• Open Access

Acetyl-l-carnitine protects yeast cells from apoptosis and aging and inhibits mitochondrial fission


  • Vanessa Palermo,

    1. Department of Cell and Developmental Biology, Pasteur Institute-Cenci Bolognetti Foundation, University of Rome “La Sapienza” Piazzale Aldo Moro, Rome, Italy
    Search for more papers by this author
  • Claudio Falcone,

    1. Department of Cell and Developmental Biology, Pasteur Institute-Cenci Bolognetti Foundation, University of Rome “La Sapienza” Piazzale Aldo Moro, Rome, Italy
    Search for more papers by this author
  • Menotti Calvani,

    1. Scientific Department, Sigma-Tau, Via Pontina Km. Pomezia, Rome, Italy
    Search for more papers by this author
  • Cristina Mazzoni

    1. Department of Cell and Developmental Biology, Pasteur Institute-Cenci Bolognetti Foundation, University of Rome “La Sapienza” Piazzale Aldo Moro, Rome, Italy
    Search for more papers by this author

Cristina Mazzoni, Department of Cell and Developmental Biology, University of Rome “La Sapienza”, Piazzale A. Moro, 5 00185 Rome, Italy. Tel.: +(39-6) 499 12257; fax: +(39-6) 499 12256; e-mail: cristina.mazzoni@uniroma1.it


In this work we report that carnitines, in particular acetyl-l-carnitine (ALC), are able to prolong the chronological aging of yeast cells during the stationary phase. Lifespan extension is significantly reduced in yca1 mutants as well in rho0 strains, suggesting that the protective effects pass through the Yca1 caspase and mitochondrial functions. ALC can also prevent apoptosis in pro-apoptotic mutants, pointing to the importance of mitochondrial functions in regulating yeast apoptosis and aging. We also demonstrate that ALC attenuates mitochondrial fission in aged yeast cells, indicating a correlation between its protective effect and this process. Our findings suggest that ALC, used as therapeutic for stroke, myocardial infarction and neurodegenerative diseases, besides the well-known anti-oxidant effects, might exert protective effects also acting on mitochondrial morphology.


Genetic studies on lifespan determination revealed a large number of potential targets for chemical compounds to slow aging. Because aging is a main risk factor in human cancer or neurodegenerative diseases, such compounds may prove useful in retarding age-related diseases.

Reducing energy intake by controlled caloric restriction or intermittent fasting increases lifespan and protects various tissues against disease. Genetics has revealed that aging in several organisms may be controlled by changes in the intracellular NAD/NADH ratio regulating sirtuins, a group of proteins linked to aging, as well by changes in metabolism and stress tolerance (Calabrese et al., 2008).

It has been demonstrated that the neuroprotective effects of dietary antioxidants, including curcumin, acetyl-l-carnitine and carnosine, operate on the activation of these redox-sensitive intracellular pathways (Calabrese et al., 2008). In particular, acetyl-l-carnitine and polyphenols activate the transcription factor Nrf2 which, after binding to the ARE (antioxidant responsive element) in the HO-1 promoter, up-regulates both heme oxygenase (HO-1) and thioredoxin reductase (TrxR) genes thus counteracting pro-oxidant conditions.

Although the notion that stress proteins are neuroprotective is broadly accepted, still much work needs to be carried out to associate protection with specific patterns of stress response.

The complex regulatory network and the often contradictory results obtained with human cell lines made the development of an easier model system desirable.

A number of compounds have been shown to extend lifespan of simple eukaryotes, such as yeast, C. elegans, and Drosophila melanogaster, and then it was demonstrated their effectiveness in preventing aging also in higher eukaryotes. Yeast represents a basic model system, widely used because of its easy cultivation and manipulation, for the study of the complex mechanisms occurring in higher eukaryotes and, more recently, also for the study of replicative and chronological aging processes.

Replicative aging is defined by the number of daughter cells a mother can give birth to and, hence, reflects aging in proliferating cells. Differently, chronological aging is defined as the capability of maintaining viability of nondividing cells during the stationary phase and is widely accepted as a feature for postmitotic tissue aging.

In both kinds of aging, the production of ROS occurs, as well the stimulation of programmed death by necrosis and apoptosis. The inactivation of the apoptotic machinery or the reduction of oxidative radicals lead both to longevity suggesting that, in yeast, aging is closely connected to apoptosis.

Although in the yeast genome there are no genes similar to those of the mammalian Bax and Bcl family, it was shown that Bax-mediated cell death in S. cerevisiae is accompanied by the typical features of apoptosis, such as externalization of phosphatidylserine at the surface of the cytoplasmic membrane, cytochrome C release, membrane blebbing, chromatin condensation and margination, DNA cleavage, and it is known that the simultaneous expression of Bcl-xL prevents these effects and cell death (Ligr et al., 1998). Moreover, yeast analogues of some crucial components of the apoptotic cascade in mammals have also been described, i.e., caspase (YCA1) (Madeo et al., 2002), Omi (Fahrenkrog et al., 2004), AIF (Wissing et al., 2004), EndoG (Buttner et al., 2007), suggesting that the basic machinery of apoptosis is indeed present and functional also in unicellular organisms.

The first evidences of yeast apoptosis emerged from experiments performed with a mutant of CDC48, the gene encoding the ER-associated ATPase that participates in retrotranslocation of ubiquitinated proteins from the ER into the cytosol for degradation by the proteasome (Madeo et al., 1997). Subsequently, the homologs of CDC48 in mammals (p97/VCP) and in C. elegans (mac-1) have been characterized as anti-apoptotic factors, indicating that studies on yeast apoptosis can bring useful information on the general molecular mechanisms of apoptosis (Shirogane et al., 1999).

During last years, it was demonstrated the involvement of the metacaspase YCA1 as the executor of cell death in mutants showing apoptotic phenotypes (Mazzoni & Falcone, 2008).

YCA1 activation seems to be related to defects in many fundamental cellular processes such as DNA replication, mitochondrial function, RNA and protein stability.

Concerning mRNA stability, we previously reported that mutations in yeast genes involved in the mRNA decapping process, such as dcp1 and dcp2, and in several components of the Lsm1-7 complex, show apoptotic phenotypes and accelerated chronological aging, and we demonstrated that in lsm4Δ1 mutant, these phenotypes can be partially rescued by the inactivation of YCA1 (Mazzoni et al., 2003, 2005a,b).

We exploited the increased sensitivity of lsm4Δ1 to apoptosis-inducing agents for the isolation of suppressor genes of programmed cell death (Mazzoni et al., 2005a, 2009). This mutant is also an interesting tool for the screening of molecules with anti-apoptotic or anti-aging properties.

Carnitine, also known as l-carnitine or levocarnitine (natural l-stereoisomer), is a quaternary ammonium compound primarily biosynthesized in the liver and kidney of mammals from the amino acid lysine and methionine (Calvani et al., 2004).

In mammalian cells, carnitine has a key role in fatty acid metabolism; in fact, beta-oxidation of fatty acids is incomplete and acetyl-CoA, propionyl-CoA and medium chain acyl- CoA are transported as carnitine esters to mitochondria where they are further oxidized to CO2 and H2O (Reddy & Mannaerts, 1994).

In yeast, degradation of fatty acid takes place exclusively in peroxisomes, and the produced acetyl-CoA molecules are imported into mitochondria where they are further oxidized through two pathways.

In the first, acetyl-CoA enters the peroxisomal glyoxylate cycle by mean of the citrate synthase Cit2p. The acetyl group is then transferred to oxalacetate to form citrate that, once transformed to succinate, enters mitochondria through the mitochondrial succinate-fumarate transporter Acr1.

In the second pathway, the peroxisomal carnitine acetyltransferase (Cat2p) transfers the acetyl group from CoA to carnitine to form acetyl-carnitine which is subsequently transported into mitochondria through the carnitine acetyl-carnitine translocase Crc1p (van Roermund et al., 1995; Huizing et al., 1997).

Mitochondrial Cat2p catalyzes the reverse reaction to form carnitine and acetyl-CoA, which enters the tricarboxylic acid cycle (TCA) for energy production. This process is referred to as the carnitine shuttle that plays an important role when yeast cells are grown on nonfermentable carbon sources, such as acetate and ethanol. The metabolism of these compounds results in the production of acetyl- CoA in the cytoplasm, which needs to be transported to mitochondria for energy production (Schmalix & Bandlow, 1993; Stemple et al., 1998).

The role of carnitine in the metabolism of S. cerevisiae has not been investigated beyond the shuttling of acetyl residues. It has been recently reported that carnitine improves cell growth in the presence of H2O2 and of weak organic acids such as lactate, and that this effect appears to be independent of the carnitine shuttle (Franken et al., 2008).

Here, we report the effect of three carnitines (L-carnitine, acetyl-l-carnitine and propionyl- L-carnitine) on longevity in both wild-type, pro-apoptotic and anti-apoptotic mutant strains.

We also found that acetyl-l-carnitine promotes the maintenance of mitochondrial network during aging and can inhibit mitochondrial fission, pointing to this organelle as the mediator of lifespan during chronological aging.


Carnitines extend lifespan in yeast

After prolonged cultivation, wild-type yeast cells lose their viability through apoptotic pathways, a phenomenon referred as chronological aging (Fabrizio et al., 2001).

In this work, we wanted to test whether L-carnitine (LC), propionyl-l-carnitine (PLC) and acetyl-l-carnitine (ALC) could influence the lifespan of wild-type and mutant yeast strains.

As shown in Fig. 1A, the presence in the media of LC, PLC and ALC extended the lifespan of the wild-type strain CML39-11A with an increase of 21.4%, 42.8% and 62.4%, respectively, suggesting that the effectiveness of protection was higher for ALC, followed by PLC and LC. Then, we also tested the effect of these three molecules in a mutant (lsm4Δ1) in the mRNA decapping protein Lsm4p. We previously reported that most of these mutant cells show a rapid loss of viability in stationary phase associated with the onset of apoptosis following a caspase-dependent pathway (Mazzoni et al., 2005a,b).

Figure 1.

 Carnitines influence chronological aging. Cells of CML39-11A (wt, panel A), MCY4/313Kllsm4Δ1 (lsm4Δ1, panel B) and CML39-9A (yca1, panel C) were grown in SD medium supplemented with L-carnitine (LC), acetyl-l-carnitine (ALC) and propionyl-l-carnitine (PLC) at the final concentration of 1mM. Viability is expressed as the percentage of micro-colony-forming units (CFU). Standard deviation was obtained from three independent experiments.

As shown in Fig. 1B, LC, PLC and ALC-treated cells showed a lifespan increase of 180%, 120% and 220%, respectively, compared to the untreated cells, indicating that the effectiveness of carnitines still followed the order ALC > LC > PLC. The striking effect of carnitines in enhancing cell viability in this pro-apoptotic mutant suggests that the protection mechanisms of these molecules are in some way connected with the onset of programmed cell death.

To confirm this, we followed the chronological aging in a yeast mutant lacking YCA1, the gene encoding the yeast metacaspase. This mutant shows an apoptotic-resistant phenotype and prolonged lifespan compared to its isogenic wild-type strain (Madeo et al., 2002).

As shown in Fig. 1C, the extension of cell viability occurred in the presence of ALC and LC, but their effect was relatively modest compared to the one observed in the wild-type and in the pro-apoptotic mutant. In fact, the best increase of lifespan was only 18% in the case of ALC-treated cells. These results suggest that carnitines may act by inhibiting the caspase-dependent pathway of cell death or/and that the absence of YCA1 reduces the protective effects of carnitines.

Considered that ALC was the best protector molecule compared to LC and PLC, we used only this compound in all the subsequent experiments.

Acetyl-l-carnitine recovers the pleiotropic phenotypes of the lsm4Δ1 pro-apoptotic mutant

Beside the shortened lifespan, the lsm4Δ1 mutant shows pleiotropic phenotypes such as the inability to grow in respiratory carbon sources and the increased sensitivity to acetic acid and caffeine (Mazzoni & Falcone, 2001; Mazzoni et al., 2005a,b). Unexpectedly, we found that the addition of ALC to lsm4Δ1 cells restored all the wild-type phenotypes (Fig. 2). Interestingly, these phenotypes are also suppressed by the inactivation of YCA1, indicating that they are dependent on the activation of a caspase-dependent cell death pathway (Mazzoni et al., 2005a,b). These data suggest that ALC might act as an anti-apoptotic molecule preventing the activation of the Yca1 activity in yeast. According to this, it has been shown that in fibroblasts undergoing apoptosis following serum deprivation, the activity of caspase 3 is greatly reduced in the presence of ALC (Pillich et al., 2005).

Figure 2.

 Effect of ALC on caffeine and acetic acid sensitivity of the lsm4Δ1 mutant. Tenfold serial dilutions of liquid cultures of CML38-11A (wt) and MCY4/313Kllsm4 Δ 1 (lsm4Δ1) were spotted on YPD solid medium containing 0.25% caffeine, 70 mm acetic acid or YPY in the absence (noALC) or in the presence of 1 mm ALC (+ALC).

ALC protection requires functional mitochondria

Increasing evidence demonstrates that aging is closely associated with mitochondrial degeneration (Balaban et al., 2005). It has been reported that LC acts as an endogenous antioxidant and reduces the formation of hydrogen peroxide (Gulcin, 2006). Moreover, LC and ALC can restore mitochondrial function and protect cells against injury induced by a variety of toxic agents (Dhitavat et al., 2002; Mazzio et al., 2003; Binienda et al., 2005).

To better understand the contribution of mitochondria to the effects observed after ALC treatment, we measured the chronological aging in rho0 strains obtained from both the wild-type CML39-11A and the lsm4Δ1 mutant. We previously reported that these rho0 strains showed significantly slower kinetics of death, probably because of a growth delay and in a reduced production of stress factors (Mazzoni et al., 2005a,b).

As showed in Fig. 3, the addition of ALC to rho0 strains, compared to the corresponding rho+ counterparts, resulted in a modest increase of lifespan for the lsm4Δ1 mutant while, in the wild-type, the lifespan in the presence of ALC was even slightly reduced. These results suggest that ALC might act through the detoxification of oxidants produced within mitochondria during respiration.

Figure 3.

 Mitochondrial function is required for ALC-chronological lifespan extension. Chronological aging of CML39-11A (wt, panel A), MCY4/313Kllsm4Δ1 (lsm4Δ1, panel B) and their isogenic rho0 mutant strains without or with acetyl-l-carnitine(+ALC). Average and standard deviations, obtained from three independent experiments, are also indicated.

Growing cells on nonfermentable carbon sources, such as glycerol and ethanol, provides a direct way to study the effect of respiratory condition on CLS. As shown in Fig. 4, growing in 3% glycerol CML39-11A and BY4741 two different laboratory wild-type strains resulted in the extension of CLS of 50% and 140%, respectively, compared to 2% glucose. These data are consistent with a model predicting that some of the benefits on lifespan caused by caloric restriction (CR) in yeast cells are because of the metabolic shift from fermentation to aerobic respiration (Lin et al., 2002; Smith et al., 2007). As also shown in Fig. 4, the presence of ALC can further increase lifespan extension in glycerol-grown cells, even though at a lesser extent than in glucose-grown cells.

Figure 4.

 ALC-mediated lifespan extension is less efficient in respiratory conditions. Chronological of two wild-type laboratory strains, CML39-11A (A) and BY4742 (B) grown in minimal medium containing 3% glycerol as carbon source. The respiration rate is expressed as nanomoles of oxygen consumed per milliliter per minute. Average and standard deviations, obtained from three independent experiments, are also indicated.

ALC protects cells from oxidative stress

ALC is known to exert an anti-oxidant effect within cells and, for this reason, we tested the sensitivity of the two wild-type strains to hydrogen peroxide treatment in the presence of this carnitine. As shown in Fig. 5(A,C), ALC clearly protected both strains from cell death at H2O2 concentrations ranging from 0.8 to 3.0 mm, suggesting, as also reported recently (Franken et al., 2008), a protecting role of ALC against oxidative stress. To confirm this, we measured the percentage of ROS-positive cells during the exponential phase of growth and after treatment with 3 mm H2O2. As shown in Fig. 5(B,D), the presence of ALC lowered the number of ROS-positive cells under the tested conditions, witnessing the capability of ALC in acting as ROS scavenger.

Figure 5.

 Effect of ALC on H2O2-treated cells. Viability of the BY4741 (A) and CML39-11A (C) wild-type strains was measured after exposure of cells to H2O2 at the indicated concentration for 4 h without (black bars) or with (gray bars) 1 mm ALC. Panels B and D show the percentage of ROS-positive cells during the exponential phase and after treatment with 3 mm H2O2. Average and standard deviations, obtained from three independent experiments, are also indicated.

ALC prevents mitochondrial fragmentation

Mitochondrial fusion and fission are relevant mechanisms for maintaining and remodeling the morphology of these organelles (Hoppins et al., 2007). It is also known that, from yeast to mammals, mitochondria play a pivotal role in determining cell aging and apoptosis (Heath-Engel & Shore, 2006) in that they represent a storage site of pro-apoptogenic factors (cytochrome c, Aif, EndoG) that are released in the cytoplasm upon apoptotic stimuli. In addition, the mitochondrial tubular network is disrupted by such stimuli, as well as during cellular aging (Bossy-Wetzel et al., 2003).

To understand the role of mitochondria in ALC protection, we followed the chronological aging in mutants in genes involved in mitochondria morphology. In detail, we analyzed mutants lacking DNM1, the homolog of the mammalian dynamine–related GTPase Drp1; MDM30, which encodes a F-box protein; FIS1, the mitochondrial fission promoting gene; YME1, encoding a subunit of the mitochondrial inner membrane i-AAA protease complex, and the mitofusin gene FZO1. We previously reported that the absence of either DNM1, MDM30 and FIS1 leads to an extension of longevity while the yme1 mutant looses viability very quickly during aging (Palermo et al., 2007).

In Table 1 is reported the lifespan (LS) of such mutants obtained from three independent experiments in the absence (−) and in the presence (+) of 1mM ALC.

Table 1.   Longevity of wild-type and mutant isogenic strains in the presence (+) and absence (-) of 1 mm ALC. Lifespan (LS) represents the time (day) at which < 0.1% cells are still viable and was calculated from three independent experiments on cell populations growing at 28°C.
StrainALCLS (days)LS variation (gt)LS variation (+ALC/−ALC)LS variation + ALC (mutants vs. wt)
  1. Lifespan variations of mutants in the absence of ALC were compared to the wild-type to highlight the differences in longevity because of genotype (gt). Moreover, for each mutant, lifespan is also reported as the percent variation between ALC-treated and untreated cells (+ALC/-ALC) and as the percent variation respect between mutant and wild-type treated cells (+ALC; mutants vs. wt).

+18 +260% 
+23 +187%+27%
+16 +220%−12.5%
+24 +242%+30%
+27 +237%+50%
+25 +317%+39%

Compared to the wild-type, the dnm1 and mdm30 mutations, per se, conferred an increase of the lifespan of about 60% and, in the case of the fis1 mutation, of 40%. Although ALC prolongs lifespan in all mutants compared to the wt, the advantage of the dnm1 and fis1 mutants was about 27% and 30%, respectively, while the advantage of the mdm30 mutant was kept on 50%. In the case of the yme1 mutant, the presence of ALC could extend LS but at a lesser extent than in the wild-type (see Table 1), indicating that ALC cannot completely suppress the absence of this subunit of the i-AAA protease complex located in mitochondrial inner membrane.

One of the described targets of Mdm30p is Fzo1p, the conserved GTPase present in the outer membrane of mitochondria involved, together with the inner membrane associated protein Mgm1, in mitochondrial fusion (Escobar-Henriques et al., 2006).

Recent data indicated that mitochondrial fusion protects cells from apoptosis (Olichon et al., 2003; Sugioka et al., 2004; Neuspiel et al., 2005), while the inhibition of Drp1-dependent mitochondrial division delays and partially inhibits intrinsic apoptosis (Frank et al., 2001; Lee et al., 2004; Jagasia et al., 2005). The fzo1 mutant showed a viability extension of about 20% compared to the wild-type, suggesting that also mitochondrial fusion could be a hazardous process during aging. Anyway, the presence of ALC extended lifespan to more than 300% and increased the advantage of the fzo1 mutation, compared to the wild-type + ALC, up to 39%.

The absence of Fzo1p in yeast causes mitochondrial membrane fragmentation and, in consequence, cells quantitatively lose mtDNA and are unable to grow on the nonfermentable carbon source glycerol. Mutations in components required for mitochondrial division, such as DNM1 or FIS1, suppress mitochondrial fragmentation and the glycerol growth defect (Mozdy et al., 2000). The latter phenotype was used in a screening to identify compounds inhibiting mitochondrial division to be used as therapeutics for stroke, myocardial infarction and neurodegenerative diseases (Cassidy-Stone et al., 2008). As shown in Fig. 6A, the presence of ALC restored the growth of the fzo1 mutant in glycerol, both at 28 and 37°C, suggesting that ALC is able to inhibit mitochondrial division.

Figure 6.

 Effect of ALC on the mitochondrial fusion mutant fzo1. (A) Tenfold serial dilutions of wild-type and mutant cultures were spotted on YPY solid medium without or with the addition of 1mM ALC(YPY+ALC). Plates were incubated at the indicated temperatures for 4 days before recording. (B) Mitochondrial morphology, revealed by mitoGFP visualization, in BY4741 (wt) and fzo1 mutant cells grown in SD medium in the absence (SD) or in the presence of 1 mm ALC.

We then analyzed the mitochondrial morphology in the wild-type and in the fzo1 mutant by the use of mitoGFP, a mitochondria-targeted green fluorescent protein (Westermann & Neupert, 2000). As shown in Fig. 6B, in wild-type cells, mitochondria formed a tubular network while they appeared fragmented and as single spots in fzo1 mutant cells. The addition of ALC allowed the formation of the tubular network also in fzo1 cells, suggesting a role of this molecule in maintaining the fission/fusion dynamic of mitochondria.

We also analyzed the mitochondrial morphology in a wild-type strain and in the pro-apoptotic mutant Kllsm4Δ1, which is known to present fragmented mitochondria (Mazzoni et al., 2005a,b). We recorded the mitochondrial morphology after 2 days of growth in the early stationary phase. At this stage of growth, wild-type mitochondria start to fragment and tubules appear shorter compared to the exponential phase (Fig. 7E) while, in ALC-treated cells, mitochondrial tubules were clearly elongated (Fig. 7F). In lsm4Δ1, more than 87% of cells showed an aberrant mitochondrial morphology with a punctuate distribution instead of the wild-type tubular shape (Fig. 7G). As shown in Fig. 7H, in the presence of ALC some tubules were still formed, suggesting that mitochondrial fragmentation was, at least in part, inhibited also in this mutant.

Figure 7.

 Effect of ALC on mitochondrial morphology during aging. Mitochondrial morphology was analyzed by mitoGFP visualization in the wild-type (A, B, E, F) and in the Kllsm4Δ1 mutant (C, D, G, H) after 2 days of growth in SD medium in the absence (E and G) or in the presence (F and H) of 1 mm ALC. Panels A, B, C and D show the correspondent phase contrast images.

Oxygen consumption in the presence of ALC

To check if ALC could extend CLS by causing a metabolic shift as induced by glycerol, we measured O2 consumption in the two wild-type strains using an oxygraph on live respiring cells at room temperature.

As shown in Fig. 8(A,B), two different wild-type strains grown in glycerol in the presence of ALC did not show evident increase in oxygen consumption compared to the growth without ALC, suggesting that this compound has not a significant effect in promoting respiration.

Figure 8.

 ALC increases oxygen consumption during aging: two wild-type laboratory strains, CML39-11A (A and C) and BY4741 (B and D), were grown in 3% glycerol up to OD600 = 0.2 (A and B) or 2% glucose for 2 days (C and D). Respiration rates are expressed as nanomoles of oxygen consumed per milliliter of solution per minute. Average and standard deviations, obtained from three independent experiments, are also indicated.

Mitochondrial morphology is closely linked to energy metabolism. Enhanced respiration correlates with an interconnected mitochondrial network, while reduced oxidative phosphorylation and enhanced glycolysis correlate with fragmented mitochondrial and matrix expansion (Alirol & Martinou, 2006).

We tested if ALC could promote respiration through its ability to maintain the mitochondrial tubular network in aged wild-type strains, which normally present fragmented mitochondria.

As shown in Fig. 8(C,D), ALC accelerated O2 consumption in wild-type strains after 2 days of growth in SD, correlating the observed effect of ALC on mitochondrial morphology during aging to a higher respiratory performance.

Lsm4Δ1 mutant cells cannot grow in glycerol, but they are still able to grow in 0.2% glucose. In this condition, oxygen consumption is very low, probably because of the altered mitochondrial morphology described in this mutant. We found that the respiration rate of the mutant was higher in the presence of ALC, suggesting again that the improvement in the mitochondrial network by ALC correlates with a faster oxygen consumption. In contrast, wild-type cells grown in the same conditions (0.2% glucose) showed similar respiratory rates with or without ALC (Fig. S1).


In this work, we studied the effects of carnitines, in particular of Acetyl-l-Carnitine (ALC), in protecting yeast cells from aging and apoptosis.

Acetyl-l-carnitine and aging

We found that ALC was able to suppress premature aging and apoptotic phenotypes in lsm4Δ1, a pro-apoptotic mutant impaired in the mRNA decapping pathway. In this mutant, mRNA degradation is delayed and cells undergo apoptosis through a caspase-dependent pathway (Mazzoni et al., 2003, 2005a,b).

We also tested the effect of ALC in a yeast mutant lacking the caspase-like YCA1 gene. This mutant is more resistant to apoptotic stimuli in that survives to hydrogen peroxide treatment better than the wild-type and shows delayed aging (Madeo et al., 2002).

The presence of ALC did not improve further the lifespan of the yca1 null mutant, suggesting that the mechanism of ALC protection might pass through the inhibition of the yeast caspase or, alternatively, may act downstream of the action of caspase. Nevertheless, yca1 cell viability during aging was higher in the presence of ALC indicating that, after prolonged stationary phase (about 9 days), ALC could prevent cell death independently of Yca1.

One might speculate that, in this case, cells died through necrosis and that ALC could suppress this kind of death. In fact, it has been recently reported that combined treatment with LC and a pan-caspase inhibitor (zVAD-fmk) effectively protected human lung epithelial cells against amiodarone-induced cell injury, which causes both necrosis and apoptosis (Yano et al., 2008).

ALC and mitochondrial metabolism

The role of carnitine in the metabolism of S. cerevisiae has not been investigated beyond the shuttling of acetyl residues. ALC protection in aging cells was higher in respiratory competent cells, while this effect was not observed in rho° cells, suggesting that major ROS production comes from active mitochondria. Nevertheless, in glycerol growing cells mitochondria are highly active but lifespan is longer compared to glucose-grown cells. This is also true for cells grown in low glucose, a condition referred as calorie restriction (CR). Under this condition, although ROS production is higher compared to 2% glucose-grown cells, aging is delayed and apoptotic phenotypes, such as cytochrome c release, DNA fragmentation and phosphatidylserine exposure are less evident. It has been also proposed that CR postpones chronological aging in yeast by turning on the RTG signaling pathway and increasing gluconeogenesis, which could explain the observed age-dependent changes in the levels of trehalose and glycogen, the major glucose stores in yeast (Goldberg et al., 2009).

In glycerol-grown cells, ALC further extended lifespan in two wild-type laboratory strains, CML39-11A and BY4741, of about 29 and 50%, respectively.

It has been recently reported that ALC improves growth in the presence of H2O2 and of weak organic acids, such as lactate, and that this effect appears to be independent of the activity of the carnitine shuttle (Franken et al., 2008).

Actually, in the presence of ALC, cells were more resistant to H2O2 and the amount of intracellular ROS following the oxidative stimulus was lowered in both the wild-type strains we used.

Our results suggest that ALC might act on some components that induce the activation of apoptosis in glucose-grown cells carrying active mitochondria. Oxygen consumption during respiration was not affected by the presence of ALC, suggesting that the extension of CLS is not caused by a metabolic shift toward respiration.

ALC and mitochondrial morphology

In most of apoptotic scenarios in mammalian cells, the interconnected mitochondrial network early converts into a fragmented morphology through the activation of the physiological fission machinery, a process known as thread–grain transition (Frank et al., 2001; Scorrano, 2005; Parone & Martinou, 2006). The inhibition of this process in mammals, that include Fis1p and Drp1, the ortholog of the yeast Dnm1p, impairs mitochondrial fragmentation, the release of cytochrome c and the process of apoptotic death. This makes Drp1 an appealing pharmaceutical target for neurodegenerative diseases, stroke and myocardial infarction, where the inhibition of apoptosis may be therapeutically beneficial.

We found that ALC favors mitochondrial fusion in fzo1 mutants and in aged yeast cells, probably as a consequence of the inhibition of the mitochondrial division machinery.

According to this, in the absence of dynamine, the increase of lifespan by ALC was modest, suggesting a relevant role of the mitochondrial fission pathway involving dynamine in limiting lifespan extension.

This could also be true for the fission-defective fis1 mutant, which showed a lifespan increase of 30% in the presence of ALC. In addition, ALC clearly helps mitochondria of aging yeast to maintain the tubular network shape, indicating that it can suppress mitochondrial fragmentation in old cells. In a recent paper (Goldberg et al., 2009), it has been reported that in aging cells under calorie restriction there is an increase of the mitochondrial fusion protein Mgm1 and an inhibition of the mitochondrial fission proteins Caf4 and Mdv1 which, together with Dnm1 and Fis1, could represent other ALC targets involved in mitochondrial division.

At present, we may hypothesize that the increased respiration rate stimulated by ALC in wild-type stationary cells and in lsm4Δ1 mutant cells, which present fragmented mitochondria, is the result of the formation of a mitochondrial tubular network operated by ALC.

In conclusion, we found that ALC protects yeast cells by inhibiting Yca1 activation, ROS production and mitochondrial thread–grain transition (Fig. 9). During aging, or upon an apoptotic stimulus, the activation of Yca1 seems to induce mitochondrial fragmentation (Mazzoni et al., 2005a,b). As ALC partially inhibits Yca1 activation, in the presence of this molecule mitochondrial fragmentation is reduced. We have also demonstrated that in fzo1, in the absence of other stimuli, ALC can restore the mitochondrial tubular network, suggesting an inhibition of the mitochondrial division machinery.

Figure 9.

 Schematic representation of the principal targets of ALC in yeast cells during aging and upon apoptotic stimuli.

Whether the interactions of ALC with Yca1 and the mitochondrial division machinery are direct or indirect, and what mediates these interactions, are still under investigation.

Experimental procedures

Yeast strains and culture media

Yeast strains used in this work are listed in Table 2.

Table 2.   Yeast strains used in this study
CML39-11AMAT a, ade1-101, his3Δ1, leu2Δ0, ura3, trp1-289Mazzoni et al., 2005a,b
CML39-9AMAT a, ade1-101, his3Δ1, leu2Δ0, ura3, trp1-289, yor197w::kanMX4Mazzoni et al., 2005a,b
MCY4/313Kllsm4Δ1MAT a, ade1-101, his3Δ1, leu2Δ0, ura3, trp1-289, LEU2-GAL1-SBD23, pRS313/lsm4Δ1Mazzoni et al., 2005a,b
BY4741MAT a, his3Δ1, leu2Δ0, met15Δ0, ura3Δ0Brachmann et al., 1998
Y05277Isogenic to BY4741 except for (MDM30)YLR368w::kanMX4EUROSCARF Collection
Y07144Isogenic to BY4741 except for (YME1) YPR024w::kanMX4EUROSCARF Collection
Y01489Isogenic to BY4741 except for (DNM1)YLL001w::kanMX4EUROSCARF Collection
Y01458Isogenic to BY4741 except for (FIS1)YIL065c::kanMX4EUROSCARF Collection
Y03319Isogenic to BY4741 except for (FZO1) YBR179c::kanMX4EUROSCARF Collection

The culture media utilized for cell growth were complete YP (1% yeast extract, 2% peptone) supplemented with 2% glucose (YPD) at 28°C or minimal medium (0.67% yeast nitrogen base) containing 2% glucose (SD). Amino acids (20 mg/l) were added to all media according to the specific requirements of each strain. Solid media were supplemented with 2% Bactoagar (Difco, Detroit, MI, USA). These media were also supplemented with L-carnitine, acetyl-l-carnitine and propionyl-l-carnitine to a final concentration of 1 mm.

Cell viability

The determination of chronological lifespan was done as described in Palermo et al., 2007. Briefly, cell suspensions (5 μL) containing approximately 6 × 106 cells mL-1 were poured on a thin layer of YPD agar on a microscope slide. A cover slip was placed over the samples and, after 24 h, viable and unviable cells were scored on the basis of their ability to form microcolonies.

H2O2 sensitivity

To determine the sensitivity to oxygen peroxide, cells growing exponentially were exposed to 0.8, 1.2 and 3 mm H2O2 at 28°C for 4 h. Cell viability was determined by counting the formation of micro-colonies.

Fluorescence microscopy

The presence of ROS was detected with DHR (Sigma D1054) as described previously (Madeo et al., 1999). For analysis of mitochondria morphology, we used plasmid pYX232-mtGFP (Westermann & Neupert, 2000) that targets GFP into mitochondrial matrix. For image acquisition of the cells containing this plasmid, we used an Axioskop2 fluorescence microscope (Carl Zeiss, Jena, Germany) equipped with a digital camera (micro-CCD).

Respiration measurements

Cell respiration rate was determined on exponential cultures grown in 3% glycerol, 0,2% glucose or on stationary cultures grown in 2% glucose minimal medium (SY, SD0.2 and SD, respectively). In 10 mm potassium phosphate buffer (pH 7.5) containing 20 mm glucose, 2.5 × 107 cells/mL was suspended. Oxygen consumption was measured at room temperature with an Oxytherm oxygraph (Hansatech Instruments, Norfolk, England). Respiration rates are expressed as nanomoles of oxygen consumed per milliliter of solution per minute.