Mitochondrial membrane lipidome defines yeast longevity

LCA enters yeast cells

In the high-throughput screen that led to the identification of LCA as an anti-aging molecule, all of the tested chemical compounds derived from several commercial libraries were dissolved in dimethyl sulfoxide (DMSO) and the final concentration of this solvent in yeast cultures was 1% (v/v) [44]. Due to the well-known ability of DMSO to create pores in the hydrophobic core of a membrane bilayer and to increase its fluidity, this amphiphilic enhancer of cell membrane permeability for drugs and DNA is traditionally used as a vehicle for delivering various chemical compounds into a cell [68]. If LCA was first dissolved in DMSO and then added to growth medium at a final concentration of 50 μM immediately following cell inoculation into the medium, this bile acid significantly extended the chronological lifespan (CLS) of yeast cells that were cultured under caloric restriction (CR) conditions on 0.2% glucose (Figure 1A; [44]).

Noteworthy, LCA at a final concentration of 50 μM displayed a similar strong beneficial effect on the CLS of yeast cells cultured under CR at 0.2% glucose if it was added to growth medium in water, i.e., under conditions that are unlikely to enhance cell membrane permeability for exogenous chemical compounds (Figure 1B; [37]). This observation suggested that, akin to the established mechanism underlying an anti-tumor effect of LCA in cultured human cells [69], a mechanism by which this bile acid extends longevity of chronologically aging yeast under CR conditions does not involve its delivery into cells. As a first step towards addressing this important spatial aspect of the longevity-extending effect of exogenously added LCA, we used a quantitative mass spectrometric analysis to compare the relative level of cell-associated LCA to that of LCA in the cultural medium. Our analysis of the age-related dynamics of changes in the levels of LCA associated with cells or remaining in the cultural medium revealed that in yeast cultures incubated with exogenously added 50 μM LCA in the presence of DMSO, only a minor portion of this bile acid (from 6.1% to 14.6% at different periods of chronological lifespan) was present in a cell-associated form (Figure 1C). In contrast, in yeast cultures incubated with exogenously added 50 μM LCA in the absence of DMSO, the major portion of this bile acid (from 76.7% to 94.8%, depending on a period of chronological lifespan) was associated with cells (Figure 1D).

We then used subcellular fractionation by differential centrifugation (Figure 1G) followed by mass spectrometric quantitation of LCA to assess its relative levels in various subcellular fractions. We found that in yeast cultured with exogenously added LCA in the presence of DMSO (i) LCA was present mostly (from 71.3% to 90.2%, depending on a period of chronological lifespan) in the cytosolic (100KgS) fraction; (ii) from 5.2% to 21.6% of LCA was recovered in the 12KgP fraction containing mitochondria, endoplasmic reticulum (ER), Golgi, vacuoles, plasma membrane (PM) and nuclei; and (iii) from 4.2% to 7.1% of LCA was associated with cell surface, as it was recovered in the 1KgP fraction known to consist of unspheroplasted cells and cell debris (Figure 1E).In contrast, in yeast cultured with exogenously added LCA in the absence of DMSO (i) LCA was mostly (from 73.3% to 92.2%) associated with cell surface (1KgP fraction); (ii) from 4.1% to 23.6% of LCA was recovered in the 12KgP fraction consisting of mitochondria, ER, Golgi, vacuoles, PM and nuclei; and (iii) only a minor portion (from 2.9% to 4.4%) of LCA was present in the cytosolic (100KgS) fraction (Figure 1F).

In sum, our quantitative analysis of LCA recovered in various subcellular fractions separated by differential centrifugation implies that in yeast cultured with exogenously added LCA in the presence of DMSO, this bile acid is evenly distributed between the cultural medium and the cytosol – perhaps because of its high solubility in DMSO present in the cultural medium and the cytosol and probably due to its passive diffusion through pores in the PM created by DMSO. Importantly, in yeast cultured with LCA in the presence of DMSO, up to 21.6% of the intracellular pool of this bile acid is confined to the 12KgP fraction containing mitochondria, ER, Golgi, vacuoles, PM and nuclei. Furthermore, in yeast cultured with exogenously added LCA in the absence of DMSO, this most hydrophobic molecular form of bile acids [44] associates mainly with cell surface - possibly because of its low solubility in water leading to its adhesion to the cell wall and perhaps due to lack of pores in the PM. Yet, even in yeast cultured with LCA in the absence of DMSO, up to 28% of this bile acid enters yeast cells and associates mainly with the 12KgP fraction consisting of mitochondria, ER, Golgi, vacuoles, PM and nuclei.

Intracellular LCA accumulates in mitochondria

To elucidate in which organelle or organelles present in the 12KgP subcellular fraction LCA resides, we subjected the mix of mitochondria, ER, Golgi, vacuoles, PM and nuclei recovered in this organellar fraction to separation by centrifugation to equilibrium in a sucrose density gradient. The shape of this gradient has been previously optimized for the purification of mitochondria devoid of contamination by other organelles present in the 12KgP fraction (Figure 2; [70]). Our mass spectrometric identification and quantitation of LCA in gradient fractions revealed that the 12KgP-associated pool of LCA is almost exclusively confined to mitochondria of yeast cultured with this bile acid in the presence of DMSO (Figures 2A and 2B) or in its absence (Figures 2C and 2D). We therefore concluded that, regardless of the presence of DMSO in yeast cultures containing exogenously added LCA, the only kind of cellular organelle this longevity-extending bile acid accumulates in is the mitochondrion.

Mitochondria-associated LCA resides mainly in the inner mitochondrial membrane (IMM)

To examine in which mitochondrial sub-compartment LCA resides, we first subjected purified mitochondria to fractionation using a swell-shrink procedure and subsequent equilibrium density gradient centrifugation. This fractionation approach enables to separate mitochondria into the intact outer mitochondrial membrane (OMM) fraction, the intact intermembrane space (IMS) fraction and the mitoplast fraction consisting of mitochondrial matrix surrounded by the intact IMM [71]. We found that, regardless of the presence of DMSO in yeast cultures containing exogenously added LCA, the bulk quantities of this bile acid (from 69.5% to 74.6% of the total pool of mitochondria-associated LCA) were confined to mitoplasts (Figures 3A and 3B). A smaller portion of LCA (from 20.7% to 26.6% of the total pool of mitochondrial LCA) was recovered in the OMM, and only minute quantities of it (from 1.4% to 8.3% of the total pool of mitochondria-associated LCA) were found in the IMS (Figures 3A and 3B).

We then quantitatively assessed the relative abundance of LCA in the membrane (i.e., the OMM and IMM) and soluble (i.e., the IMS and matrix) mitochondrial sub-compartments by subjecting purified mitochondria to fractionation using sonication and subsequent differential centrifugation. This fractionation approach results in a reconstruction of so-called submitochondrial particles (SMP). These vesicular particles (i) are depleted of protein and other components of the mitochondrial matrix and IMS; and (ii) are surrounded by the IMM and OMM resealed in the inside-out orientation [71]. High-speed centrifugation of sonicated mitochondria yields the pellet of SMP, whereas protein and other components of the mitochondrial matrix and IMS remain in the supernatant. We found that, in yeast cultured with LCA in the presence of DMSO or in its absence, this bile acid was confined mostly (from 92.8% to 98.5% of the total pool of mitochondrial LCA) to the pellet of SMP (Figures 3C and 3D). Minor quantities of LCA (from 1.5% to 7.2% of the total pool of mitochondria-associated LCA) were recovered in the supernatant consisting of mitochondrial matrix and the IMS (Figures 3C and 3D).

Altogether, our quantitative analysis of LCA recovered in various mitochondrial sub-compartments implies that, regardless of the presence of DMSO in yeast cultures containing exogenously added LCA, the mitochondria-associated pool of LCA resides mainly in the IMM. A smaller portion of this bile acid (from 20.7% to 26.6% of the total pool of mitochondrial LCA) associates with the OMM.

Mitochondrial membranes of yeast cultured in the presence of LCA exhibit altered concentrations of various glycerophospholipid species

Glycerophospholipid compositions of both mitochondrial membranes depend on several processes that are integrated into an intricate network governing lipid dynamics within mitochondria and the ER. These processes include (i) the synthesis of the glycerophospholipids phosphatidic acid (PA), cytidine diphosphate-diacylglycerol (CDP-DAG), phosphatidyl-serine (PS), phosphatidylcholine (PC) and phosphatidyl-inositol (PI) by enzymes residing in the ER; (ii) the synthesis of the glycerophospholipids CDP-DAG, phosphatidylglycerol (PG), cardiolipin (CL) and monolysocardiolipin (MLCL) by enzymes confined to the IMM; (iii) a bidirectional movement of glycerophospholipids via mitochondria-ER junctions (also called mitochondria-ER contact sites), which represent zones of close apposition between the OMM and the mitochondria-associated membrane (MAM) domain of the ER; and (iv) a CL-dependent inhibition of PA transport from the OMM to the IMM by Ups1p, a protein that shuttles PA between the two mitochondrial membranes (Figure S1) [18, 20-22]. Based on our observation that in yeast cells cultured in the presence of exogenous LCA this most hydrophobic molecular form of bile acids accumulates in the IMM and is also present in the OMM (Figure 3), we hypothesized that LCA may alter the relative concentrations of various glycerophospholipid species in mitochondrial membranes – perhaps by modulating activities of enzymes involved in glycerophospholipid synthesis within the IMM, affecting glycerophospholipid exchange between the mitochondrial and ER membranes via mitochondria-ER junctions, and/or impinging on the Ups1p-driven shuttling of PA between the two mitochondrial membranes.

To assess the validity of our hypothesis that mitochondria-confined LCA alters glycerophospholipid composition of mitochondrial membranes, we used mass spectrometry to compare the membrane lipidomes of mitochondria purified from yeast cultured under CR conditions with or without LCA, either in the presence of DMSO or in its absence. We found that, regardless of the presence of DMSO in yeast cultures containing exogenously added LCA, this bile acid causes an increase in the glycerophospholipid/protein ratio (calculated as nmoles of glycerophospholipid/mg of protein) of mitochondrial membranes. It should be emphasized that LCA affects this ratio is in an age-dependent manner. Indeed, LCA (i) had no effect on the glycerophospholipid/protein ratio of mitochondrial membranes in cells recovered at diauxic (D) growth phase on day 2 of cell culturing; (ii) caused a moderate (but significant) increase in this ratio in cells recovered at post-diauxic (PD) growth phase on day 4 of cell culturing; and (iii) significantly elevated the glycerophospholipid/protein ratio of mitochondrial membranes in cells recovered at stationary (ST) growth phase on day 7 of cell culturing, upon entry into a quiescent state (Figure 4).

It needs to be emphasized that, regardless of the presence of DMSO in yeast cultures containing exogenously added LCA, this bile acid exhibited differential effects on the concentrations of different molecular forms of mitochondrial membrane glycerophospholipids; moreover, these effects of LCA were age-dependent. Indeed, we found that (i) LCA causes a rise in the levels of mitochondrial PA, PG, PS, PC and PI calculated as nmoles of glycerophospholipid/mg of protein; (ii) the extent of such effect of LCA on PA, PG, PS, PC and PI levels (calculated as nmoles of glycerophospholipid/mg of protein) increases with the chronological age of yeast cells; and (iii) for the PG and PI species of mitochondrial membrane glycerophospho-lipids, this effect of LCA can be seen only in cells recovered at PD or ST growth phase on day 4 or 7 (respectively) of cell culturing (Figure 5).

In contrast, our mass spectrometric identification and quantitation of mitochondrial membrane glycerophospholipids revealed that (i) LCA elicits a decline in the levels of mitochondrial CL, MLCL and PE calculated as nmoles of glycerophospholipid/mg of protein; (ii) the extent of such effect of LCA on CL, MLCL and PE levels (calculated as nmoles of glycerophospholipid/mg of protein) decreases with the chronological age of yeast cells; and (iii) for the MLCL species of mitochondrial membrane glycerophospholipids, this effect of LCA can be seen only in cells recovered at PD or ST growth phase on day 4 or 7 (respectively) of cell culturing (Figure 5).

It should be stressed that, although a similar (but not identical) trend of the differential effect of LCA on the concentrations of different molecular forms of mitochondrial membrane glycerophospholipids was observed if their relative levels were calculated as mol% of all glycerophospholipids, the extent to which LCA increased the relative levels of mitochondrial PA, PG, PS and PC or decreased the relative levels of mitochondrial CL, MLCL and PE gradually progressed with the chronological age of yeast cells (Figure 6). Moreover, LCA did not alter the relative level of PI if it was calculated as mol% of all glycerophospholipids (Figure 6).

Importantly, while LCA elicited substantial differential effects on the relative levels of different molecular forms of mitochondrial membrane glycerophospholipids (calculated as mol% of all glycerophospholipids), it did not cause a significant change in the “unsaturation index” for any of the glycerophospholipid species (Figure S2). This index is calculated as the “glycerophospholipids with one, two, three or four unsaturated acyl chains (i.e., C_n:1, C_n:2, C_n:3 and C_n:4 species)/glycerophospholipids without unsaturated acyl chains (i.e., C_n:0 species)” ratio [72]. We found that the unsaturation index was high for each molecular form of mitochondrial membrane glycerophospholipids, ranging from 4.2 to 63.3 (Figure S2). For PG, PS, PC, PI and PE, the unsaturation index is known to play a pivotal role in defining whether each of these glycerophospholipid species acquires the bilayer forming shape of a cylinder or it attains the non-bilayer forming shape, either that of a cone or an inverted cone (Figures 7A-7D) [18, 73, 74]. On note, PA, CL and MLCL are known to be always present in the non-bilayer forming shape of a cone, regardless of their unsaturation indexes (Figures 7A-7D) [18, 73]. Thus, the observed differential effects of LCA on the relative levels of different molecular forms of mitochondrial membrane glycerophospholipids (Figure 6; calculated as mol% of all glycerophospholipids) and the demonstrated lack of its effect on the unsaturation index for each of these molecular forms (Figure S2) imply that this bile acid alters only the relative levels of bilayer forming and non-bilayer forming glycerophospholipid species, but does not affect the molecular shape of any of them.

Mitochondrial membranes of yeast cultured in the presence of LCA display reduced concentrations of non-bilayer forming glycerophospholipids and elevated concentrations of their bilayer forming species

The shape of a glycerophospholipid molecule is defined by the relative sizes of the cross-sectional areas of its hydrophilic head group and hydrophobic acyl chains (Figures 7A-7C) [18, 73, 74]. If the cross-sectional areas of the hydrophilic head group and hydrophobic acyl chains of a glycerophospholipid are equally sized, it has a cylindrical shape (Figure 7C) [18, 73, 74]. A glycerophospholipid whose hydrophilic head group is smaller than the cross-sectional area of the hydrophobic acyl chains has a cone shape, whereas a glycerophospholipid that exhibits the opposite trend is shaped as an inverted cone (Figure 7C) [18, 73, 74].

The relative levels of cylinder-, cone- and inverted cone-shaped glycerophospholipids in a membrane are known to define membrane curvature; thus, changes in the relative levels of the differently shaped lipids within a membrane can alter its curvature [18, 73, 74]. For the IMM, a rise in the relative levels of glycerophospholipids having the bilayer forming shape of a cylinder reduces the extent of membrane curving [18, 73, 74]. This, in turn, causes the following three kinds of simultaneous changes in mitochondrial morphology: (i) an increase in the abundance of the IMM domains exhibiting so-called “flat” bilayer conformation; (ii) a reduction in the abundance of the IMM domains displaying negative curvature (i.e., membrane curving away from the mitochondrial matrix) characteristic of mitochondrial contact sites between the IMM and OMM; and (iii) a decline in the abundance of the IMM domains having positive curvature (i.e., membrane curving towards the mitochondrial matrix) characteristic of mitochondrial cristae formed by the IMM (Figure 7D) [18, 73, 74]. Furthermore, the concurrent rise in the relative levels of cone- and inverted cone-shaped glycerophospholipids, both being non-bilayer forming species, elevates the extent of membrane curving to increase the abundance of mitochondrial contact sites and mitochondrial cristae (Figure 7D) [18, 73, 74].

Based on our data on mass spectrometric identification and quantitation of mitochondrial membrane glycerophospholipids, we calculated the relative levels of their species having the non-bilayer forming shape of a cone or an inverted cone as well as the relative levels of glycerophospholipid species exhibiting the bilayer forming shape of a cylinder. The relative levels of non-bilayer forming and bilayer forming glycerophospholipids were calculated as mol% of all membrane glycerophospholipids. We found that, regardless of the presence of DMSO in yeast cultures containing exogenously added LCA, this bile acid reduces the relative levels of non-bilayer forming glycerophospholipids (Figures 7E and 7F) and elevates the relative levels of their bilayer forming species (Figures 7G and 7H). We therefore hypothesized that LCA may (i) reduce the abundance of the IMM domains displaying negative curvature characteristic of mitochondrial contact sites between the IMM and OMM; (ii) decrease the abundance of the IMM domains exhibiting positive curvature typical of mitochondrial cristae formed by the IMM; and (iii) increase the abundance of the IMM domains having flat bilayer conformation.

Mitochondria of yeast cultured in the presence of LCA are enlarged, their number is reduced and their morphology is altered

Membrane lipids are known to influence the physical properties of biological membranes, thereby having a significant impact on their structure and function [18, 72, 75, 76]. As for mitochondrial membranes, their glycerophospholipid compositions and the relative abundance of several glycerophospholipid species within the IMM and OMM have been shown to define the morphology of mitochondria [18, 77-79]. Because LCA causes an age-related increase in the glycerophospholipid/protein ratio of mitochondrial membranes and differentially affects (in an age-dependent manner) the relative levels of different molecular forms of glycerophospholipids within these membranes, we sought to investigate how LCA influences mitochondrial morphology and abundance in chronologically aging yeast. Our electron microscopical analysis revealed that in yeast cells grown under CR on 0.2% glucose and recovered upon entry into a quiescent state on day 7 of cell culturing, LCA (i) causes a substantial enlargement of mitochondria (Figures 8C and 8I); and (ii) significantly reduces the number of these organelles (Figures 8D and 8J). Noteworthy, LCA elicited these characteristic changes in mitochondrial size and number regardless of the presence of DMSO in yeast cultures containing exogenously added LCA (compare Figures 8A-8D to Figures 8G-8J). It is likely that the substantial enlargement of mitochondria and the resulting expansion of both mitochondrial membranes observed in yeast cultured with exogenously added LCA (Figures 8A-8C and Figures 8G-8I) were due to the age-related increase in the glycerophospholipid/protein ratio (and, thus, in glycerophospholipid abundance) of mitochondrial membranes seen in these cells (Figure 4). Moreover, it is conceivable that the significant decrease in mitochondrial number in yeast grown in the presence of exogenous LCA (Figures 8A, 8B, 8D, 8G, 8H and 8J) was caused by the differential effects of this bile acid on the relative levels of different glycerophospholipid species (calculated as mol% of all glycerophospholipids) within mitochondrial membranes (Figure 6). Indeed, LCA increased the relative level of mitochondrial PA (Figure 6), a glycerophospholipid known to elicit a reduction in mitochondrial number by stimulating fusion of small mitochondria [18, 80, 81].

In yeast cells grown under CR on 0.2% glucose and recovered upon entry into a quiescent state on day 7 of cell culturing, LCA also alters the morphological appearance of the IMM and mitochondrial cristae. Indeed, our electron microscopical analysis revealed that this bile acid (i) substantially lowers the proportion of mitochondria with cristae extending from the IMM by reducing the extent of connectivity between cristae and the inner boundary membrane (Figures 8A, 8B, 8E, 8G, 8H and 8K); and (ii) significantly increases the total length of mitochondrial cristae (including both cristae extending from the IMM and cristae disconnected from the inner boundary membrane) relative to the total length of the OMM (Figures 8A, 8B, 8F, 8G, 8H and 8L). The formation of mitochondrial cristae by the IMM domains having positive curvature (i.e., membrane curving towards the mitochondrial matrix) is known to require both glycerophospholipids having the non-bilayer forming shape of a cone and glycerophospholipids exhibiting the non-bilayer forming shape of an inverted cone (Figure 7D) [18, 73, 74]. Therefore, it is conceivable that the substantial changes in morphology of the IMM and mitochondrial cristae observed in yeast cultured with exogenously added LCA (Figures 8A, 8B, 8E-8H, 8K and 8L) were due to the significant reduction in the relative levels of non-bilayer forming (i.e., cone- and inverted cone-shaped) mitochondrial glycerophospholipids seen in these cells (Figures 7E and 7F). Moreover, it is plausible that the observed in yeast grown in the presence of exogenous LCA build-up within the mitochondrial matrix of cristae disconnected from the IMM and thus exhibiting flat bilayer conformation (Figures 8A, 8B, 8E-8H, 8K and 8L) was caused by the substantial rise in the relative levels of bilayer forming (i.e., cylinder-shaped) mitochondrial glycerophospholipids seen in these cells (Figures 7G and 7H).

Mitochondria of yeast cultured in the presence of LCA exhibit altered functionality

Because LCA causes major changes in mitochondrial morphology and abundance in chronologically aging yeast, we sought to examine its effect on the age-related dynamics of longevity-defining cellular processes that are confined to and regulated by mitochondria. We found that in yeast cells grown under CR on 0.2% glucose and recovered at different periods of chronological lifespan, LCA alters the age-related chronology of changes in vital mitochondrial processes, including (i) mitochondrial respiration; (ii) the maintenance of electrochemical potential across the IMM; and (iii) the establishment of a steady-state level of intracellular reactive oxygen species (ROS) known to be generated mainly as by-products of mitochondrial respiration [82]. In yeast cultured in the absence of exogenous LCA, with or without DMSO, the efficacies of these three processes were (i) greatly amplified when yeast entered D growth phase on day 2 of cell culturing; and (ii) sharply declined through the subsequent PD and ST growth phases (Figures 9A-9C and 9E-9G). In yeast cultured in the presence of exogenous LCA, with or without DMSO, the efficacies of all these processes were (i) increased to a much lesser extent during D phase than they were increased in yeast cultured in the absence of LCA; and (ii) reached a plateau in PD phase and remained mainly unchanged during the subsequent ST phase (Figures 9A-9C and 9E-9G). It needs to be emphasized that, following entry of yeast cells into a quiescent state (i.e., during ST phase), the efficacies of all three monitored mitochondrial processes in yeast cultured in the presence of exogenous LCA (with or without DMSO) significantly exceeded those in yeast cultured in its absence (Figures 9A-9C and 9E-9G).

Furthermore, in yeast grown under CR on 0.2% glucose and recovered at different periods of chronological lifespan, LCA altered the age-related dynamics of changes in the cellular level of ATP, which in yeast limited in calorie supply is known to be produced mainly in mitochondria [83]. Indeed, we found that during PD and ST phases the level of ATP in yeast cells cultured under CR conditions in the presence of exogenous LCA (with or without DMSO) considerably exceeds that detected in yeast cultured in its absence (Figures 9D and 9H).

Altogether, these findings imply that in chronologically aging yeast LCA alters the age-related chronology of four longevity-defining cellular processes confined to and regulated by mitochondria. It is conceivable that mitochondria in chronologically “old” cells cultured with exogenous LCA respire more efficiently, maintain higher electrochemical potential across the IMM, sustain greater steady-state concentration of intracellular ROS and preserve elevated cellular level of ATP (as compared to mitochondria in age-matched cells cultured without LCA; Figure 9) because of the substantial changes in morphology of the IMM and mitochondrial cristae (Figure 8), which in cells exposed to LCA are likely to be caused by the significant reduction in the relative levels of non-bilayer forming mitochondrial glycerophospholipids (Figure 7). Indeed, the non-bilayer forming glycerophospholipids PE and CL have been implicated in regulating mitochondrial respiration, membrane potential, ROS homeostasis and ATP synthesis, the vital processes driven by protein complexes that reside in the IMM and mitochondrial cristae [18, 77, 84].

Noteworthy, the observed ability of mitochondria in chronologically “old” cells cultured with exogenous LCA to sustain intracellular ROS at the steady-state level exceeding that in age-matched cells cultured without LCA (Figures 9C and 9G) is likely to play an important role in longevity extension by this bile acid. Indeed, recent findings imply that mitochondria-derived ROS can delay chronological aging in yeast if their steady-state cellular concentrations are maintained at a certain “optimal”, non-toxic level [9, 13, 41]. This “optimal”, sub-lethal level of mitochondria-derived ROS is insufficient to cause a substantial oxidative damage to cellular constituents but can stimulate “stress-response hormesis” by activating an intricate signaling network that in response specifically alters epigenetic and gene expression patterns to cause lifespan extension [10, 11, 13, 50, 82, 85].

Yeast strains, media and growth conditions

The wild-type strain Saccharomyces cerevisiae BY4742 (MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0) [Thermo Scientific/Open Biosystems; #YSC1054] was grown in YP medium (1% yeast extract, 2% peptone) [both from Fisher Scientific; #BP1422-2 and #BP1420-2, respectively] containing 0.2% glucose [#D16-10; Fisher Scientific] as carbon source. Cells were cultured at 30°C with rotational shaking at 200 rpm in Erlenmeyer flasks at a “flask volume/medium volume” ratio of 5:1.

Pharmacological manipulation of chronological lifespan

Chronological lifespan assay and pharma-cological manipulation of chronological lifespan by addition of lithocholic acid (LCA) [Sigma; #L6250] were performed as previously described [37, 44]. The stock solution of LCA in dimethyl sulfoxide (DMSO) was made on the day of adding this compound to cell cultures. LCA was added to growth medium in DMSO or water at the final concentration of 50 μM immediately following cell inoculation into the medium. The final concentration of DMSO in yeast cultures supplemented with LCA (and in the corresponding control cultures supplemented with compound vehicle) was 1% (v/v).

Quantitative analysis of LCA associated with cells or remaining in the cultural medium

Yeast cells cultured in the presence of LCA were pelleted by centrifugation for 5 min at 3,000 × g at room temperature. The recovered supernatant of the cultural medium was stored at room temperature. The recovered pellet of cells was washed twice with distilled water; the supernatant recovered after each centrifugation was combined with the cultural medium. An established procedure [90] was used for lipid extraction from the washed cells and from the cultural medium combined with the two washes. Mass spectrometric identification and quantitation of LCA in the extracts of lipids was performed as previously described [37].

Quantitative analysis of LCA recovered in subcellular fractions separated by differential centrifugation

Subcellular fractionation of yeast cell homogenates through sequential centrifugation steps of increasing force and duration [91], lipid extraction from the recovered subcellular fractions [90], and mass spectrometric identification and quantitation of LCA in the extracts of lipids [37] were performed according to established procedures.

Quantitative analysis of LCA recovered in organelles pelleted by centrifugation at 12,000 × g and then separated by equilibrium density gradient centrifugation

A mix of mitochondria, endoplasmic reticulum, Golgi, vacuoles, plasma membrane and nuclei was recovered in a 12,000 × g pellet by sequential centrifugation of yeast cell homogenates at 1,000 × g and 12,000 × g as previously described [91]. An established procedure [70] was used for separating the recovered mix of organelles by centrifugation to equilibrium in a sucrose density gradient. Lipid extraction from equal volumes of gradient fractions [90], and mass spectrometric identification and quantitation of LCA in the extracts of lipids [37] were performed according to established procedures.

Subfractionation of purified mitochondria using a swell-shrink procedure and equilibrium density gradient centrifugation

Mitochondria purified as previously described [70] were subjected to subfractionation using a modified swell-shrink procedure and subsequent equilibrium density gradient centrifugation [71]. Briefly, the purified mitochondria were resuspended in hypotonic buffer KH (10 mM KCl, 2 mM HEPES, pH 7.2) at 2 mg/ml with gentle agitation for 20 min on ice. This caused the mitochondria to swell, ruptured the outer mitochondrial membrane (OMM) and resulted in the formation of mitoplasts consisting of mitochondrial matrix surrounded by the intact inner mitochondrial membrane (IMM). One-third volume of hypertonic buffer SAMH (1.8 M sucrose, 2 mM ATP, 2 mM MgSO₄, 2 mM HEPES, pH 7.2) was then added and the mitoplasts attached to ruptured OMM vesicles were subjected to gentle homogenization for an additional 10 min. This caused the mitoplasts to shrink, disrupted the OMM/IMM contact sites and resulted in the release of ruptured OMM vesicles. The swollen/shrunk mitochondrial suspension was then subjected to centrifugation on a discontinuous sucrose (26, 34.2, 45.2 and 61.6%; wt/v in KH buffer) gradient at 75,000 × g for 3 h at 4°C in a SW28 rotor (Beckman Coulter). 19 fractions of 2 ml each were collected. Equal volumes of gradient fractions were subjected to SDS-PAGE and immunoblotting with antibodies to Por1p (a protein marker of the OMM fraction), Ccp1p (a protein marker of the intermembrane space [IMS] fraction), Cox2p (a protein marker of the IMM fraction) and Mge1p (a protein marker of the matrix fraction). Lipid extraction from equal volumes of gradient fractions [90], and mass spectrometric identification and quantitation of LCA in the extracts of lipids [37] were performed according to established procedures.

Subfractionation of purified mitochondria using sonication and differential centrifugation

Mitochondria purified as previously described [70] were subjected to subfractionation using a modified procedure of sonication and subsequent differential centrifugation [71]. The procedure results in a reconstruction of so-called submitochondrial particles (SMP) depleted of protein and other components of the mitochondrial matrix and IMS; these vesicular particles are surrounded by the IMM and OMM that are resealed in the inside-out orientation [71]. To reconstruct SMP, the purified mitochondria were gently resuspended in buffer SEH (250 mM sucrose, 2 mM EGTA, 5 mM HEPES, pH 7.2) at 2 mg/ml on ice. The suspension of mitochondria was sonicated 3 × 1 min on ice with 1 min intervals. The sonicated suspension was then subjected to centrifugation for 10 min at 12,000 × g at 4°C to remove unbroken mitochondria. The recovered supernatant was spun for 45 min at 120,000 × g at 4°C to yield the pellet of SMP and the supernatant containing protein and other components of the mitochondrial matrix and IMS. The pellet of SMP was gently resuspended in ice-cold SEH buffer and spun for 45 min at 120,000 × g at 4°C to wash and re-pellet SMP.

Miscellaneous procedures

Electron microscopy and the morphometric analysis of the resulting images were performed as previously described [92]. Protein concentration in samples of purified mitochondria was determined with an RC DC protein assay kit (#500-0122; Bio-Rad) following the manufacturer's instructions. SDS-PAGE and immunoblotting using a Trans-Blot SD semi-dry electrophoretic transfer system (#170-3940; Bio-Rad) were performed according to established procedures [93]. Blots were decorated with monoclonal antibodies raised against Por1p (#459500; Invitrogen), Nsp1p (#ab4641; Abcam), Dpm1p (ab113686; Abcam), Vps10p (#A21274; Invitrogen), Pma1p (ab4645; Abcam), Pho8p (ab113688; Abcam) or polyclonal antisera raised against Cox2p (a kind gift of Dr. Carla Koehler, University of California, Los Angeles) or Ccp1p (a kind gift of Dr. Ann English, Concordia University, Montreal). Antigen-antibody complexes were detected by enhanced chemilumines-cence using an Amersham ECL Plus Western Blotting Detection Reagents (#RPN2132; GE Healthcare). Extraction of lipids from purified mitochondria and following mass spectrometric identification and quantitation of various lipid species were performed according to established procedures [37]. The “unsaturation index” for each molecular form of glycerophospholipids was calculated as previously described [72]. For the PA, PG, PS, PE, PC and PI species of glycerophospholipids, the unsaturation index was calculated as the “glycerophospholipids with one or two unsaturated acyl chains (i.e., C_n:1 and C_n:2 species)/glycerophospholipids without unsaturated acyl chains (i.e., C_n:0 species)” ratio. For the MLCL species of glycerophospholipids, the unsaturation index was calculated as the “glycerophospholipids with one, two or three unsaturated acyl chains (i.e., C_n:1, C_n:2 and C_n:3 species)/glycerophospholipids without unsaturated acyl chains (i.e., C_n:0 species)” ratio. For the CL species of glycerophospholipids, the unsaturation index was calculated as the “glycerophospholipids with one, two, three or four unsaturated acyl chains (i.e., C_n:1, C_n:2, C_n:3 and C_n:4 species)/glycerophospholipids without unsaturated acyl chains (i.e., C_n:0 species)” ratio. Cellular respiration assay [41], monitoring of the mitochondrial membrane potential [41] and ROS measurement [94] were performed according to established procedures. Preparation of cellular extracts and a microanalytic biochemical assay for measuring ATP have been described elsewhere [95].