Inhibitor treatment of peripheral mononuclear cells from Parkinson’s disease patients further validates LRRK2 dephosphorylation as a pharmacodynamic biomarker

Activating mutations in leucine-rich repeat kinase 2 (LRRK2) are strongly associated with increased risk of Parkinson’s disease (PD). Thus, LRRK2 kinase inhibitors are in development as potential Parkinson’s disease therapeutics. A reduction in the constitutive levels of phosphorylation on leucine-rich repeat kinase 2 (LRRK2) is currently used to measure target engagement of LRRK2 kinase inhibitors in cell and animal models. We aimed to determine if reduced phosphorylation of LRRK2 following inhibitor treatment is also a valid measure of target engagement in peripheral mononuclear cells from Parkinson’s disease patients. Peripheral mononuclear cells from idiopathic Parkinson’s disease patients and controls were treated ex vivo with two structurally distinct inhibitors of LRRK2, at four different doses, and immunoblotting was used to assess the reduction in LRRK2 phosphorylation at Ser910, Ser935, Ser955 and Ser973. Both inhibitors showed no acute toxicity in primary cells and both inhibitors reduced the constitutive phosphorylation of LRRK2 at all measured residues equally in both control and Parkinson’s disease groups. Measuring the reduction in LRRK2 phosphorylation resulting from LRRK2 kinase inhibition, is thus a valid measure of acute peripheral target engagement in Parkinson’s disease patients. This is important if LRRK2 kinase inhibitors are to be used in a clinical setting.

Scientific RepoRts | 6:31391 | DOI: 10.1038/srep31391 It is important to note however, that all four of these residues are not direct LRRK2 auto-phosphorylation sites. Rather, their phosphorylation is regulated indirectly by LRRK2 kinase activity, likely via intermediate signaling kinases and/or phosphatases 13,[17][18][19] . This introduces a level of complexity into the pharmacodynamic assay and thus careful validation is required. This is particularly important if further translation of LRRK2 inhibitors from cellular studies and animal models to human clinical trials is warranted. Therefore, in the present study, we have investigated the dose-dependent decrease in the constitutive phosphosphorylation of LRRK2 Ser910, Ser935, Ser955 and Ser973 in primary human peripheral blood mononuclear cells from idiopathic PD patients and matched controls. We show using structurally distinct inhibitors of LRRK2 ex vivo, that the reduction of LRRK2 phosphorylation at all measured residues is similar between the PD and control groups. Moreover, LRRK2 inhibitors had no cytotoxic effects on primary immune cells. These results suggest that the reduction in LRRK2 phosphorylation seen with inhibitor treatment may be a viable peripheral pharmacodynamic biomarker for potential clinical trials of LRRK2 kinase inhibitors.

Methods
Viability and toxicology assays using PBMCs from healthy blood donors. Peripheral blood mononuclear cells (PBMCs) were isolated from buffy coat obtained from healthy volunteer blood donors to the Red Cross blood service by layering onto ficoll paque plus (GE healthcare) and centrifuging at 400 × g for 30 min. All studies with buffy coat were approved by the University of NSW human research ethics advisory panel (reference #HC14226) and the methods carried out in accordance with the approved guidelines and regulations. Cell count and viability by trypan blue exclusion was performed with an automated cell clounter (Countess, Life Technology) and 2 × 10 6 cells were used for experiments. PBMC's were then treated with LRRK2 inhibitors, dissolved in DMSO, PF06447475 (Sigma) 20 , GSK2578215A (Tocris) 21 , LRRK2-IN1(a kind gift from Dario Alessi) 15 at 0.5 μ m for 24 h. Untreated and DMSO alone treatments were included as controls. The release of lactate dehydrogenase in the tissue culture media was measured using a CytoTox 96 ® Non-Radioactive Cytotoxicity Assay (Promega) according to the manufacturer's protocol. ATP levels were measured using a CellTitre-Glo ATP assay (Promega), and plates for both ATP and LDH assays were read on a Polarstar Omega plate reader (BMG labtech). Apoptosis measurements were carried out according to the manufacturer's protocol using an Annexin V/PI kit (Miltenyi Biotech). At least 20,000 events were acquired with a FacsCANTO II flow cytometer (BD Biosciences) and analysed using FlowJo version 10.0.7 (Treestar Inc.). Treated cell pellets were also collected for immunoblot analysis of total LRRK2 levels as described below.
Subject recruitment. Subjects were recruited with informed consent and the study was approved by the Macquarie University Human Research Ethics Committee (reference 5201100874). Some additional control subjects were also recruited with informed consent through the Forefront program at Neuroscience Research Australia (approved by South Eastern Sydney Local Health District Human Research Ethics Committee, reference 10/92). All methods were performed in accordance with relevant guidelines and regulations. Clinic visits occurred between 9-11 am to minimize the impact of diurnal variation and subjects with less than 8 years from initial symptom onset were chosen. The control group predominantly comprised of unaffected spouses of PD subjects. Patients with a strong family history or early onset PD were excluded and no patient or control subject had an immune, inflammatory or neurological disorder other than PD. Clinical severity of PD was assessed by the Movement Disorders Society Unified Parkinson's Disease Rating Scale (MDS-UDPRS) 22 . Use of dopamine replacement medication was recorded and levodopa equivalent dose calculated 23 . Demographic data are presented in Table 1.

Isolation and treatment of PBMCs from participants. Blood from participants was collected into
Vacutainer CPT cell preparation tubes with sodium heparin (BD Biosciences) and then centrifuged at 1600 × g for 20 min at room temperature in a swing bucket rotor (Multifuge 3SR, Thermofisher). The layer of peripheral blood mononuclear cells was removed by sterile Pasteur pipette and diluted with 10 volumes of RPMI-1640 medium before being pelleted by centrifugation at 300 × g for 7 min to reduce platelet contamination. The supernatant was discarded and PBMCs resuspended in 10 ml RPMI-1640 media for cell counting and viability measurements using trypan blue exclusion and an automated cell counter (Countess, Life Technology). 1 × 10 6 cells were then treated with LRRK2 inhibitors GSK2578215A and LRRK2-IN1 at indicated concentrations for 1 hr. Cells were then centrifuged at 300 × g for 5 min and the pellets lysed in buffer consisting of 50 mM Tris.HCL pH 7.5, 1 mM EGTA, 1 mM EDTA, 1 mM sodium orthovanadate, 50 mM sodium fluoride, 5 mM sodium pyrophosphate, 0.27 M sucrose, 1 mM benzamidine, 1 mM PMSF and 1% (v/v) Triton X-100. The lysates were snap frozen and stored at − 80 °C until analysis.
Immunoblotting. Samples were thawed on ice and clarified by centrifuging at 10,000 × g for 20 min at 4 °C.
Protein concentration was determined by bicinchoninic assay (Thermofisher Scientific) and samples made up in LDS sample buffer (Life Technologies). 10 μ g of sample was separated with 4-12% Novex Tris-glycine gels (Life Technologies) and transferred onto nitrocellulose membrane (Biorad). Membranes were blocked with 5% skim milk powder in Tris buffered saline with 0.1% (v/v) Tween 20 (TBST). Membranes were probed overnight at 4 °C for LRRK2 (N241A/34, Neuromab) and phosphorylated LRRK2 P-Ser935 (UDD2, Abcam), P-Ser955 (MJF-R11, Abcam), P-Ser973 (MJF-R12, Abcam), P-Ser910 (UDD1, Abcam). All primary antibodies were used at 1:1000 dilution in 5% skim milk in TBST. All LRRK2 antibodies have been substantially validated for specificity using immunoblot 16,17,24 . After overnight incubation membranes were washed in TBST and then anti-rabbit and anti-mouse horseradish peroxidase (HRP) secondary antibodies (Biorad) were used at 1:5000 dilution in 2.5% skim milk in TBST. Enhanced chemiluminescence reagent (GE Healthcare) was used for detection using a Chemidoc MP Imaging system (Biorad). Immunoblot quantitation was performed using Imagelab software 5.1 (Biorad). Levels of phosphorylated LRRK2 were normalized to levels of total LRRK2 and expressed as a percentage of the untreated control PBMCs. For western blot analysis of total LRRK2 levels in healthy donor PBMCs,   ANOVA with Tukey's posthoc test was used to analyze the effect of LRRK2 inhibitors on LRRK2 phosphorylation in the control and PD patient PBMCs. All analyses were performed using Prism software (Graphpad) with significance accepted at p < 0.05.

LRRK2 inhibitors are not toxic to primary human immune cells.
We first assessed any acute toxicity of LRRK2 kinase inhibitors in primary human PBMCs obtained from healthy blood donors. Cells were incubated for 24 h with relatively high concentrations of inhibitors (0.5 μ M). None of the three structurally distinct LRRK2 inhibitors investigated had any effect on cell viability as assessed by trypan blue exclusion (Fig. 1A) or cellular ATP levels (Fig. 1B). None of the inhibitors induced the extracellular release of lactate dehydrogenase (Fig. 1C) or the plasma membrane expression of annexin V (Fig. 1D), indicating no induction of necrosis or apoptosis. These results suggest that LRRK2 kinase inhibitors have no acute toxicity in primary human immune cells.
LRRK2 inhibitors reduce LRRK2 expression in cultured PBMCs. Studies using cell culture and animal models have suggested that LRRK2 kinase inhibitors may promote the proteosomal degradation of LRRK2 protein 25 . Thus we investigated the effect of inhibitor treatment on total levels of LRRK2 in primary human PBMCs obtained from healthy blood donors over a 24 h time course. All inhibitors induced a significant loss of LRRK2 phosphorylation at Ser935 within the first 2 h time point, and phosphorylation remained diminished over the 24 h time course (Fig. 2). Levels of total LRRK2 were stable over most of the time course, however for all three inhibitors there was a 30-40% reduction in total LRRK2 protein by the 24 h time point. This reduction was not seen with the DMSO treated cells (Fig. 2A). In the case of GSK2578215A and LRRK2-IN1 the decrease at 24 h was statistically significant. These results suggest that continued exposure to LRRK2 inhibitors can reduce LRRK2 protein levels in peripheral human immune cells.  Figure 2) inhibitors reduced LRRK2 phosphorylation to similar extents in both the control and PD patient PBMCs for all four phosphorylation sites. With the GSK2578215A inhibitor, the least variance was obtained with the P-Ser935 site (Fig. 3B), whilst under the same conditions the P-Ser955 site appeared to be a more sensitive readout (Fig. 3C) and P-Ser973 the least sensitive readout, in terms of percentage reduction in LRRK2 phosphorylation with inhibitor treatment (Fig. 3D). For the LRRK2-IN1 inhibitor, maximal dephosphorylation of all four sites was achieved with the 0.5 μ M concentration. To determine if PD patient PBMCs continued to respond similarly to control PBMCs with lower doses of LRRK2 inhibitors, we repeated the experiments in a second cohort of subjects, this time using GSK2578215A at 0.125 and 0.063 μ M and LRRK2-IN1 at 0.25 and 0.125 μ M. Again, both GSK2578215A (Fig. 4) and LRRK2-IN1 (Supplementary Figure 3) reduced LRRK2 phosphorylation to a similar extent in both the PD and control subject PBMCs at all four phosphorylation sites. These results show that at least at the concentrations employed, LRRK2 kinase inhibitors will work equally well to reduce LRRK2 phosphorylation in primary immune cells from control and PD patients. Thus, monitoring the phosphorylation status of these residues may have utility as a pharmacodynamic biomarker in clinical trials.

Discussion
Recent years have seen a number of potent and selective inhibitors of LRRK2 kinase activity developed, including by major pharmaceutical and biotechnology companies. The development of these potential drugs has been driven by discoveries that PD-causing missense mutations in LRRK2, especially the relatively common G2019S mutation, increase the enzyme's kinase activity 10 . Some early studies using pre-clinical models of LRRK2-associated PD also support the concept that LRRK2 kinase inhibition may be of therapeutic benefit [26][27][28][29] . PD caused by the activating LRRK2 G2019S mutation is largely indistinguishable from idiopathic PD 6 . Moreover, genome-wide association studies also link LRRK2 variants to idiopathic PD 1-3 . Thus, although it remains to be established whether LRRK2 kinase activity is increased in idiopathic PD, there is a possibility that LRRK2 kinase inhibitors may also be of benefit for this more common form of PD.
In the event that clinical trials of LRRK2 kinase inhibitors are ultimately warranted, it is important to have robust readouts of inhibitor target engagement. The most commonly used readout of LRRK2 target engagement involves monitoring the loss of phosphorylation on a cluster of serine residues located near the N-terminus of LRRK2, particularly Ser935, which occurs with inhibitor treatment 13 . However, LRRK2 kinase activity is not the sole regulator of phosphorylation at these residues, rather it is a yet to be fully elucidated result of a complex interplay between LRRK2 and other upstream kinases and phosphatases 13,[17][18][19]30 . It is therefore important to establish whether LRRK2 inhibitor-induced reductions in serine phosphorylation still occur in cells from PD patients, and if so is it similar in extent to controls.
In this study we have therefore examined PBMCs from control and PD subjects. It is established that LRRK2 is highly expressed in PBMCs, particularly monocytes and B-lymphocytes 31 , providing a convenient source of peripheral primary cells. It has also been demonstrated that levels of LRRK2 and its phosphorylation at Ser910 and Ser935 are unaltered in idiopathic PD PBMCs 32 . Moreover, the levels were not influenced by disease duration, disease severity, levodopa medication or age, suggesting that LRRK2 in PBMCs is stable and suitable for use as a pharmacodynamic biomarker 32 . Our new data using ex vivo treatment of PBMCs from control and PD patients now demonstrates that inhibitor-induced dephosphorylation of Ser910, Ser935, Ser955 and Ser973 is also similar between control and PD patients. This suggests that the complex pathways regulating LRRK2 serine phosphorylation are unperturbed in idiopathic PD PBMCs, and adds further validity to the use of these residues as peripheral pharmacodynamic readouts.
This study employed concentrations of inhibitors previously shown to induce a dose-dependent reduction in LRRK2 phosphorylation in model cell systems 15,21 . In the primary human PBMCs however, these same concentrations caused near maximal loss of phosphorylation, suggesting these cells may be more sensitive to LRRK2 inhibitors. Even when lower concentrations of inhibitors were used we still did not find a significant difference between control and PD patient cells however, it may be prudent to perform IC50 curves with actual clinical trial drugs once these have been established and a better idea of the therapeutic window of LRRK2 drugs is known. Moreover, validation using higher throughput readouts of LRRK2 phosphorylation, such a LRRK2 P-Ser935 ELISA assay 33 , would be important for larger scale studies. Our results also suggest a mild effect of LRRK2 inhibitors at high concentrations for 24 h to reduce total LRRK2 levels, as has been noted in cell culture studies 25 . It may thus also be important to monitor total LRRK2, particularly if treatments are extended over days. It is also noteworthy that the highest inhibitor concentrations used were well tolerated by the cells with no acute toxicity observed.
In summary we provide evidence that reduced phosphorylation of LRRK2 on residues Ser910, Ser935, Ser955 and/or Ser973 can be used as a measure of peripheral target engagement for LRRK2 inhibitors in primary human cells from PD patients.