A widely accepted hypothesis suggests that intrinsic toxicity of α-synuclein and γ-synuclein is a prerequisite for the demise of dopaminergic neurons in the SNpc of MPTP-treated mice. A logical extrapolation of this presumption would be that SNpc neurons of triple α/β/γ-synuclein null mutant (TKO) mice should be at least as resistant to the toxic effect of MPTP as neurons lacking α-synuclein and γ-synuclein. However, morphometric analysis of the number of TH-positive neurons in the SNpc of wild type (WT) and TKO mice treated with a subchronic MPTP regimen [10, 12], revealed the same degree of neuronal loss for both genotypes (Fig. 1). This result clearly demonstrated that none of the three synucleins are required for MPTP-induced death of dopaminergic neurons in the SNpc and suggests a protecting role of β-synuclein may in the single and double α/γ-synuclein knockout scenarios.
The sensitivity of dopaminergic neurons to MPTP depends on the ability of its active toxic metabolite, MPP+, to enter the cell via the dopamine transporter (DAT)- and to avoid being sequestered in synaptic vesicles by a vesicular monoamine transporter (VMAT-2)-driven mechanism [32, 33]. Thus, dopaminergic neurons with a higher DAT/VMAT-2 ratio, such as SNpc neurons, are more sensitive to MPTP than neurons with a lower ratio, for instance ventral tegmental area (VTA) neurons. We have previously demonstrated that the function of DAT in synapses of SNpc neurons is not affected by the absence of synucleins [25, 34]. This was further confirmed by studies of in vitro dopamine uptake by synaptosomes isolated from the striatum of wild type and TKO mice that revealed no difference in this DAT-dependent uptake between the two animal groups (Additional file 1). These observations suggest that increased sensitivity of midbrain dopaminergic neurons of ТКО mice to MPTP may be related to attenuated VMAT-2-dependent MPP+ uptake by vesicles lacking these proteins. Consistent with the well-documented ability of synucleins to interact with various biological and synthetic membranes, particularly with high curvature vesicles [35–37], we showed that all three family members co-purify with striatal synaptic vesicles despite being predominantly cytosolic proteins in the striatum of wild type mice (Fig. 2a). To check whether vesicle-associated synucleins affect VMAT-2-dependent uptake we compared the ability of synaptic vesicles isolated from the striatum of 5-month old male TKO and wild type mice to take up 3H-dopamine. It has previously been shown that the basal level of cytosolic dopamine in dopaminergic neurons is below 0.1 µM, which is the detection limit of the intracellular patch electrochemistry technique [38]. Therefore, to imitate physiological conditions, uptake was measured with 10 nM dopamine in the reaction mixture. We observed a 38.5 ± 4.71% reduction of tetrabenazine-sensitive (i.e. VMAT-2-dependent) dopamine uptake by synaptic vesicles isolated from the striatum of 5-month old male TKO mice when compared to synaptic vesicles from wild type mice of the same age and sex (Fig. 2b). A similar difference between genotypes was found in an independent set of experiments with synaptic vesicles isolated from ageing 14-month old male (39.2 ± 1.29% reduction) or 18-month old female mice (39.0 ± 4.79% reduction), suggesting that the effect is not age- or sex-dependent. Further analysis demonstrated decreased Vmax (9.97 vs 12.29) and Km (117.4 vs 196.1) values for this uptake in the absence of synucleins. These changes were not due to a decrease in either the number of synaptic vesicles in dopaminergic synapses or levels of VMAT-2 in the striatum of TKO mice [25] and in synaptic vesicles isolated from the striatum of TKO mice (Fig. 2b, inset).
Reduced dopamine uptake efficiency of synuclein-depleted synaptic vesicles and the resultant decrease of the vesicular dopamine pool explains the attenuated response of TKO mice to amphetamine observed in our previous study [25]. Moreover, this functional deficiency should lead to the accumulation of free cytosolic dopamine within the presynaptic terminals because TH activity and dopamine re-uptake by DAT are unaffected in the striatum of TKO mice [25]. The increased striatal DOPAC/dopamine ratio suggests that to prevent toxicity of free cytosolic dopamine, these mice utilise a compensatory increase of dopamine degradation [25].
It is feasible that synucleins differ in their ability to potentiate vesicular uptake of dopamine and other, structurally similar, molecules e.g. MPP+. Thus, functional substitution for the loss of a family member(s) by another, more efficacious member might significantly affect both vesicular dopamine uptake and sensitivity of dopaminergic neurons to MPTP toxicity. The resistance of α/γ-synuclein-deficient ([12] and Fig. 3a), but sensitivity of α/β-, β/γ- and α/β/γ-synuclein-deficient (Figs. 1 and 3a) neurons to this drug suggest that within the family only β-synuclein can efficiently potentiate vesicular uptake. In accordance with this assumption, reduced dopamine uptake, similar to the level of uptake by TKO vesicles, was observed for synaptic vesicles isolated from the striatum of β-synuclein but not α-synuclein or γ-synuclein null mutant mice (Additional file 2). To confirm the pivotal role of β-synuclein in potentiating the efficiency of vesicular uptake, we first compared the sensitivity of SNpc dopaminergic neurons to subchronic MPTP administration in three groups of β-synuclein deficient mice and in three groups of mice lacking the other two synucleins singularly or in combination. In single β-synuclein and both double β/γ-synuclein and α/β-synuclein null mutant mice these neurons were sensitive to the drug to approximately the same degree as neurons of wild type mice (Fig. 3a). In contrast, this protocol of MPTP administration did not cause loss of dopaminergic neurons in the SNpc of α-synuclein, γ-synuclein and α/γ-synuclein null mutant mice (Fig. 3a), consistent with previous observations [10, 12].
Next, we attempted to restore β-synuclein expression in midbrain neurons of TKO mice using in vivo lentiviral delivery of an expression construct. Stereotaxic injection of viral particles into the SNpc region resulted in β-synuclein expression and its transport via the nigrostriatal tract (Additional file 3) to the dorsal striatum where the protein could be detected by Western blotting (Fig. 3b). Vesicular dopamine uptake was significantly higher in vesicles isolated from the ipsilateral than from contralateral striatum (Fig. 3c), suggesting that β-synuclein potentiates this process. To assess whether β-synuclein directly affects vesicular dopamine uptake, we performed the uptake reaction by TKO mouse striata-derived vesicles in the presence of purified recombinant β-synuclein. Preincubation of isolated vesicles (i.e. from P4 fraction) with the recombinant protein before adding 3H-dopamine did not affect the uptake (Fig. 4a). However, when the vesicle-containing S3 supernatant was incubated with recombinant β-synuclein followed by spinning down synaptic vesicles (i.e. obtaining P4 fraction from β-synuclein preincubated S3 supernatant) and using them to measure dopamine uptake, a statistically significant increase was observed compared to the uptake by vesicles from a mock-incubated S3 supernatant (Fig. 4b). This effect was not observed for recombinant α- or γ-synuclein, suggesting that the presence of β-synuclein alongside one or more proteins contained within the S3 supernatant are required for efficient synaptic vesicle dopamine uptake.
In an attempt to provide an insight into how supplementing with exogenous β-synuclein improves dopamine uptake by striatal synaptic vesicles from synuclein-free mice, we assessed the proteome of β-synuclein-bound vesicles by a combination of the crosslink immunoprecipitation (CLIP) and mass spectrometry techniques. Synaptic vesicle-containing S3 supernatant obtained from the striata of TKO mice was preincubated with Strep-tagged human β-synuclein followed by incubation with a cleavable crosslinking agent (DTSSP) and subsequent high-speed centrifugation to spin down synaptic vesicles (for details see Methods). Synaptic vesicles isolated from the same S3 supernatant sample preincubated without β-synuclein but similarly crosslinked were used as a non-specific binding control. Membranes of both control and β-synuclein-bound vesicles were lysed in a non-ionic detergent solution and protein complexes containing β-synuclein were pulled down using Tactin magnetic beads that have high affinity for Strep-tagged proteins. Proteins incorporated in these complexes were eluted from the beads by breaking crosslinking bonds with dithiothreitol, leaving Strep-tagged β-synuclein attached to beads. Protein composition of the resulting CLIP proteome was analysed by mass spectrometry (MS). Proteins identified by the presence of at least two unique peptides in the β-synuclein-CLIP and undetectable in the control CLIP were further filtered using the CRAPome tool (crapome.org) to remove common non-specific proteins often present in MS immunoprecipitation experiments.
Not surprisingly, around half of the 224 proteins included in the final analysis (thereafter, β-synuclein-CLIP proteome) were proteins with predominant localisation to, or function at, neuronal membranous structures and synaptic vesicles in particular (Fig. 5a). Presynapse (GO:0098793) appeared to be the top Cellular Component with a false discovery rate (FDR) of 3.44e-27. Mechanistically, proteins involved in presynaptic functions were also vastly overrepresented in this β-synuclein-CLIP proteome. Gene Ontology analysis revealed Synaptic Vesicle Cycle (GO:0099504) and Neurotransmitter Transport (GO:0006836) as the most represented Biological Processes with FDR of 4.91e-12 and 3.35e-11, respectively. Similarly, Synaptic Vesicle Cycle (MMU4721; FDR 4.85e-12) appeared to have the highest representation between KEGG Pathways, as well as Transmission Across Chemical Synapse (MMU112315) and Neurotransmitter Release Cycle (MMU112310) – between Reactome Pathways (FDR 6.88e-12 and 5.71e-12, respectively).
The improved dopamine uptake after addition of exogenous β-synuclein was observed only when synaptic vesicles were preincubated with β-synuclein in the presence of cytosolic factors but not after these factors were removed by high-speed centrifugation at the final step of vesicle purification (Fig. 4). A logical interpretation of these data is that upon interaction with the synaptic vesicle membrane via its N-terminal lipid-binding domain, β-synuclein attracts certain proteins, principally residing in the cytosol of dopaminergic presynaptic terminals, to the outer surface of these vesicles, and that the resulting molecular complexes improve the efficiency of the VMAT-2 transporter. The absence of VMAT-2 itself between proteins included in the β-synuclein-CLIP proteome provides additional evidence that its direct interaction with β-synuclein is not a plausible reason for this improved efficiency.
A number of proteins, which are not intrinsic constituents of synaptic vesicles at any stage of their presynaptic cycle, have been previously detected in brain synaptic vesicle fractions. It has been suggested that some of these proteins can affect vesicular uptake of dopamine, for example by increasing local availability of ATP in the case of glycolytic enzymes [39] or direct modulation of VMAT-2 activity [40–42]. Amongst the various, principally cytosolic, proteins identified in our analysis of the β-synuclein-CLIP proteome, the most probable candidates for augmentation of synaptic dopamine uptake are TH and AADC (also known as DOPA decarboxylase, DDC). Both proteins were detected in the β-synuclein-CLIP but not in the control CLIP samples by Western blot, confirming the findings of the MS analysis (Fig. 5b). These proteins are not only involved in presynaptic dopamine production but can also form a transient complex with VMAT-2, which creates spatial coupling of dopamine synthesis and loading into synaptic vesicles [43, 44]. Although this coupling cannot explain the improved 3H-dopamine uptake in the in vitro assay nor improved synaptic vesicle uptake of MPP+ in the striatum of MPTP-treated mice, the β-synuclein-triggered formation of the TH/AADC/VMAT-2 complex might have an allosteric effect on transporter function.
Although the TH/AADC/VMAT-2 complex is the most obvious scenario, a contribution from other proteins interacting with the vesicle-bound β-synuclein cannot be excluded and it is feasible that improved function of VMAT-2 in the presence of β-synuclein is the result of a cumulative effect of several multiprotein interactions rather than of any one particular complex. Indeed, our data suggest that β-synuclein, via interaction with a number of vesicular and cytosolic proteins, can potentiate formation of various molecular complexes on the surface of synaptic vesicles. This is similar to α-synuclein, which is known to interact with a variety of proteins and function as a chaperone or scaffold for the assembly of multiprotein complexes on the surface of synaptic vesicles, implicating α-synuclein in a number of molecular processes at several stages of the synaptic vesicle cycle [1–3, 5, 45–47]. However, these do not include potentiation of vesicular uptake and therefore, α-synuclein, as well as γ-synuclein, are unable to compensate for the loss of this particular function of β-synuclein. All three members of the synuclein family can interact with phospholipids at the outer surface of synaptic vesicles via their conserved N-terminal repeat domain, but the number of functional multiprotein complexes each member can form on a given vesicle is limited. It is feasible that in the presence of all three synucleins (i.e. in the synaptic terminals of dopamine neurons of WT animals), this competition for space results in a limited number of β-synuclein-triggered complexes formed at the vesicle surface. This is not sufficient to potentiate VMAT-2-dependent uptake to the level that can efficiently sequester MPP+ in synaptic vesicles and thus, prevent its toxic effect. However, in the absence of one or two other family members, β-synuclein occupies vacant sites on the vesicle surface, triggering the formation of more complexes that efficiently potentiate vesicular uptake of dopamine and MPP+, making dopaminergic neurons more robust to MPTP-induced toxicity.