The effect of copy number on α-synuclein’s toxicity and its protective role in Bax-induced apoptosis, in yeast

Parkinson’s disease (PD) is a severe neurodegenerative disorder which is categorised by bradykinesia (slow movement), postural instability, muscular rigidity and resting tremors [1,2]. Many PD cases are from an obscure cause, though some are known to be caused by missense mutations in the α-synuclein gene [3,4]. In PD, α-synuclein accumulates into protein aggregates that show, both in in vitro and in vivo, many features of amyloid formation [4,5]. α-synuclein is water-soluble, and it is made up of 140 amino acid residues. It contains seven 11-residue repeats in the positively charged N-terminal area, followed by a vastly acidic 40-residue C-terminal tail. α-synuclein is known to be associated with mitochondrial function [6,7]. However, it was also found in red blood cells that lack mitochondria [8–10]. α-synuclein is present predominantly in the presynaptic region of neuronal cells. However, the actual function of α-synuclein has not yet been defined.

Apoptosis serves as a mechanism that helps to restructure tissues through development, considering this definition/description, it was assumed that apoptosis could not be found in modest eukaryotes, i.e. yeast. However, when apoptosis started gaining link with the removal of damaged cells, major regulatory proteins involved in programmed cell death were recognised in yeast [11]. The α-synuclein protein is conspicuously expressed in the central nervous system [12], and functions in vesicular tracking and lipid metabolism [10,13]. It physically interacts with vesicular membrane proteins [3]. Overproduced α-synuclein inhibits phospholipase D, thereby inducing accumulation of lipid droplets which eventually adversely affects vesicle trafficking [3]. Overexpression of α-synuclein is associated with PD and also multiple system atrophy which is a rare neurological disorder [14]. Yeast is useful in the study of human apoptosis; in yeast, apoptosis can be induced by both intrinsic and extrinsic factors [15].

When overexpressed in yeast, α-synuclein protein migrates to the periplasmic membrane, just like in human cells. As in human cells, it forms cytoplasmic inclusion bodies (which are referred to as Lewy bodies in human cells) when overexpressed in yeast, resulting in the display of apoptotic features in yeast [5,16]. The process of apoptosis in human cells is complex. It is regulated by the Bcl-2 family of proteins which have both pro- and anti-apoptotic members [17]. It can occur dependent or independent of the tumour suppressor protein, p53 [17]. The absence of yeast homologues of mammalian Bcl-2 family members and also p53 initially undermined yeast studies linked to human apoptosis. However, identification of a yeast caspase (named metacaspase) made further studies, with human proteins like α-synuclein, plausible [18]. In human cells, activation of caspases underlines a commitment to apoptosis.

Overproduction of reactive oxygen species (ROS) controls the initiation of the apoptotic pathway and is thought to be a hallmark of mammalian apoptosis. Intracellular ROS is produced during the process of electron transport across the endoplasmic reticulum (ER) or nuclear membranes [18,19]. Like human cells, yeast cells’ response to ROS varies with the amounts of cellular ROS that are present [20]. When the amounts of ROS are high, some antioxidant proteins, such as Msn2, Yap1 and 4p transcription factors, are activated within yeast cells [18,20]. Activation of these proteins results in a delay in yeast cell division. In the presence of even higher doses of ROS, some, if not all, cells undergo apoptosis [19]. α-synuclein overexpression in yeast is known to induce high levels of ROS.

α-synuclein is usually present in the cytosol and nucleus. It is associated with outer membranes of neurons affected by PD [21]. As noted above, periplasmic membrane association is also observed in yeast during α-synuclein overexpression that leads to its aggregation/misfolding and which culminates in yeast cell death [3]. Pathogenic neuronal cell demise stemming from misfolding/aggregation of α-synuclein involves DNA degradation triggered by the mitochondrial endonuclease G. The translocation of endonuclease G from mitochondria to the nucleus results in cell death [21]. Nuclear localisation of endonuclease G is cytotoxic to dopaminergic neurons of patients with PD. In flies and nematodes, endonuclease G is also crucial for α-synuclein-induced degeneration of dopaminergic neurons [21].

Besides α-synuclein, other proteins are linked to PD. They include Parkin, DJ-1 and PINK1. Like α-synuclein, they are all linked to mitochondrial function [13,22]. α-synuclein has been shown to interact physically with neuronal mitochondrial membranes. Consequently, overexpression of α-synuclein interferes with the normal function of mitochondria resulting in neuronal apoptosis [23–25]. It is, therefore, reasonable to say that one or more mitochondrial factors trigger α-synuclein-induced apoptosis in neuronal cells.

Mitochondrial dysfunction plays a major role in PD [26]. A systemic shortage of mitochondrial complex I activity has been reported in PD patients [16]. Hence, it has been proposed that the normal function of α-synuclein involves sequestration of cytochrome c within mitochondria. Only during its overexpression are Lewy bodies formed through α-synuclein aggregation, and the release of cytochrome c from mitochondria to the cytosol leading to neuronal apoptosis [16]. Apoptosis, which controls the overall turnover of cells without damaging the cellular environment, was discovered as having a vital role in the maintenance of cell number and function by Wyllie and colleagues [27]. The process of apoptosis, like all biochemical processes, is coordinated.

Human Bax protein-induced cell death in the yeast Saccharomyces cerevisiae shows the characteristics of apoptosis, which include membrane blebbing, DNA fragmentation, phosphatidylserine externalisation at the cytoplasmic membrane surface, marginalisation and condensation of chromatin, and vacuolisation of cytoplasm [28]. Bax expression also causes the release of cytochrome c from the mitochondria and decreases levels of cytochrome c oxidase [29]. As the mitochondria are interconnected and elongated [30], damaged mitochondria are removed through fission via a conserved mechanism [31]. Cyclin C, the activating partner of the cell cycle kinase Cdk8, translocates in response to stress to the mitochondria from the nucleus, suggesting that cyclin C may have a role to play in programmed cell death and mitochondrial fission [32].

α-synuclein produces a three-way complex with anionic lipids, like cytochrome c and cardiolipin. The complex induces peroxidase activity that leads to the enhancement of hetero-oligomerisation of α-synuclein with cytochrome c ultimately forming a huge molecular weight aggregate [16]. The aggregate induces activation of caspases and formation of the apoptosome, which represents a commitment to apoptosis [16]. Pro-apoptotic factors are released via damage to presynaptic mitochondria which serves as a threat to the survival of all neurons [33].

α-synuclein can halt the oxidative chain reaction, thereby hypothetically playing a vital handy role in averting brain lipid oxidative damage [8]. It has been claimed that aggregation of α-synuclein protein could be inevitable, but the circumstances which warrant this aggregation in cells is not yet well understood [9,34]. This could be due to the poor understanding of α-synuclein’s true function, although it is known that it is associated with vesicular membranes, and other membrane interactions [9,34].

The present study’s aim was to study the characteristics of two pro-apoptotic human proteins, Bax and α-synuclein, in the baker’s yeast S. cerevisiae. The two pro-oxidant human proteins, Bax and α-synuclein, which are deeply involved in the induction of apoptosis in human cells, were used as apoptotic triggers in yeast and were co-expressed.

The yeast strain W303-1A Mata (ATCC #208352), is auxotrophic for the genes ADE2, HIS3, LEU2, TRP1 and URA3. New yeast strains were derived from W303-1A by transforming integrative plasmids (Supporting Information, Sections 1 and 2) which would express α-syn from the MET25 or GAL1 promoter.

Plasmids bearing α-syn gene expression cassettes under the control of either the methionine-repressible MET25 or galactose-inducible GAL1 promoter (MET25p/GAL1p; see Supporting Information, Sections 1 and 2) were used for genomic integration at the HIS3, TRP1 and URA3 chromosomal loci of the yeast strain to yield strains that contain 1–3 copies of α-syn. Similarly, plasmids bearing Bax-α gene expression cassettes under the control of galactose-inducible GAL1 promoter was used for genomic integration at the LEU3 chromosomal locus of the yeast strain. The integrative transformation was carried out using a published protocol [35,36].

Rhodamine 1 2 3 is a cationic fluorescent, cell-permeable dye. It is readily sequestered within active mitochondria. Rhodamine 1 2 3 dye from the yeast mitochondrial staining kit was used (Molecular Probes, Y7530), after the expression of protein or proteins of interest. A total of 1 × 10⁶ cells were suspended in 1 ml of 50 mM sodium citrate buffer, containing 2% w/v glucose, at pH 5. A final concentration of 50 µM of Rhodamine 1 2 3 was added to cells and incubated at room temperature for 30 min. Fluorescent cells were visualised using a fluorescence microscope.

Cell death was assessed by staining cells with the red dye Phloxine B (Sigma, P-4030-25G) [37]. Live cells expel the dye, whereas it is accumulated in dead cells. This can be observed by fluorescence microscopy. Staining experiments were performed as published earlier [36].

AAT Bioquest Fluorimetric Intracellular Total ROS Activity Assay Kit (#22901) was used for measuring ROS. Experiments were performed as published earlier [36].

AAT Bioquest JC-10 Mitochondrial Membrane Potential Assay kit (#22800) uses water-soluble JC-10 to determine mitochondrial membrane potential (MMP) quantitively. Experiments were conducted as per the published protocol [36].

AAT Bioquest TUNEL Apoptosis Assay kit (#22844) was used for the detection of nuclear DNA fragmentation (NDF). The assays were performed as described earlier [36].

Western blotting was carried out using standard protocols (Haar, 2007), using primary antibodies specific to α-syn (Proteintech; #10842-1-AP), c-Myc mouse monoclonal antibody (Thermo Scientific, #MA 1-980) or β-actin (Proteintech; #60008-1-Ig) [38].

The result of the toxicity mediated by wildtype human α-synuclein protein from two different inducible promoters; and mutant human α-synuclein proteins (A30P and A53T) when defined copy numbers of its gene were expressed, were presented below. At first, yeast cells were transfected with an episomal plasmid containing α-synuclein on a PGK1p promoter (pSYE239/α-synuclein-HA), protein expression from the PGK1p can occur in media containing either glucose or galactose as a carbon source. The control plasmid pSYE239, which includes no α-synuclein gene, was also transformed into yeast cells (Figure 1).

Episomal plasmids, in theory, should provide multiple copies of the plasmid within yeast cells. However, it has been published that yeast strains bearing episomal plasmids that contain GAL1p-α-synuclein expression cassettes offer less than one copy because of α-synuclein’s inherent toxicity in yeast [3]. Hence, the effect of increasing copy number was studied by integrating α-synuclein expression cassettes at different chromosomal loci so that defined 1 copy, 2 and 3 copies of the α-synuclein gene could be expressed. Also, mutant α-synuclein genes (A30P and A53T) were integrated into yeast cells in 1–3 copies. Alongside, various control strains were generated.

The results seen in Figure 1 (where α-synuclein protein expression is switched on in the presence of both galactose and glucose), clearly show that human α-synuclein, when expressed from an episomal 2-micron (2µ) plasmid, does not stop cell growth and therefore is not toxic to yeast. Compared with the control strain, there is no significant retardation of growth. Outeiro and Lindquist (2003) have reported that yeast cells expressing GFP-tagged α-synuclein from a 2-µ plasmid also grew when expressed from the galactose-inducible GAL promoter. Expression of α-synuclein, controlled by the MET25 promoter, from a 2-µ plasmid, also did not inhibit cell growth [22].

Defined copies of α-synuclein would allow control of its amount at a particular time. The results in Figure 2 show that in contrast to the expression of α-synuclein from an episomal plasmid, expression of a single integrative copy of the human α-synuclein gene from the galactose-inducible GAL1 promoter blocks growth of yeast cells, likewise two and three copies of α-synuclein. These results obtained on a solid agar plate, containing galactose as the sole carbon source for growth, were corroborated with cells grown in minimal liquid medium containing galactose (Figure 2B). In Figure 2C,D, α-synuclein induced significant cell death in yeast cells. α-synuclein was then tagged with eGFP to allow the visible study of α-synuclein aggregation. Results from Figure 3 show that the expression of only three copies of α-synuclein-eGFP fusion protein retards/stop cell growth on galactose-containing SG agar plates. Figure 3B show the cells expressing three copies of α-synuclein-eGFP fusion protein have the most considerable retardation of growth. However, the growth retardation is not significant compared to cells carrying one copy and two copies of α-synuclein-eGFP; this could be due to the large size of eGFP fusion protein, the expressed protein appears to be aggregated (Figure 3C,D). It was seen that as the copy number of α-synuclein-eGFP gene increases, there was significantly more aggregation of α-synuclein. The results below indicate that, although α-synuclein-eGFP fusion protein does not block cell growth to any appreciable degree, it imposes apoptotic cell death which increases with increasing copy number. Figure 2E shows that expression of one copy, two and three copies of α-synuclein produces much more ROS than control cells which contain empty plasmid that bears no α-synuclein gene, similarly, this was observed in eGFP strains (Figure 3L). Figures 2F and 3E shows that there is a drop of MMP in yeast cells expressing one, two, and three copies of α-synuclein-HA and α-synuclein-eGFP compared with their respective control, indicating that both HA- and eGFP tagged α-synuclein expression has a deleterious effect on mitochondrial outer and inner membranes and thereby affected MMP which is a hallmark of apoptosis [39]. Figures 2G,H and 3H,I shows that DNA fragmentation was high in cells that express three copies of α-synuclein-(HA and eGFP) compared with strains expressing one copy and two copies but was very low in control cells which contained empty plasmids. The results in Figures 2 and 3 indicate that expression of HA-tagged α-synuclein, from a chromosomal locus, induces apoptosis in yeast.

The mutant α-synuclein (A30P) harbouring strains did grow in galactose like the control strains that contained 1–3 copies of empty plasmids, without any α-synuclein gene. The results on plates show that the mutant α-synuclein (A30P) protein is non-toxic, the observations were corroborated in cells that were grown in a liquid medium; see Figure 4A,B. Contrary to the effects of wildtype α-synuclein expression, the A30P mutant showed no difference in growth and cell death between the strains expressing mutant A30P α-synuclein in 1–3 copies, and when compared with their respective controls. Mutant α-synuclein gene (A53T) α-synuclein in 1–3 copies did not grow in galactose. The results on plates show that like the wildtype α-synuclein protein, the mutant α-synuclein (A53T) protein is toxic. The observations were corroborated in cells that were grown in a liquid medium; see Figure 4G,H. Phloxine B staining of cells expressing 1–3 copies of the α-synuclein mutant (A53T) in Figure 4I,J, suggest that A53T mutant imposes toxicity on yeast cells while A30P mutant does not; wildtype α-synuclein and A53T mutant α-synuclein seems to have similar properties in terms of toxicity/cell death induction. Same results were obtained for yeast cells bearing α-synuclein on MET25p promoter, and α-synuclein with no tag (α-synuclein-No HA) (Figure 5).

α-synuclein is part of the SNARE complex at the synapses, and this is its physiological role. When it misfolds, α-synuclein becomes pathological, as seen in PD. In previous sections, the toxicity of wildtype and mutant (A53T) α-synuclein have been shown in different yeast strains. It was reported that only after increasing the dosage from one copy to two copies of α-synuclein does α-synuclein cause cell death in yeast [3,10]. However, we see that a single integrative copy of the wildtype α-synuclein (tagged or untagged) has already a profound effect on cell growth and death.

Having seen that yeast strains harbouring the GAL1 promoter-driven, chromosomally integrated copy of the Bax expression cassette did not grow on galactose containing media (Figure 6A–C), it was confirmed that pro-apoptotic Bax protein is toxic to yeast. The episomal plasmid containing α-synuclein (pSYE239/α-synuclein-HA) and the empty vector (pSYE239), as shown in (Figure 6D), was introduced into the yeast strain,:: Bax(LEU2), that already contains an integrated copy of the Bax expression cassette at yeast’s LEU2 chromosomal locus, Figure 6D. Yeast cells were rescued from death in cells co-expressing Bax (from an integrated copy) and α-synuclein (from a 2µ-plasmid) (see Figure 6D–G).

Copies of the α-synuclein-HA, α-synuclein-eGFP, A30P and A50T (mutant) genes expression cassette were also integrated into the yeast strain harbouring Bax(LEU2) at its TRP1, HIS3 and URA3 loci to create 1, 2 and 3 copies of the respective α-synuclein strains. Galactose allows expression of Bax and α-synuclein, both from the GAL1 promoter. It seems that cells which express one copy of HA-tagged α-synuclein struggle to overcome Bax-induced block in growth while the cells with the empty plasmid were completely blocked in Growth because of Bax protein's toxic effects. The results show that cells which express two copies of HA-tagged α-synuclein can overcome Bax-induced apoptosis while the cells with empty plasmid were completely blocked in growth or dead because of Bax’s toxicity. These results have been corroborated by the growth of cells in SG liquid medium (Figure 7A,B). Three copies of HA-tagged α-synuclein could not overcome Bax-induced block in growth anymore. It behaved just like cells with one copy. The third copy seemed to increase the toxicity in the yeast cells.

Co-expression of two copies of HA-tagged α-synuclein with Bax results in nearly complete protection from death. The rescue by two copies of α-synuclein was significant. Although three copies of HA-tagged α-synuclein could not overcome blockage in cell growth induced by Bax, it seems that co-expression of 3 copies of HA-tagged α-synuclein with Bax in Bax(LEU2)::α-syn-HA(TRP1, HIS3, URA3) cells still protects cells from Bax induced death. This can be seen when percentage cell death (Figure 7C,D) is compared with control cells Bax(LEU2)::–(TRP1, HIS3, URA3) which express Bax alone but contain three empty plasmids that do not express any α-synuclein. There appears to be a significant reduction in ROS when two copies of HA-tagged α-synuclein were co-expressed with Bax. The results would suggest that two copies of α-synuclein protect cells from excess ROS generated through Bax expression in yeast (Figure 7E).

Figure 7F show that expression of one-copy of HA-tagged α-synuclein is protective of mitochondrial membranes damaged by Bax expression in yeast. It also indicates that there is a further increase in MMP in cells co-expressing Bax and two copies of α-synuclein-HA. This suggests that two copies of α-synuclein truly protects mitochondrial membranes from Bax-mediated mitochondrial damage. The TUNEL results (Figure 7G) show that expression of two-copies of HA-tagged α-synuclein prevents Bax-mediated cell death in yeast. The bar chart (Figure 7G) clearly indicates that two copies of HA-tagged α-synuclein rescue the majority of cells from Bax mediated apoptosis.

In Figure 8A,B, the 1-copy and 2-copies α-synuclein-eGFP containing strains grew upon full expression of the α-synuclein-eGFP fusion protein in the presence of Bax. The 2-copies strain was having more growth than 1-copy, while the 3-copy strain did not grow at all when both proteins, Bax and α-synuclein-eGFP, were fully expressed. Hence, it suggests that only 1 and 2 copies of α-synuclein-eGFP can rescue the toxic effects of Bax. The results confirmed the earlier observations presented in Figure 7. The result shows that the percentage of cell death is least in the strain that co-expressed Bax, and two copies of α-synuclein-eGFP (Figure 8C,D), the difference in cell death is significantly low in the strain that co-expressed Bax and two copies of α-synuclein-eGFP.

The strain that co-expressed Bax and two copies of α-synuclein-eGFP produces the lowest amount of ROS with a significant difference from strains with 1 and 3 copies (Figure 8E). Cells in the strain which have two copies of α-synuclein-eGFP have the highest MMP, followed by strain with one copy and then strain with three copies of α-synuclein-eGFP (Figure 8F). Figure 8G,H shows that DNA fragmentation was high in the strain containing three copies of α-synuclein-eGFP, followed by the strain with one copy. The strain with two copies of α-synuclein-eGFP has the least DNA fragmentation. The presence of α-synuclein aggregation was seen in the strains expressing both 1-copy and 3-copies of α-synuclein-eGFP (Figure 8J). α-synuclein is barely seen in the strain that co-expressed 2-copies of α-synuclein-eGFP together with Bax.

In unstimulated cells, monomeric Bax-α (referred to as just Bax) protein, a pro-apoptotic member of the Bcl-2 family, usually resides in the cytosol and the periphery of intracellular membranes which include the mitochondria. Bax only inserts into mitochondrial membranes following a death signal [17]. Only homodimers of Bax are known to be pro-apoptotic. Malfunctioning of apoptosis can lead to diseases like viral infections, autoimmune diseases, cancer, while heightened apoptosis can lead to ischemic disease, neurodegenerative diseases, and AIDS. After co-expressing two human pro-apoptotic proteins α-synuclein and Bax, surprisingly, α-synuclein manifests an anti-apoptotic role in the presence of Bax, rescuing Bax-mediated cell death in baker’s yeast.

Figure 9A–E shows co-expression of Bax with 1, 2 or 3 copies of mutant α-synuclein (A30P and A53T) genes. None of the yeast strains that co-express Bax and α-synuclein (A30P and A53T) proteins grew on expression in galactose. Hence, mutant α-synuclein (A30P and A53T) does not rescue yeast cells from Bax-induced cell growth block/death. It should be noted that independent of the copy number, yeast strains expressing mutant α-synuclein (A30P) alone are not restricted in cell growth and are therefore not toxic to yeast. In contrast, mutant α-synuclein (A53T) was as toxic as the wildtype (Figure 4).

When copies of α-synuclein were expressed, the α-synuclein protein seemed to localise to the mitochondria. It has been suggested that it is bonding to the mitochondrial outer membrane could affect the linkage of other proteins to the mitochondria, thereby causing modifications that eventually give way to permeability translocation pore dependent endonuclease G release [21]. Misfolded α-synuclein can spread from one neuron to another in the brain with the development of PD [40]. In a non-pathogenic nerve terminal, there is an equilibrium between monomeric, unfolded cytosolic α-synuclein and a multimeric, membrane-bound, α-helical form of α-synuclein which accompany SNARE complexes [7,40]. The relation between the conformation of the endemic form of α-synuclein and its aggregated pathological form is not entirely understood [14,40]. It is unclear if the aggregation of α-synuclein is from its membrane-bound oligomeric or cytosolic monomeric forms.

It was observed that the yeast strain co-expressing one copy of both HA-tagged and eGFP-tagged α-synuclein, and three copies of eGFP-tagged α-synuclein with a copy of Bax produced relatively large amounts of both α-synuclein and Bax proteins. In contrast, the strains containing 2 and 3 copies of α-synuclein-HA and 2 copies of α-synuclein-eGFP produced much lesser amounts of both proteins. The β-actin loading control shows that equal amounts of proteins from cell lysates were indeed loaded; this suggests some form of protein–protein interaction in the two proteins.

From the Western blots, it would appear that protein degradation had occurred with the wildtype α-synuclein and Bax proteins when 2 and 3 copies of α-synuclein were co-expressed. The differences, between the levels of α-synuclein and Bax in the 1-copy α-synuclein strain on one hand and the 2 or 3-copy α-synuclein strains on the other, suggest that there is likely to be strong communication and/or interaction between the wildtype α-synuclein and Bax proteins during co-expression of 1–3 copies of α-synuclein and Bax. The strain co-expressing 2-copy α-synuclein-eGFP with Bax produces the least amount of both α-synuclein-eGFP and Bax proteins. This was also observed in yeast cell expressing α-synuclein under the control of MET25 promoter and α-synuclein untagged with HA (Figure 10).

With an increase in α-synuclein copy number, it was observed that there was a significant effect on cell growth, cell death, ROS, NDF and mitochondrial potential when α-synuclein is expressed in yeast cells from the GAL1 and MET25 promoters. This further confirmed that the toxicity of α-synuclein is proportional to an increase in gene dosage (i.e. an increase in copy number). However, the rescue of the yeast cell was seen when co-expressed with Bax. Figure 10I shows Rhodamine 1 2 3 stainings of cells from yeast strains that co-express Bax and one, two or three copies of untagged α-synuclein. Depending on the polarisation of the membranes (which includes mitochondrial membranes), the Rhodamine 1 2 3 stain can accrue on live cell membranes. This stain indicates the state of membrane structures. Intact membranes show durable stain in contrast with cells which have lost membrane polarisation and therefore stain very weakly or not all.

The results above clearly show that the untagged α-synuclein protein behaves similarly to the HA-tagged protein. Increase in untagged α-synuclein gene dosage showed a significant effect on cell growth and mitochondrial potential when α-synuclein was expressed from the GAL1 promoter, and this further confirmed that the toxicity of α-synuclein is proportional to an increase in copy number.

Can mutant α-synuclein proteins rescue yeast cells from the toxic effects of Bax? This was the question that came to our mind. The quantitative estimates of these Phoxine B staining results, as shown in Figure 9, clearly show that the mutant α-synuclein (A30P and A53T) has wholly lost wildtype α-synuclein’s ability to protect cells from Bax-mediated death. Figure 9E,J depicts the expression of the two proteins Bax and α-synuclein (A30P and A53T), as monitored by Western blots, in the three different strains that express 1–3 copies of α-synuclein (A30P and A53T). In contrast with cells which co-express α-synuclein proteins that protect cells from death, cells expressing A30P and A53T mutants show undiminished levels of α-synuclein and Bax proteins on Western blots.

Contrary to its pathogenic role, α-synuclein exercises substantial neuroprotection in mice, overexpression of α-synuclein prevents rapidly progressive neurodegeneration mediated by deletion of CSPα, a presynaptic co-chaperone which is part of the SNARE complex [41]. Moreover, it has been suggested that α-synuclein cooperates with CSPα in preventing neurodegeneration [14]. Wildtype α-synuclein has also been reported to protect neurons from undergoing apoptosis through inhibition of caspase-3 [42]. It has been suggested that when misfolding of α-synuclein protein occurs; there is usually an increase in its toxicity and a decrease in its protective functions [43].

Bax forms oligomers at the outer membrane of the mitochondria which results in the release of cytochrome c (from the mitochondrial inter-membrane space) and other proteins (i.e. Nuc-1, Ndi-1, AIF, cytochrome c) from the mitochondria. Inhibitor of apoptosis protein (IAP) is also released into the cytosol. IAP typically suppresses caspases by blocking caspase activities [44]. Once caspases are activated, they use multiple pathways to achieve apoptosis. Bcl-2 blocks the action of Bax typically, but p53 inhibits Bcl-2. Alteration in protein quality control (PQC) pathways has also been linked to mediate α-syn misfolding, accumulation, and aggregation [45].

Rescue of apoptosis could target some of the pathways stopping apoptosis from occurring (Figure 11), this could include the restoration of mitochondrial function which is essential, as it will stop every other downstream process. Restoration of mitochondrial function by an anti-apoptotic protein could also mean blocking pores made on the mitochondria, which would lead to the prevention of mitochondrial protein translocation (Figure 11B). Inhibiting/preventing the activation of caspases, for example, preventing the conversion of pro-caspase-3 to caspase-3 could also be an anti-apoptotic intervention. Similarly, interruption of AIF, NUC-1 and Ndi-1 may be necessary for the prevention of apoptosis. Other possible rescue pathways could involve protein-protein interactions between pro and anti-apoptotic proteins. Mopping up of oxidative stress or ROS in cells could be another channel for rescue.

Results of the present study show an interesting trend. With increasing copy number of α-synuclein, when co-expressed with Bax, there was a progression in rescue from one copy to two copies, but then rescue did not occur with three copies of α-synuclein. Protein–protein interaction could have led to degradation (as seen in two copies of α-synuclein when co-expressed with Bax), on the introduction of the third copy, rescue activity decreases significantly, owing to more or over the aggregation of α-synuclein, this suggests that the level of α-synuclein protein present at a point in time dictates its behaviour (pro or anti-apoptotic).

Expression from episomal plasmids in yeast had failed to provide conclusive results regarding α-synuclein’s toxicity in yeast, the effect of an increasing number of defined copies of wildtype α-synuclein is indeed toxic to yeast. Amongst the two α-synuclein mutants, A30P and A53T, A30P appears not to be toxic to yeast even when three defined chromosomal copies were expressed. In contrast, the A53T mutant was found to be toxic just like wildtype α-synuclein when expressed from chromosomal loci. The present study shows a very different aspect of wild type α-synuclein; it was quite interesting to see the dual nature of α-synuclein. It acts as a pro-apoptotic agent inducing blockage of cell growth and apoptosis (Figures 2 and 3). But α-synuclein seems to act as an anti-apoptotic agent when it is co-expressed with Bax, which is known to be a potent inducer of apoptosis both in dividing and non-dividing cells. Mutant α-synuclein proteins, A30P and A53T, when co-expressed with Bax, do not rescue yeast cells from Bax-induced apoptosis.

Overall, one can conclude that independent of the promoter, α-synuclein expressed from a 2µ plasmid does not have any toxic effects on yeast. It can be concluded that the mutant A30P and A53T α-synuclein proteins, in contrast with the wildtype α-synuclein protein, can not overcome Bax-mediated toxicity in yeast. The inability of these mutants to rescue Bax’s toxicity somehow helps to maintain the levels of mutant α-synuclein proteins and Bax, as seen in Western blots. We would like to suggest that the mutant α-synuclein proteins are the causative agents of PD, responsible for the destruction of neuronal cells. In contrast, the wildtype α-synuclein exerts a protective influence on neurons deletion. However, we have also observed that, even in the course of protecting yeast cells by wildtype α-synuclein, overexpression of wildtype can be damaging too.

The authors declare that there are no competing interests associated with the manuscript.

This work was supported by the De Montfort University.

D.D.A. performed all the experiments. B.C. coordinated the study. B.C. and D.D.A. wrote the manuscript.

DA acknowledges De Montfort University for the support provided.

GAL1p-α-synuclein-HA

GAL1 promoter-driven HA-tagged α-synuclein expression cassette

IAP

inhibitor of apoptosis protein

MMP

mitochondrial membrane potential

NDF

nuclear DNA fragmentation

PD

Parkinson’s disease

ROS

reactive oxygen species

SNARE

SNAP REceptor

SNAP

soluble NSF attachment proteins

TUNEL

terminal deoxynucleotidyl transferase dUTP nick end labeling