Intra-mitochondrial proteostasis is directly coupled to alpha-synuclein and Amyloid β 1-42 pathology

Mitochondria have long been implicated in Parkinson’s disease (PD), however, it is not clear how mitochondrial impairment and alpha-synuclein pathology are coupled. We report here that intra-mitochondrial protein homeostasis plays a major role in alpha-synuclein fibril elongation, as interference with intra-mitochondrial proteases and mitochondrial protein import significantly aggravate alpha-synuclein aggregation. In contrast, direct inhibition of mitochondrial complex I, increase in intracellular calcium concentration or formation of reactive oxygen species (ROS), all of which have been associated with mitochondrial stress, did not affect alpha-synuclein pathology. We further demonstrate that similar mechanisms are involved in Amyloid β 1-42 (Aβ42) aggregation, suggesting that mitochondria are directly capable of influencing cytosolic protein homeostasis of aggregation-prone proteins.


Introduction
mitochondria is higher at the beginning of the experiment (t0) than in the control group (Aβ42 218 + mito 3538 +/-15 ps compared to Aβ42 control with 3380 +/-93 ps), since in the control Aβ42 219 starts to aggregate immediately upon preparation. The decrease in Aβ42 Hylite TM  showing that mitochondrial proteases can influence aggregate dissolution and that 233 aggregation-prone proteins are directed to mitochondrial import (42). There is one report 234 showing mitochondrial import of alpha-synuclein (47), but it is discussed critically that 235 aggregation-prone proteins like alpha-synuclein and Aβ42 are directly imported into 236 mitochondria (48,49). Thus, in order to prove whether we see alpha-synuclein residing within 237 mitochondria, we immuno-gold labelled YFP-alpha-synuclein in SH-SY5Y cells, and found 238 specific staining within mitochondria, which was mainly located at the inner mitochondrial 239 membrane ( Fig. 6A and Supplementary Fig. 2). In addition, we isolated mitochondria from 240 wild-type adult rat brain and probed them for the presence of endogenous alpha-synuclein 241 after protein K (PK) digestion. We saw that alpha-synuclein was still present after PK 242 treatment, indicating that alpha-synuclein resides within the organelle since PK is not able to 243 degrade proteins protected by organelle membranes. This is further supported, since 244 incubation with 0.1% Triton X-100 during PK treatment, which is capable to solubilize 245 mitochondrial membranes (50), enabled complete alpha-synuclein degradation ( Fig. 6B and 246 C). 247 248 Since the above results indicate that alpha-synuclein is localized to mitochondria, we 249 hypothesized that inhibition of mitochondrial protein import might have a similar effect on 250 alpha-synuclein pathology as the inhibition of proteases. Thus, using MitobloCK-6, a small 251 molecule inhibitor of protein translocation into mitochondria (51), we equally observed 252 increased alpha-synuclein seeding in YFP-alpha-synuclein SH-SY5Y cells (Fig. 6D). Testing 253 MitobloCK-6 on Aβ42-mCherry overexpressing HEK cells also showed increased aggregation 254 The effect of BAPTA-AM on alpha-synuclein seeding took us to study the role of mitochondrial 261 proteostasis on the aggregation of amyloidogenic proteins. We demonstrate here, that 262 inhibition of the mitochondrial proteases HtrA2 and Lon, as well as inhibition of mitochondrial 263 protein import enhances alpha-synuclein pathology. However, downstream effects of 264 mitochondrial dysfunction, induced without mitochondrial integrity disturbance, did not 265 12 recapitulate the increased seeding. Inhibition of HtrA2 as well as inhibition of mitochondrial 266 protein import also increased Aβ42 pathology, and the overexpression of HtrA2 was able to 267 decrease Aβ42 aggregation notably. It was reported recently that mitochondria are able to 268 influence the degradation and protein homeostasis of cytosolic proteins, which was shown in 269 yeast cells upon heat shock (42), however mitochondria may also play an important role for 270 the degradation of amyloidogenic proteins, since mitochondrial proteostasis seems to be 271 clearly coupled to the pathology of alpha-synuclein and Aβ42. HtrA2 appears of particular 272 interest, since it has previously been linked to PD by genetics (52-56) and shows a 273 neuroprotective effect upon overexpression in mice (57,58). 274

275
So far the effect of amyloid proteins on mitochondria has been interpreted only as a secondary 276 pathological hallmark, with alpha-synuclein as well as Aβ exacerbating mitochondrial 277 dysfunction (48,49,(59)(60)(61)(62)(63). But amyloidogenic proteins might be directed to mitochondrial 278 uptake deliberately, which then disrupts overall mitochondrial function if the uptake is 279 overloaded. So vice versa, an initial failure in mitochondrial function, i.e. by complex I 280 inhibition or upon disturbance of mitophagy, could eventually lead to increased levels of 281 alpha-synuclein, having important implications for sporadic forms of the disease. Our recently 282 published review provides more insights on the mitochondrial uptake of alpha-synuclein and 283 Aβ, on the interaction with mitochondrial translocases as well as background on mitochondrial 284 proteases (64). 285

286
There still remains the argument that inhibition of mitochondrial proteases just causes an 287 unspecific mitochondrial dysfunction which then per se leads to increased alpha-synuclein 288 aggregation. However, it seems that this effect is not mediated via the known downstream 289 13 events of mitochondrial dysfunction. So it has been shown that ATP depletion is not able to 290 increase alpha-synuclein aggregation, as seen in our study using short term complex I 291 inhibition via MPP + reducing ATP without causing major mitochondrial fragmentation and also 292 previously when ATP levels were reduced independently from mitochondrial respiration using 293 2-Desoxyglucose, which inhibits glycolysis (65). Also increased calcium concentrations, when 294 induced acutely via calcium influx through the plasma membrane using the ionophore 295 ionomycin, did not influence alpha-synuclein seeding. This seems to stand in contrast to our 296 previous study (25), where we have shown that calcium affects alpha-synuclein aggregation 297 in-vitro. However, here mainly the nucleation rate was increased, thus calcium may still 298 contribute to PD via alpha-synuclein seed formation. Oxidative stress has been discussed as a 299 likely mechanism, since antioxidants are able to reduce dopaminergic neuron death and 300 alpha-synuclein accumulation after complex I inhibition (65), however also a general 301 protective impact on mitochondria might play a role. 302 303 Though, this does not mean that complex I inhibition, calcium dysregulation and oxidative 304 stress do not play an important role during the course of the disease. Chronic complex I 305 inhibition has clearly been shown to lead to dopaminergic neuron death and alpha-synuclein 306 accumulation (12-16) and is a major factor implicating mitochondrial dysfunction in sporadic 307 Parkinson's disease. Chronic complex I inhibition would lead to reduced mitochondrial fitness 308 and in a similar way high calcium loads do when specific neuronal subtypes are subjected to 309 Flp-InTM expression vector as described previously (44,69). Cells were maintained in DMEM 335 high glucose media (Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS), 2 mM 336 15 glutaMAX, and 1 % antibiotic-antimycotic (all Thermo Fisher Scientific). Cells were grown at 337 37°C under a 5% CO2 atmosphere. Cells were plated at 35 000 cells per well in NUNC 24 well 338 plates, and construct expression was induced for 3 days using media above with 1 µg/mL 339 tetracycline (Sigma Aldrich) added. All cell lines were tested for mycoplasma contamination 340 were acquired using PeakForce Quantitative Nanomechanical Property mapping mode with 381 ScanAsyst-Fluid+ probes (Bruker AXS GmbH). Images were flattened and exported using 382 NanoScope Analysis software, version 1.8. 383

Preformed fibril (PFF) assay 385
For the induction of alpha-synuclein seeding, YFP-alpha-synuclein overexpressing SH-SY5Y 386 cells were incubated with sonicated preformed alpha-synuclein fibrils as described by Luk et 387 al. (26). Briefly, cells plated in MatTek dishes were washed with Neurobasal medium and 388 subsequently changed to 500 µL Neurobasal medium supplemented with 2 % B27 and 0.5 mM 389 GlutaMAX (all Thermo Fisher Scientific). Cells were preincubated for 1 hour, either using 0. Cells were fixed as described above, blocking and permeabilisation were performed using 5 % 423 donkey serum in 0.05 % Tween-20 in phosphate-buffered saline (PBS) for 1 h. Primary 424 antibodies were incubated overnight at 4°C, followed by 5 washes with PBS. Secondary 425 antibodies were incubated for 1 hour at room temperature, followed by 5 washes with PBS. Germany). Experiments were repeated three times with four replicates for each condition. 487 488

Mitochondrial fragmentation 489
To label mitochondria SH-SY5Y cells were incubated overnight with 1:1000 CellLight TM 490 Mitochondria-RFP (Thermo Fisher Scientific) and imaged by SIM or a widefield microscope for 491 quantification. Images were taken randomly by automated imaging of a grid, images were 492 analyzed from 3 biological repeats. The mitochondrial length was evaluated using the NIEL 493 Mito algorithm (86,87).

Statistics 571
Statistical analysis was performed using GraphPad Prism 6.07 (GraphPad Software, Inc., La 572 Jolla, CA, USA). Values are given as mean ± SEM unless otherwise stated. Normal distribution 573 was tested using Shapiro-Wilk test. Two-tailed unpaired t-test was used upon normal 574 distribution, two-tailed Mann-Whitney U test was used when no normal distribution was 575 given. For multiple comparisons either one-way ANOVA with Dunnett's post hoc correction 576 upon normal distribution or Kruskal-Wallis test with Dunn's multiple comparison when no 577 normal distribution was given were performed. Significance was considered at p < 0.05. 578 579 Data availability. 580 All relevant data are available from the corresponding authors.

Conflict of interest 602
The authors declare no conflict of interest. 603 604 605