogenase (Pda1p) is mutated. This enzyme, although not uniquely involved in leucine biosynthesis, is located in the mitochondrial matrix and also influences the stability of mtDNA (WENZEL et al. 1992 ).

3. The first unique step in leucine biosynthesis is catalyzed by -isopropylmalate synthase, an enzyme encoded by LEU4 (BAICHWAL et al. 1983 ) and LEU9 (CASALONE et al. 2000 ). leu4 is synthetic with each of leu5 (BAICHWAL et al. 1983 ), leu6, leu7, leu8 (DRAIN and SCHIMMEL 1988 ), and leu9 (CASALONE et al. 2000 ) for complete leucine auxotrophy. The proteins encoded by leu6, leu7, and leu8 have not yet been identified. The recent identification of Leu5p as the transporter of CoA (or a precursor thereof) into mitochondria (PROHL et al. 2001 ) explains the absolute requirement of leu4leu5 strains for leucine since the remaining -isopropylmalate synthase, encoded by LEU9 (CASALONE et al. 2000 ) and reported to be localized exclusively to the mitochondrion, requires higher mitochondrial CoA levels for function than can be achieved in leu5 strains (PROHL et al. 2001 ). Furthermore, the curious pet-like phenotype of leu5 strains (DRAIN and SCHIMMEL 1986 ) is also specified by leu6 (DRAIN and SCHIMMEL 1988 ), thus implicating leu6 as well in mitochondrial function.

The mitochondria of two yeast mating partners, which form a true diploid zygote, have been convincingly demonstrated to rapidly fuse to form a single reticulum: proteins that reside in the mitochondrial inner membrane (NUNNARI et al. 1997 ; OKAMOTO et al. 1998 ), outer membrane (OKAMOTO et al. 1998 ), and matrix (AZPIROZ and BUTOW 1993 ; OKAMOTO et al. 1998 ) from one partner of mating pairs were found to be distributed throughout the entire mitochondrion of zygotes well before the emergence of the first bud. My interpretation of the cytoduction data in Fig 2 and Fig 3 assumes that in the heterokaryotic zygotes formed in cytoductive matings, the mitochondria contributed by the two mating partners remain separate. This assumption is difficult to reconcile with such rapid protein homogenization. However, it appears that the true diploid zygotes examined in the protein mixing studies behave differently with respect to mitochondrial mixing than do the heterokaryotic kar1 x KAR1 zygotes used in cytoductions. Indeed, the first detailed study of the use of kar1 as a tool for shuttling mtDNA into different nuclear backgrounds found that in KAR1 ⁺ cap ^r oli ^r par ^r x kar1 ⁺ cap ^s oli ^s par ^s crosses (cap, oli, and par are mtDNA loci), there was a strong tendency of a bud of a given nucleotype to retain the mitochondrial genotype with which it was originally associated. In addition, that study showed kar1 to cause a low degree of recombination between mitochondrial markers (Table 6a of LANCASHIRE and MATTOON 1979 ). A second study yielded similar findings (SENA 1982 ). In contrast, rapid mixing of parental mitochondrial genomes—allowing substantial formation of nonparental mitochondrial genotypes by recombination—is well established to occur in true diploid zygotes (reviewed by DUJON 1981 ). The maintenance, in kar1 x KAR1 zygotes, of two distinct mitochondria, which are concluded here to exist in alternative structural states, is therefore plausible.

These earlier studies demonstrating coinheritance of nuclear and mitochondrial genotypes in kar1 crosses (LANCASHIRE and MATTOON 1979 ; SENA 1982 ) imply that in heterokaryotic zygotes each of the haploid nuclei remain tethered to their respective parental mitochondria, which also remain unfused. Such an interpretation explains why in the ^° [Leu⁺] x ^° [Leu^P] crosses discussed here (data not shown), no cytoduction of [Leu^P] was observed. In contrast, when one of the parents was instead ^- [Leu⁺] (Fig 2) or ^- [Leu^P] (Fig 3), the trait could be cytoduced from that strain to a ^° mating partner. This difference implies that the inheritance advantage of ^- mitochondria compared to ^° mitochondria in kar1 x KAR1 zygotes overrides the association of the parental mitochondrion with their respective haploid nuclei. The well-established ability for ^- genomes to be cytoduced into ^° hosts is also consistent with the preferential inheritance of ^- over ^° mitochondria. Moreover, ^- [Leu⁺] mitochondria are apparently far better at displacing ^° [Leu^P] mitochondria (Fig 2) than ^- [Leu^P] mitochondria are at displacing those that are ^° [Leu⁺] (Fig 3). This asymmetry suggests a lower degree of tethering between the nucleus and [Leu^P] vs. [Leu⁺] mitochondria. The coinheritance of nucleus and mitochondrion is also suggested by a completely different line of investigation: FISK and YAFFE 1997 identified three distinct classes of alleles of mdm1, which affect the inheritance of the nucleus, the mitochondrion, or both. The segregation of the two types of organelle therefore shares at least one component and more extensive overlap in segregation mechanisms is plausible. Protein mixing experiments using heterokaryotic zygotes would be quite informative.

In the first publication describing petite, small colonies were shown to give rise, on rare occasion, to what were termed "revertants" (EPHRUSSI et al. 1949 ). Such reversion was the subject of an entire subsequent article (EPHRUSSI and HOTTINGUER 1951 ) and was also discussed in detail in a monograph (EPHRUSSI 1953 ; reprinted in EPHRUSSI 1999 ). Since petite strains are now known to have lost mtDNA, their reversion to respiratory competence is impossible. Indeed, the respiratory capacity of yeast was not measured in those earliest studies of petite. In light of the present report, it is likely that two changes occurred in Ephrussi's strains to cause petite: the loss of mtDNA and the appearance of [Leu^P]. The reversion described by EPHRUSSI et al. 1949 appears to be analogous to the conversion of strains 3Z and 16kar/0 from [Leu^P] to [Leu⁺]. Finally, the description here, to my knowledge for the first time since 1953, of the existence of ^° and ^- derivatives that grow as well as do their ⁺ parents, is consistent with the preference of yeast for fermentative over respiratory growth (LAGUNAS 1986 ).

Two distinct aspects of [Leu^P] warrant examination at the molecular level. First, what accounts for the phenotypic differences between [Leu^P] and [Leu⁺] strains? The partial leucine requirement may be the result of a slowing of either the Leu1p-catalyzed or the Leu4p/Leu9p-catalyzed steps of leucine biosynthesis. Both steps involve transport across the mitochondrial membranes, a process that the data of Fig 5 suggest may be defective. Second, why is [Leu^P] heritable? That is, how is the structural difference between [Leu^P] and [Leu⁺] mitochondria propagated? The incomplete dominance of [Leu^P] and [Leu⁺] and their independence of Hsp104p suggest the prion model not to be applicable. On the other hand, the ability of [Leu^P] strains to be converted at a low frequency to those with normal growth properties is consistent with a prion model, but is also consistent with other models in which two states interconvert. One such model is the positive feedback loop first proposed by NOVICK and WEINER 1957 . Interconversion is inconsistent with models wherein [Leu^P] is caused by the loss of structural information. For example, the activities of at least two enzymes, which function exclusively in the mitochondrial matrix—Hsp60p (CHENG et al. 1990 ) and Yah1p (LANGE et al. 2000 )—are required for the production of additional active enzyme. The loss of proper localization of either of these proteins is therefore an irreversible change. Models of higher complexity that involve proteinaceous and/or membranous supramolecular assemblies capable of adopting alternate conformational states that correspond to [Leu^P] and [Leu⁺] can also be envisioned. If a complex situation of this type holds for [Leu^P], it may be necessary to identify new examples of nonprionic structural inheritance that are more amenable to molecular characterization before [Leu^P] is understood at that level.

	ACKNOWLEDGMENTS

I am grateful to Don Hawthorne for his encouragement and scientific advice and for performing the cytoduction experiment (unpublished) that led to this project. This work was initiated in the laboratory of Walt Fangman, to whom I am grateful for the introduction to mitochondria. I thank Stan Fields for supporting this work while I was in his lab. Marty Lee played a critical role early in the project by pointing out the difference between detective data and judge/jury data. Brian Kennedy graciously provided lab space and support for the final phase of this work. Thanks go to Barbara Garvik (Hartwell lab), Mark Johnston, Peter Pryciak, Bob Sclafani, and Alexander Tzagoloff for providing strains and to Ron Butow, Yuri Chernoff, and Dan Gottschling for providing plasmids. I thank Brian Kennedy, Margo Murphy, Dan Gottschling, Andrew Kirsh, Breck Byers, and Jim Thomas for making suggestions on the manuscript. This work was supported in part by award 9982929 from the National Science Foundation to Donald C. Hawthorne and D.L.

Manuscript received December 13, 2001; Accepted for publication May 7, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ADAMS, A., D. E. GOTTSCHLING, C. A. KAISER and T. STEARNS, 1997 Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

AZPIROZ, R. and R. A. BUTOW, 1993  Patterns of mitochondrial sorting in yeast zygotes. Mol. Biol. Cell 4:21-36.[Abstract]

BAICHWAL, V. R., T. S. CUNNINGHAM, P. R. GATZEK, and G. B. KOHLHAW, 1983  Leucine biosynthesis in yeast. Curr. Genet. 7:369-377.

BEISSON, J. and T. M. SONNEBORN, 1965  Cytoplasmic inheritance of the organization of the cell cortex in Paramecium aurelia.. Proc. Natl. Acad. Sci. USA 53:275-282.[Free Full Text]

BERGER, K. H., L. F. SOGO, and M. P. YAFFE, 1997  Mdm12p, a component required for mitochondrial inheritance that is conserved between budding and fission yeast. J. Cell Biol. 136:545-553.[Abstract/Free Full Text]

BERTRAND, H., 2000  Role of mitochondrial DNA in the senescence and hypovirulence of fungi and potential for plant disease control. Annu. Rev. Phytopathol. 38:397-422.[Medline]

CASALONE, E., C. BARBERIO, D. CAVALIERI, and M. POLSINELLI, 2000  Identification by functional analysis of the gene encoding alpha-isopropylmalate synthase II (LEU9) in Saccharomyces cerevisiae.. Yeast 16:539-545.[Medline]

CHENG, M. Y., F. U. HARTL, and A. L. HORWICH, 1990  The mitochondrial chaperonin hsp60 is required for its own assembly. Nature 348:455-458.[Medline]

CHERNOFF, Y. O., S. L. LINDQUIST, B. ONO, S. G. INGE-VECHTOMOV, and S. W. LIEBMAN, 1995  Role of the chaperone protein Hsp104 in propagation of the yeast prion-like factor. Science 268:880-884. [psi+].[Abstract/Free Full Text]

COLLINS, R. A., L. L. STOHL, M. D. COLE, and A. M. LAMBOWITZ, 1981  Characterization of a novel plasmid DNA found in mitochondria of N. crassa.. Cell 24:443-452.[Medline]

CONDE, J. and G. R. FINK, 1976  A mutant of Saccharomyces cerevisiae defective for nuclear fusion. Proc. Natl. Acad. Sci. USA 73:3651-3655.[Abstract/Free Full Text]

COUSTOU, V., C. DELEU, S. SAUPE, and J. BEGUERET, 1997  The protein product of the het-s heterokaryon incompatibility gene of the fungus Podospora anserina behaves as a prion analog. Proc. Natl. Acad. Sci. USA 94:9773-9778.[Abstract/Free Full Text]

COX, B. S., 1965  , a cytoplasmic suppressor of super-suppressor in yeast. Heredity 20:505-521.

DERKATCH, I. L., M. E. BRADLEY, P. ZHOU, Y. O. CHERNOFF, and S. W. LIEBMAN, 1997  Genetic and environmental factors affecting the de novo appearance of the [PSI+] prion in Saccharomyces cerevisiae.. Genetics 147:507-519.[Abstract]

DRAIN, P. and P. SCHIMMEL, 1986  Yeast LEU5 is a PET-like gene that is not essential for leucine biosynthesis. Mol. Gen. Genet. 204:397-403.[Medline]

DRAIN, P. and P. SCHIMMEL, 1988  Multiple new genes that determine activity for the first step of leucine biosynthesis in Saccharomyces cerevisiae.. Genetics 119:13-20.[Abstract/Free Full Text]

DUJON, B., 1981 Mitochondrial genetics and functions, pp. 505–635 in The Molecular Biology of the Yeast S. cerevisiae: Life Cycle and Inheritance, edited by J. N. STRATHERN, E. W. JONES and J. R. BROACH. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

EPHRUSSI, B., 1953 Nucleo-Cytoplasmic Relations in Micro-Organisms. Oxford University Press, New York.

EPHRUSSI, B., 1999 Nucleo-cytoplasmic relations in micro-organisms, pp. 39–94 in The Early Days of Yeast Genetics edited by M. N. HALL and P. LINDER. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

EPHRUSSI, B. and H. HOTTINGUER, 1951  Cytoplasmic constituents of heredity: on an unstable cell state in yeast. Cold Spring Harbor Symp. Quant. Biol. 16:75-85.

EPHRUSSI, B., H. HOTTINGUER, and A. M. CHIMENES, 1949  Action de l'acriflavine sur les levures. I. La mutation "petite colonie. Ann. Inst. Pasteur 76:351-364.

FISK, H. A. and M. P. YAFFE, 1997  Mutational analysis of Mdm1p function in nuclear and mitochondrial inheritance. J. Cell Biol. 138:485-494.[Abstract/Free Full Text]

FRANKEL, J., 1989 Pattern Formation. Ciliate Studies and Models. Oxford University Press, New York.

GRIMES, G. S., and K. J. AUFDERHEIDE, 1991 Cellular Aspects of Pattern Formation: The Problem of Assembly. Karger, Basel, Switzerland.

HOBBS, A. E., M. SRINIVASAN, J. M. MCCAFFERY, and R. E. JENSEN, 2001  Mmm1p, a mitochondrial outer membrane protein, is connected to mitochondrial DNA (mtDNA) nucleoids and required for mtDNA stability. J. Cell Biol. 152:401-410.[Abstract/Free Full Text]

JONES, B. A. and W. L. FANGMAN, 1992  Mitochondrial DNA maintenance in yeast requires a protein containing a region related to the GTP-binding domain of dynamin. Genes Dev. 6:380-389.[Abstract/Free Full Text]

KISPAL, G., P. CSERE, C. PROHL, and R. LILL, 1999  The mitochondrial proteins Atm1p and Nfs1p are essential for biogenesis of cytosolic Fe/S proteins. EMBO J. 18:3981-3989.[Medline]

KOHLWEIN, S. D., 2000  The beauty of the yeast: live cell microscopy at the limits of optical resolution. Microsc. Res. Tech. 51:511-529.[Medline]

LAGUNAS, R., 1986  Misconceptions about the energy metabolism of Saccharomyces cerevisiae.. Yeast 2:221-228.[Medline]

LANCASHIRE, W. E. and J. R. MATTOON, 1979  Cytoduction: a tool for mitochondrial genetic studies in yeast. Utilization of the nuclear-fusion mutation kar 1–1 for transfer of drug^r and mit genomes in Saccharomyces cerevisiae. Mol. Gen. Genet. 170:333-344.[Medline]

LANGE, H., A. KAUT, G. KISPAL, and R. LILL, 2000  A mitochondrial ferredoxin is essential for biogenesis of cellular iron-sulfur proteins. Proc. Natl. Acad. Sci. USA 97:1050-1055.[Abstract/Free Full Text]

LIEBMAN, S. W. and I. L. DERKATCH, 1999  The yeast [PSI+] prion: making sense of nonsense. J. Biol. Chem. 274:1181-1184.[Free Full Text]

LOCKSHON, D., S. G. ZWEIFEL, L. L. FREEMAN-COOK, H. E. LORIMER, and B. J. BREWER et al., 1995  A role for recombination junctions in the segregation of mitochondrial DNA in yeast. Cell 81:947-955.[Medline]

MOREIRA-LEITE, F. F., T. SHERWIN, L. KOHL, and K. GULL, 2001  A trypanosome structure involved in transmitting cytoplasmic information during cell division. Science 294:610-612.[Abstract/Free Full Text]

MORIYAMA, H., H. K. EDSKES, and R. B. WICKNER, 2000  prion propagation in Saccharomyces cerevisiae: requirement for chaperone Hsp104 and curing by overexpressed chaperone Ydj1p. Mol. Cell. Biol. 20:8916-8922. [URE3].[Abstract/Free Full Text]

MOUNOLOU, J. C., H. JAKOB, and P. P. SLONIMSKI, 1966  Mitochondrial DNA from yeast "petite" mutants: specific changes in buoyant density corresponding to different cytoplasmic mutations. Biochem. Biophys. Res. Commun. 24:218-224.[Medline]

NAGLEY, P. and A. W. LINNANE, 1970  Mitochondrial DNA deficient petite mutants of yeast. Biochem. Biophys. Res. Commun. 39:989-996.[Medline]

NAGLEY, P. and A. W. LINNANE, 1972  Biogenesis of mitochondria. XXI. Studies on the nature of the mitochondrial genome in yeast: the degenerative effects of ethidium bromide on mitochondrial genetic information in a respiratory competent strain. J. Mol. Biol. 66:181-193.[Medline]

NANNEY, D. L., 1968  Cortical patterns in cellular morphogenesis. Differences in cortical patterns in ciliates may be hereditary, but independent of genic differences. Science 160:496-502.[Free Full Text]

NG, S. F. and J. FRANKEL, 1977  180 degrees rotation of ciliary rows and its morphogenetic implications in Tetrahymena pyriformis. Proc. Natl. Acad. Sci. USA 74:1115-1119.[Abstract/Free Full Text]

NOVICK, A. and M. WEINER, 1957  Enzyme induction as an all-or-none phenomenon. Proc. Natl. Acad. Sci. USA 43:553-565.[Free Full Text]

NUNNARI, J., W. F. MARSHALL, A. STRAIGHT, A. J. MURRAY, and J. W. SEDAT et al., 1997  Mitochondrial transmission during mating in Saccharomyces cerevisiae is determined by mitochondrial fusion and fission and the intramitochondrial segregation of mitochondrial DNA. Mol. Biol. Cell 8:1233-1242.[Abstract]

OKAMOTO, K., P. S. PERLMAN, and R. A. BUTOW, 1998  The sorting of mitochondrial DNA and mitochondrial proteins in zygotes: preferential transmission of mitochondrial DNA to the medial bud. J. Cell Biol. 142:613-623.[Abstract/Free Full Text]

OSHEROVICH, L. Z. and J. S. WEISSMAN, 2001  Multiple Gln/Asn-rich prion domains confer susceptibility to induction of the yeast [PSI⁺] prion. Cell 106:183-194.[Medline]

PATINO, M. M., J. J. LIU, J. R. GLOVER, and S. L. LINDQUIST, 1996  Support for the prion hypothesis for inheritance of a phenotypic trait in yeast. Science 273:622-626.[Abstract]

PAUSHKIN, S. V., V. V. KUSHNIROV, V. N. SMIRNOV, and M. D. TER-AVANESYAN, 1996  Propagation of the yeast prion-like [psi+] determinant is mediated by oligomerization of the SUP35-encoded polypeptide chain release factor. EMBO J. 17:3127-3134.

PROHL, C., W. PELZER, K. DIEKERT, H. KMITA, and T. BEDEKOVICS et al., 2001  The yeast mitochondrial carrier Leu5p and its human homologue Graves' disease protein are required for accumulation of Coenzyme A in the matrix. Mol. Cell. Biol. 21:1089-1097.[Abstract/Free Full Text]

PRUSINER, S. B., 1982  Novel proteinaceous infectious particles cause scrapie. Science 216:136-144.[Abstract/Free Full Text]

RAPAPORT, D., M. BRUNNER, W. NEUPERT, and B. WESTERMANN, 1998  Fzo1p is a mitochondrial outer membrane protein essential for the biogenesis of functional mitochondria in Saccharomyces cerevisiae.. J. Biol. Chem. 273:20150-20155.[Abstract/Free Full Text]

RYAN, E. D., J. W. TRACY, and G. B. KOHLHAW, 1973  Subcellular localization of the leucine biosynthetic enzymes in yeast. J. Bacteriol. 116:222-225.[Abstract/Free Full Text]

SENA, E. P., 1982  Effects of the Kar gene on cytoplasmic mixing and mitochondrial genome suppressiveness, and consequences for cytoduction of petite DNA in S. cerevisiae.. Curr. Genet. 5:47-52.

SHEPARD, K. A. and M. P. YAFFE, 1999  The yeast dynamin-like protein, Mgm1p, functions on the mitochondrial outer membrane to mediate mitochondrial inheritance. J. Cell Biol. 144:711-720.[Abstract/Free Full Text]

SONDHEIMER, N. and S. L. LINDQUIST, 2000  Rnq1: an epigenetic modifier of protein function in yeast. Mol. Cell 5:163-172.[Medline]

TARTAR, V., 1961 The Biology of Stentor. Pergamon, New York.