Quantitative Characterization of α-Synuclein Aggregation in Living Cells through Automated Microfluidics Feedback Control

摘要

class="head no_bottom_margin" id="sec1title">Introductionα-Synuclein, encoded by the SNCA gene, is a small (14.5 kDa), intrinsically disordered protein expressed abundantly in a healthy brain. The precise physiological functions of α-synuclein remain poorly understood (), although recent findings point to a role in vesicle trafficking and synaptic physiology (, ). In the human brain, an abnormal increase of α-synuclein expression levels may result in the aggregation of the protein into large complexes and amyloidogenic fibrils with the formation of intraneuronal proteinaceous inclusions known as Lewy bodies (), linked to the Parkinson’s disease (PD) pathogenesis (). Although the matter is still the subject of debate, it is thought that inclusions in the cell are generated by the impairment of degradative pathways and activation of the protein quality-control system (). The mechanisms underlying the formation of protein aggregates seem to be concentration dependent (). Indeed, either duplication or triplication of the wild-type (WT) α-synuclein gene locus is sufficient to cause familial PD (, , ). Moreover, missense mutations in the SNCA gene cause early-onset (A53T, E64K, A30P, G51D, and A53E) and late-onset (H50Q) forms of PD (, , , , , href="#bib38" rid="bib38" class=" bibr popnode">Kiely et al., 2013, href="#bib51" rid="bib51" class=" bibr popnode">Pasanen et al., 2014, href="#bib45" rid="bib45" class=" bibr popnode">Martikainen et al., 2015).Cell-free, cellular, and animal models of PD have been developed to study the formation of inclusions (href="#bib72" rid="bib72" class=" bibr popnode">Visanji et al., 2016, href="#bib39" rid="bib39" class=" bibr popnode">Koprich et al., 2017, href="#bib41" rid="bib41" class=" bibr popnode">Lázaro et al., 2017). Pioneering studies have dissected aggregate and fibril formation in cell-free systems using purified α-synuclein protein (href="#bib27" rid="bib27" class=" bibr popnode">Giasson et al., 1999, href="#bib12" rid="bib12" class=" bibr popnode">Conway et al., 1998). These earlier in vitro studies were semiquantitative in that they did not quantify threshold concentrations for aggregation nor the difference between WT and mutant α-synuclein proteins. Subsequent in vitro studies, building on these previous works, have now precisely quantified the molecular steps of α-synuclein fibril formation and rate constants of associated reactions, thus greatly contributing to current understanding of α-synuclein pathobiology (href="#bib28" rid="bib28" class=" bibr popnode">Giehm et al., 2011, href="#bib9" rid="bib9" class=" bibr popnode">Cohen et al., 2011, href="#bib10" rid="bib10" class=" bibr popnode">Cohen et al., 2012, href="#bib4" rid="bib4" class=" bibr popnode">Buell et al., 2014, href="#bib26" rid="bib26" class=" bibr popnode">Garcia et al., 2014, href="#bib44" rid="bib44" class=" bibr popnode">Lorenzen et al., 2014, href="#bib24" rid="bib24" class=" bibr popnode">Galvagnion et al., 2015, href="#bib25" rid="bib25" class=" bibr popnode">Galvagnion et al., 2016, href="#bib21" rid="bib21" class=" bibr popnode">Flagmeier et al., 2016, href="#bib33" rid="bib33" class=" bibr popnode">Iljina et al., 2016). These in vitro studies have also shown that α-synuclein aggregation kinetics are strongly affected by the presence of lipid vesicles, thus highlighting the importance of studying such processes in whole cells, because the cellular environment is much more complex than the commonly used in vitro conditions (href="#bib21" rid="bib21" class=" bibr popnode">Flagmeier et al., 2016, href="#bib24" rid="bib24" class=" bibr popnode">Galvagnion et al., 2015).Biological processes involved in α-synuclein inclusion formation and clearance are well conserved across evolution, hence yeast can be used to elucidate the molecular basis of the human disease and to screen for therapeutic drugs (href="#bib46" rid="bib46" class=" bibr popnode">Menezes et al., 2015, href="#bib64" rid="bib64" class=" bibr popnode">Schneider et al., 2018). Since its inception (href="#bib49" rid="bib49" class=" bibr popnode">Outeiro and Lindquist, 2003), the yeast PD model with heterologous expression of α-synuclein has been successfully used not only to study molecular mechanisms of the PD but also for high-throughput drug and genetic screenings (href="#bib76" rid="bib76" class=" bibr popnode">Zabrocki et al., 2008, href="#bib46" rid="bib46" class=" bibr popnode">Menezes et al., 2015, href="#bib8" rid="bib8" class=" bibr popnode">Chen et al., 2017). In this model, α-synuclein is expressed from the galactose-inducible promoter, and protein inclusions form with ensuing growth defects and cell death (href="#bib49" rid="bib49" class=" bibr popnode">Outeiro and Lindquist, 2003, href="#bib13" rid="bib13" class=" bibr popnode">Cooper et al., 2006, href="#bib54" rid="bib54" class=" bibr popnode">Petroi et al., 2012).The main limitation of the yeast PD model is that despite forming α-synuclein cytoplasmic inclusions (href="#bib49" rid="bib49" class=" bibr popnode">Outeiro and Lindquist, 2003, href="#bib13" rid="bib13" class=" bibr popnode">Cooper et al., 2006), which are also found in human neurons (href="#bib27" rid="bib27" class=" bibr popnode">Giasson et al., 1999, href="#bib13" rid="bib13" class=" bibr popnode">Cooper et al., 2006), these are not comprised of insoluble α-synuclein amyloid fibrils as found in Lewy bodies (href="#bib67" rid="bib67" class=" bibr popnode">Soper et al., 2008). Rather, inclusions in yeast consist of clusters of vesicles that contain α-synuclein monomers, as well as α-synuclein aggregates formed by large oligomeric species (href="#bib67" rid="bib67" class=" bibr popnode">Soper et al., 2008, href="#bib63" rid="bib63" class=" bibr popnode">Sancenon et al., 2012, href="#bib70" rid="bib70" class=" bibr popnode">Tenreiro et al., 2014, href="#bib46" rid="bib46" class=" bibr popnode">Menezes et al., 2015). Interestingly, soluble oligomers of α-synuclein are also enriched in PD patients, and they likely represent an early aggregated form of the protein that over time transitions to much larger, insoluble aggregates and amyloid fibrils (href="#bib65" rid="bib65" class=" bibr popnode">Sharon et al., 2003).Because α-synuclein controls vesicular dynamics and recycling in neurons, its basic functions seem to be maintained when the protein is expressed in yeast, thus making yeast a relevant model to study the biology and pathobiology of α-synuclein (href="#bib76" rid="bib76" class=" bibr popnode">Zabrocki et al., 2008).So far, the process of α-synuclein aggregation and inclusion formation in vivo in the yeast PD model has been characterized only qualitatively in terms of the number of integrated genome copies because of technological limitations (href="#bib46" rid="bib46" class=" bibr popnode">Menezes et al., 2015, href="#bib49" rid="bib49" class=" bibr popnode">Outeiro and Lindquist, 2003, href="#bib57" rid="bib57" class=" bibr popnode">Popova et al., 2015). Hence, the most in-depth characterization currently available is that three copies of WT α-synuclein, versus two copies of A53T α-synuclein, are needed to observe α-synuclein inclusions in the yeast PD model (href="#bib54" rid="bib54" class=" bibr popnode">Petroi et al., 2012).Here, we overcame current technical limitations enabling characterization of the yeast PD model quantitatively by dynamically regulating α-synuclein protein expression over time in the very same cell population, that is, without the need of using strains with different numbers of genomic integrations of α-synuclein. In so doing, we demonstrated that α-synuclein inclusion formation in yeast is strictly concentration dependent, but not time dependent. We precisely measured both the WT and disease-associated A53T α-synuclein protein aggregation thresholds, we studied the effects of inclusions on cell-cycle progression, and we dissected the contributions of proteasomal and autophagic pathways on the dynamics of A53T α-synuclein inclusions’ clearance.In addition to contributing to the biology of PD, our results have relevance for therapeutic applications because the yeast PD model is extensively used in large-scale screening of therapeutic small molecules and modifier genes (href="#bib13" rid="bib13" class=" bibr popnode">Cooper et al., 2006, href="#bib8" rid="bib8" class=" bibr popnode">Chen et al., 2017).