Polo-like kinase 2 inhibition reduces serine-129 phosphorylation of physiological nuclear alpha-synuclein but not of the aggregated alpha-synuclein

Accumulation of aggregated alpha-synuclein (α-syn) is believed to play a pivotal role in the pathophysiology of Parkinson’s disease (PD) and other synucleinopathies. As a key constituent of Lewy pathology, more than 90% of α-syn in Lewy bodies is phosphorylated at serine-129 (pS129) and hence, it is used extensively as a marker for α-syn pathology. However, the exact role of pS129 remains controversial and the kinase(s) responsible for the phosphorylation have yet to be determined. In this study, we investigated the effect of Polo-like kinase 2 (PLK2) inhibition on formation of pS129 using an ex vivo organotypic brain slice model of synucleinopathy. Our data demonstrated that PLK2 inhibition has no effect on α-syn aggregation, pS129 or inter-neuronal spreading of the aggregated α-syn seen in the organotypic slices. Instead, PLK2 inhibition reduced the soluble pS129 level in the nuclei. The same finding was replicated in an in vivo mouse model of templated α-syn aggregation and in human dopaminergic neurons, suggesting that PLK2 is more likely to be involved in S129-phosphorylation of the soluble physiological fraction of α-syn. We also demonstrated that reduction of nuclear pS129 following PLK2 inhibition for a short time before sample collection improves the signal-to-noise ratio when quantifying pS129 aggregate pathology.


98
Immediately before injection, an aliquot of S129A PFFs was thawed at room temperature (RT) 99 and sonicated for 30 seconds using Branson Sonifier 250 with settings adjusted to 30% duty cycle, 100 output control 3. The sonicator is customized and equipped with a water jacket cooling system to 101 avoid sample heating during sonication. OHSCs were microinjected with S129A PFFs or sterile 102 phosphate buffered saline (PBS) at DG after 7 days in culture. Microinjection pipettes (item 103 #1B200F-4 (with Filament), WPI) were pulled using a micropipette puller (P-1000, Sutter 104 Instrument). For injection, a Pulse Pal v2 (#1102) was set to phase 1 voltage 5V, phase 1 duration 105 0.01 seconds, pulse interval 0.5 seconds. The pipette was loaded using Eppendorf microloader 106 pipette tips (ThermoFisher). A final volume of 0.1 µL of either S129A PFFs (1 mg/mL) or PBS 107 was injected at DG under microscopic guidance as described in [13].

233
For evaluation of a possible effect of the PLK2i-treatment on spreading of aggregation throughout 234 the hippocampal slices (Fig. 3), the previous images were stitched together using the Stitching 235 plugin in Fiji (Fiji Is Just ImageJ, NIH) [32], and the DG and CA1 regions were defined manually 236 from the DAPI-staining (Fig. 3a). First, the DG was bound by a rectangle, and second, projections 237 along one side of the rectangle and through one diagonal were used to define the limits of the CA1 238 region. The subsequent definition of aggregates and analysis of aggregate areas was performed as 239 described above, based on the MJFR-14-6-4-2 staining.

240
For M83 mouse brain, 2-5 images per brain region were taken randomly with an X10 (2 images,

241
region at frontal cortex), an X40 oil (3 images, region at DG and CA1 of the hippocampus) or an  For primary hippocampal neurons (Supplementary fig. 4)

278
PLK2 is reported to be responsible for phosphorylation of α-syn in murine brain and may

286
To assess the effect of PLK2 inhibition on PFF-seeded aggregation using OHSCs, slices were 287 treated continuously with 20 µM PLK2i (compound 37), from 24 hours prior to S129A PFFs 288 injection until the end of experiment. PLK2i was added at each change of medium to ensure 289 continuous inhibition of PLK2 function during the experiment (Fig. 1a). Seven days following 290 S129A PFF-injection, slices were collected for biochemical analysis and immunostaining, at which 291 time point the organotypic slice cultures normally express robust levels of total and pS129 α-syn 11 292 (Fig. 1b). This time point allows the templating of α-syn aggregates in neurons at the site of 293 injection at DG and the spreading of α-syn aggregate pathology to neurons in the CA3 and CA1 294 regions as described earlier and the use of S129A PFF can ensure that the signals detected using 295 antibodies against pS129 is detecting only the de novo generated aggregates in the slices and not 296 detecting the injected PFF materials as described earlier in [13].

297
Immunoblotting of slice cultures at 7 days post injection (dpi) showed that PLK2i treatment 298 increased the level of mouse α-syn in the soluble fraction (Fig. 1c & e) while the level of likely 299 membrane associated pS129 in the RIPA-soluble fraction was reduced in a PFF-independent 300 manner ( Fig. 1c & d). By contrast, PLK2 inhibition did not affect the formation of insoluble α-syn 301 aggregates as detected by immunoblotting using rodent-specific α-syn antibody D37A6 and pS129 302 extracted by 7M urea from the of RIPA-insoluble material of PFF-injected slice cultures (Fig. 1f). α-syn following S129A PFF-microinjection. Aggregates were detected by the aggregate-specific 325 α-syn antibody, MJF-14, and pS129-α-syn (Fig. 2a). Staining for pS129 using the monoclonal 326 antibody 11A5 revealed two forms of pS129 immunoreactive signals. One is bright and intense, 327 aggregation-specific pS129-signals that co-localizes with MJF-14-signal and is only detected in 328 PFF-injected slices (Fig. 2a, arrows). The other type is a fainter, diffuse non-aggregate specific 329 pS129-signal, which is abundant in nuclei and co-localizes with DAPI-signal but never with MJF-330 14-signal, and is detected in both PFF and PBS-injected slices (Fig. 2a, arrowheads) but not in the 331 PLK2i treated slices. This indicates that PLK2i-treatment significantly reduced the diffuse, nuclear 332 pS129-signal in both PFF-and PBS-injected slice cultures (Fig. 2a, e). In contrast, PLK2 inhibition 333 did not influence PFF-induced α-syn aggregation, as detected with aggregate specific MJF-14 334 antibody (Fig. 2b, c), and it did not prevent or reduce S129-phosphorylation of the aggregates 335 generated in axons (Fig. 2d). The lack of dissociation between MJF-14 and pS129-signals in the 336 PFF-induced aggregates following the PLK2i treatment demonstrates that aggregate-associated 337 pS129 is unaffected by PLK2 inhibition (Fig. 2). Quantification analysis revealed an 338 approximately 30% reduction of nuclear pS129 following PLK2i-treatment without reducing 339 aggregate-specific pS129 inside MJF-14 signals (Fig. 2d & e). This led to an approximate doubling 340 of the signal-to-noise ratio of aggregate-specific pS129 signal, thereby facilitating the study of 341 pS129 positive α-syn aggregates without interference of non-aggregate-specific pS129 (Fig. 2f). (white boxed area) shows nuclear pS129 signal that co-localizes with DAPI (arrowheads). ii. The 348 nuclear pS129-staining was removed following PLK2i treatment. iii. S129A PFF-induced α-syn 349 aggregates, phosphorylated at S129, are detected with MJF-14 and pS129 antibodies (arrows).

350
pS129-staining of non-aggregated α-syn is also seen, predominantly located in the nuclei 351 (arrowheads). iiii. PLK2i treatment shows no effect on the S129-phosphorylation of aggregated α-

PLK2 inhibition does not hinder inter-neuronal spreading of PFF-induced α-syn aggregates 365
To test the influence of PLK2 inhibition on the spreading of PFF-induced α-syn aggregate 366 pathology, aggregate signals at CA1 region was identified for quantification (Fig. 3a). PLK2i-367 treatment of slices showed no significant influence on aggregates formed at CA1 following the 368 PFF injection in the DG when compared to DMSO-treated slices and their phosphorylation as 369 determined by aggregate-and pS129-specific antibodies (Fig. 3b, arrow), but instead only reduced 370 the nuclear pS129 signals (arrow head) (Fig. 3b-d). A tendency towards increased pathology load 371 was observed, as also seen in Fig. 2b, although this was not statistically significant. To obtain a 372 measure for the relative inter-neuronal spreading in each slice, the aggregate signal at CA1 was 373 normalized to the aggregate signal at DG (Fig. 3e).  (Fig. 4). PLK2i for 48 hours before sacrifice. Aggregate pathology in the models was initiated by injection 418 of mouse recombinant PFFs in the hind limb gastrocnemius muscle at 12 months [29].

419
Immunostaining of the brain using anti-pS129 antibody showed a significant reduction in nuclear 420 pS129 intensity as demonstrated in representative images from hippocampus and frontal cortex 421 (Fig. 5a-c). The level of reduction varied between regions, with nuclear pS129 α-syn of pyramidal 422 neurons of the CA1 region in the hippocampus appearing particularly sensitive to PLK2 inhibition 423 (Fig. 5a, d). In contrast, the frontal cortex appeared to contain both PLK2i-sensitive and PLK2i-424 resistant nuclear pS129-signals (Fig. 5c, f). Quantification the pS129-positive aggregated fibrillary 425 signals detected in the hind brain region showed that treatment of the mice with PLK2i for 48 426 hours -unsurprisingly -had no effect compared to the control group, considering the short time 427 of the treatment (Supplementary fig. 3a & b). However, PLK2 inhibition strikingly increased the 428 signal-to-noise ratio, facilitating the detection of aggregate-specific signals in the PLK2i treated 429 mice (Supplementary fig. 3c). represent the mean ± SD, n = 5 mice in each group. * p < 0.05, ** p < 0.01, as determined by an 442 unpaired Student's T test.

444
We also conducted a short-term PLK2 inhibition on cultured human dopaminergic neurons, where 445 aggregation of endogenous α-syn was initiated by adding α-syn S129A PFFs (Fig. 6a). Addition 446 of PLK1-3 inhibitor (BI2536) four hours before fixing the cells effectively reduced nuclear pS129 447 intensity, especially in the tyrosine hydroxylase-positive neurons, which displayed intense nuclear 448 pS129-staining in DMSO-treated cultures (Fig. 6b & c). No modulation of aggregate levels 449 following PLK2 inhibition was identified in the cultures treated with S129A PFFs (Fig. 6b & d).

450
Collectively, these results highlight the ability of short-term PLK2 inhibition prior to fixation to To further explore how short a period of PLK2 inhibition is needed to reduce nuclear pS129 levels, 470 we tested PLK inhibition from 10 minutes to 4 hours prior to fixation and analysis in primary 471 hippocampal neurons (Supplementary fig. 4a). As little as 30 minutes treatment with 10 nM 472 BI2536 prior to fixation was sufficient to decrease nuclear pS129 staining to a minimal level,  be due to the use of models were either α-syn or the studied kinases are over expressed or expressed 485 in a mutated forms, including our own earlier report where mutated form of α-syn (S129G) that 486 cannot be phosphorylated at S129 residue were expressed using viral vectors in organotypic slices 487 made from SNCA genes knock out pups [13].

488
In the current paper, we investigate whether PLK2, which has been identified as an efficient S129- and BI2536 was equally able to reduce nuclear pS129 staining in M83 +/and human neuronal 513 models following treatment for 48 hours and 4 hours prior to tissue or cell collection, respectively.

514
In none of the models did the short-term PLK2 inhibition affect the phosphorylation of aggregated Naturally, attention must be paid to other effects of the short-term PLK2 inhibition that was not 534 tested here, but if the treatment is brief and the primary read-out is α-syn aggregates, which are 535 considered fairly stable structures as demonstrated in the tested models in this study, then this 536 concern should be of minor importance.

537
The presence of S129-phosphorylated α-syn in the nucleus has been questioned due to the reported reporting the nuclear localization. Although our data do not identify the roles of nuclear pS129 α-549 syn, they provide a novel experimental strategy to investigate this enigmatic α-syn species that has 550 been associated to processes covering histone acetylation, neurotoxicity, transcriptional regulation The findings of this study demonstrate that PLK2 is involved in significant S129-phosphorylation 554 of physiological α-syn but not the phosphorylation of serine-129 on aggregated α-syn. Moreover, 555 short-term PLK2 inhibition can be used as an easy experimental procedure to facilitate specific 556 detection of aggregated α-syn in different models of templated α-syn pathology.