Graduate School of Cellular and Molecular Neuroscience, University of Tübingen, Tübingen, Germany
German Center for Neurodegenerative Diseases, University of Tübingen, Tübingen, Germany
Laboratory of Functional Neurogenetics, Department of Neurodegeneration, Hertie Institute for Clinical Brain Research, University of Tübingen, Tübingen, Germany
Address correspondence and reprint requests to Philipp Kahle, Laboratory of Functional Neurogenetics, Department of Neurodegeneration, German Center for Neurodegenerative Diseases and Hertie Institute for Clinical Brain Research, University of Tübingen, Otfried Müller Str. 27, 72076 Tübingen, Germany. E-mail: firstname.lastname@example.org
Read the Editorial Highlight for this article on page 895.
The gene encoding leucine-rich repeat kinase 2 (LRRK2) comprises a major risk factor for Parkinson's disease. Recently, it has emerged that LRRK2 plays important roles in the immune system. LRRK2 is induced by interferon-γ (IFN-γ) in monocytes, but the signaling pathway is not known. Here, we show that IFN-γ-mediated induction of LRRK2 was suppressed by pharmacological inhibition and RNA interference of the extracellular signal-regulated kinase 5 (ERK5). This was confirmed by LRRK2 immunostaining, which also revealed that the morphological responses to IFN-γ were suppressed by ERK5 inhibitor treatment. Both human acute monocytic leukemia THP-1 cells and human peripheral blood monocytes stimulated the ERK5-LRRK2 pathway after differentiation into macrophages. Thus, LRRK2 is induced via a novel, ERK5-dependent IFN-γ signal transduction pathway, pointing to new functions of ERK5 and LRRK2 in human macrophages.
Leucine-rich repeat kinase 2 (LRRK2) is a major risk factor for the development of Parkinson's disease (PD). However, the role of LRRK2 in the affected neurons remains enigmatic. Recently, LRRK2 has been reported to be strongly expressed in the immune system. Here, we demonstrate that LRRK2 is induced by Interferon gamma via extracellular signal-regulated kinase 5 (ERK5) in macrophages, thus providing new insights in LRRK2 and ERK5 biology.
signal transducer and activator of transcription 1
Tris-buffered saline + 0.1% Tween 20
The gene encoding leucine-rich repeat kinase 2 (LRRK2) comprises the most common autosomal-dominant Parkinson's disease (PD) locus PARK8 (Paisán-Ruíz et al. 2004; Zimprich et al. 2004). Moreover, LRRK2 is a major genetic risk factor for PD (Satake et al. 2009; Simón-Sánchez et al. 2009). Initial work on LRRK2 has focused on the nervous system (Cookson 2010), but the pathomechanisms of PARK8 still remain surprisingly enigmatic. This could be partly because of the relatively low expression levels of LRRK2 in the brain regions affected by PD (Galter et al. 2006). More recently, an important role of LRRK2 in the immune system has emerged (Greggio et al. 2012). LRRK2 is abundantly expressed in different immune cells. In addition to PD, LRRK2 was found in genome-wide association studies as a risk factor for immune system-associated diseases, viz., leprosy (Zhang et al. 2009), a chronic infectious disease, and Crohn's disease (Barrett et al. 2008), an inflammatory bowel disease. In this context, LRRK2 was identified as a suppressor of the nuclear factor of activated T cells, sequestering this transcription factor in the cytosol (Liu et al. 2011).
Within the immune system, LRRK2 is expressed in lymphocytes, dendritic cells and macrophages (Gardet et al. 2010). Stimulation of toll-like receptor 4 in mouse bone marrow-derived macrophages led to an induction of LRRK2 (Hakimi et al. 2011), whereas opposite findings were observed in another study (Liu et al. 2011). In human monocytic cells, interferon-γ (IFN-γ) strongly induced LRRK2 expression (Gardet et al. 2010). However, the signal transduction pathway(s) inducing LRRK2 expression in IFN-γ-stimulated macrophages is not known. Here, we used the human acute monocytic leukemia cell line THP-1 differentiated into macrophage-like cells by phorbol ester treatment to study LRRK2 induction pathway(s). Exposure to the kinase inhibitor (IN)-1 (Deng et al. 2011) abolished IFN-γ induction of LRRK2, initially suggesting an autoregulatory loop for the LRRK2 kinase. However, none of the other LRRK2 kinase inhibitors used showed this effect. Thus, IN-1 may interfere with cellular effects by cross-reacting with kinases other than LRRK2. As IN-1 stands out among the LRRK2 inhibitors used as having cross-reactivity with the extracellular signal-regulated kinase, ERK5 (also known as BMK1: big mitogen-activated protein kinase (MAPK) 1; systematic name MAPK7), we considered ERK5 as the mediator of LRRK2 induction. Indeed, pharmacological inhibition as well as RNA interference (RNAi) of ERK5 reduced IFN-γ mediated induction of LRRK2 in differentiated THP-1 cells. Likewise, ERK5 inhibition suppressed LRRK2 induction in IFN-γ-stimulated human monocyte-derived macrophage (MDM) cultures. These findings point to a novel signal transduction pathway in macrophages.
Antibodies, cytokines, and other reagents
Antibodies and other reagents were obtained as follows: ERK5 (ab40809; Abcam, Cambridge, UK), LRRK2 (ab133518; Abcam), LRRK2 phospho-S935 (# 5099-1; Epitomics, Burlingame, CA, USA) LRRK2 phospho-S910 (# 5098-1; Epitomics), signal transducer and activator of transcription (STAT)1 (#9172; Cell Signaling Technology, Beverly, MA, USA), STAT1 phospho-S727 (#9177; Cell Signaling), STAT1 phospho-Y701 (#9167; Cell Signaling), vinculin (# V 9131; Sigma-Aldrich, St. Louis, MO, USA), glyceraldehyde-3-phosphate dehydrogenase (clone 6C5; Meridian Life Science, Memphis, TN, USA), secondary horseradish peroxidase (HRP)-conjugated antibody for western blot analysis (Jackson Immuno-Research, West Grove, PA, USA) and secondary AlexaFluor-488 conjugated antibody for immunofluorescence (Invitrogen, Carlsbad, CA, USA).
Recombinant human IFN-γ and IFN-β were purchased from Peprotech (Hamburg, Germany), IFN-α from PBL Interferon Source (Piscataway, NJ, USA). Tumor necrosis factor-α and granulocyte–macrophage colony-stimulating factor (GM-CSF) were purchased from ImmunoTools (Friesoythe, Germany). Phorbol-13-myristate-12-acetate (PMA), LRRK2 IN-1, and XMD8-92 were all from Tocris Bioscience/R & D Systems (Wiesbaden-Nordenstadt, Germany). Janus kinase (JAK) IN-1, H-1152, K-252a, and CZC-25146 were all from Calbiochem/EMD Millipore (Billerica, MA, USA). BIX02189 and TAE684 were both from Selleckchem (Munich, Germany) and Sunitinib from Sigma-Aldrich.
THP-1 cells were maintained at 5 × 105 cells/mL in RPMI-1640 medium (ATCC, Manassas, VA, USA) supplemented with 10% fetal calf serum at 37°C and 5% CO2. To induce differentiation into a more macrophage-like state, cells where seeded in six-well plates (1.5 × 105 cells/well) treated with 10–100 ng/mL PMA for 24–48 h. For RNAi experiments, lentiviral transduction particles were purchased (Sigma-Aldrich) and stable ERK5 knockdown (KD) cells were generated according to manufacturer's instructions. In brief, THP-1 cells were transduced by spinfection and pooled stable KD cells were selected by puromycin resistance. KD efficiency was tested by immunoblot analysis. Non-targeting small hairpin RNA (shRNA) served as a negative control.
Human peripheral blood mononuclear cells were isolated by Ficoll-Paque density centrifugation from whole blood donated by healthy individuals. This was approved by the ethics commission of the medical faculty of the University of Tübingen. The cells were washed and resuspended in RPMI-1640 containing 10% pooled heat-inactivated normal human serum and 1% pen/strep and seeded into six-well plates at the concentration of 10–15 × 106 cells/well. Non-adherent cells were removed by gentle but intensive washing after 2 h of incubation at 37°C. Adherent cells were allowed to further differentiate for 7 days into MDMs in the presence of 100 U/mL GM-CSF, which was given at days 1, 3, and 5 of differentiation. MDMs were then directly subjected to stimulations/treatments.
Western blot analysis
Cells were lysed for 15 min on ice in radioimmunoprecipitation assay buffer (50 mM Tris/HCl pH 8.0; 150 mM NaCl; 1% NP-40; 0.5% deoxycholate; 0.1% sodium dodecyl sulfate) containing cOmplete protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany), and phosSTOP phosphatase inhibitor cocktail (Roche Diagnostics) when probing phosphoproteins. Cell debris was removed by centrifugation (15 min at 14 000 g and 4°C) and protein concentrations of lysates were determined using a bicinchoninic acid protein assay kit (Pierce, Rockford, IL, USA). Protein samples were boiled in Laemmli buffer at 95°C for 5 min. A total of 15 μg of denatured proteins were then separated on 6% or 12.5% polyacrylamide gels and transferred onto Immobilon-P polyvinylidene difluoride membrane (EMD Millipore). Membranes were pre-incubated 2–16 h at 4°C in 5% skim milk in Tris-buffered saline + 0.1% Tween 20 (TBST), 5% bovine serum albumin/TBST or 5% Western Blocking Reagent/TBST and incubated with primary antibody in Western Blocking Reagent (Roche Diagnostics) at 4°C overnight. This was followed by incubation with HRP-conjugated secondary antibodies for 2 h at 20°C. Detection of proteins was performed with the Immobilon Western Chemiluminescent HRP Substrate (EMD Millipore) on Amersham Hyperfilm™ for enhanced chemiluminescence (GE Healthcare, Little Chalfont, UK).
RNA was isolated using the RNeasy Mini kit (Qiagen, Hilden, Germany), according to manufacturer's instructions. Total RNA (1000 ng) was reverse transcribed with anchored oligo-dT primer using the Transcriptor High Fidelity cDNA Synthesis kit (Roche Diagnostics). cDNA was used as template for transcription amplification in a 25-μL reaction with 5 μL 5x GoTaq buffer, 0.1 μL GoTaq Polymerase (Promega, Madison, WI, USA), and 2 mM primer (LRRK2 For_4 Ex46+: 5′-AAACTCTGTGGACTAATAGACTGCGT-3′, LRRK2 Rev_4 Ex46+: 5′-TTTAAGGCTTCCTAGCTGTGCTGTC-3′, β-actin For_1: 5′-AAGATCAAGATCATTGCTCCTCC-3′, β-actin Rev_1: 5′-GTCATAGTCCGCCTAGAAGCA-3′). Amplified PCR products were subjected to electrophoresis in a 1.5% agarose gel stained with ethidium bromide.
THP-1 cells were seeded on poly-d-lysine (Sigma-Aldrich) coated coverslips and treated as indicated. Then, the cells were fixed with ice-cold methanol for 10 min at −20°C, and blocked 2 h with 10% normal goat serum. Primary antibody incubation was performed in 1% bovine serum albumin/phosphate-buffered saline overnight at 4°C. Cells were incubated with secondary Alexa-Fluor conjugated antibodies in the dark for 1 h at 20°C. Nuclei were counterstained with Hoechst 33342 (2 μg/mL/phosphate-buffered saline) for 10 min at 20°C. Coverslips were mounted in fluorescence mounting medium (Dako, Hamburg, Germany) onto microscope slides. Cells were analyzed with ApoTome Imaging system and processed with AxioVision software (Zeiss, Oberkochen, Germany).
IN-1 reduces IFN-γ-mediated induction of LRRK2 in macrophage-like cells
THP-1 cells were differentiated into macrophage-like cells by 48 h pre-treatment with 10 ng/mL phorbol ester PMA. Subsequent exposure to 200 U/mL IFN-γ strongly induced LRRK2 mRNA (Fig. 1a) and protein (Fig. 1b) expression, as expected. Interestingly, when IN-1 was coadministered, LRRK2 induction was inhibited in a dose-dependent manner (Fig. 1a). LRRK2 protein levels were greatly diminished after 48 h cotreatment with IFN-γ and the maximal dose of IN-1 (Fig. 1b). The inhibitory effect on LRRK2 protein induction appeared somewhat stronger than on the mRNA level. Thus, LRRK2 expression might be regulated at the transcriptional and post-transcriptional level.
At the high dose of 3 μM IN-1, reduction of IFN-γ-stimulated LRRK2 protein was emerging between 5 and 10 h and robust by 24 h, while the suppressive effect of the lower dose 0.3 μM IN-1 was seen at the later time point 48 h (Fig. 1c and d). IN-1 inhibited LRRK2 kinase activity throughout the time course, as determined by the LRRK2 phosphorylation state at S935 (Dzamko et al. 2010).
To check proximal IFN-γ signaling events, we measured the phosphorylation state of STAT1 at Y701. As expected, IFN-γ stimulated STAT1 Y701 phosphorylation, and this was not affected by IN-1 treatment (Fig. 1e). Thus, IN-1 did not interfere with proximal IFN-γ signaling itself, but rather a downstream step in the signal transduction pathway leading to LRRK2 induction. Blocking the immediate effector of IFN-γ with a selective JAK inhibitor blocked STAT1 Y701 phosphorylation and also LRRK2 up-regulation in a dose-dependent manner (Fig. 1e). Combining LRRK2 IN-1 and JAK IN-1 practically abolished LRRK2 induction. IFN-β, which signals through a different receptor system as IFN-γ, did not stimulate STAT1 Y701 phosphorylation and also did not consistently up-regulate LRRK2 under these conditions (Fig. 1e). Taken together, LRRK2 expression seems to be under control of a specific IFN-γ signal transduction pathway in macrophage-like cells, which can be suppressed by IN-1.
Other LRRK2 inhibitors fail to suppress IFN-γ-mediated induction of LRRK2
To generalize the above finding, we tested a panel of additional LRRK2 inhibitors, including CZC-25146 (Ramsden et al. 2011), TAE684 (Zhang et al. 2012), H-1152 and Sunitinib (Nichols et al. 2009). Surprisingly, all of these inhibitors failed to show the same LRRK2 suppressive effect in IFN-γ-stimulated macrophage-like cells, although they did inhibit LRRK2 kinase activity as detected with phospho-S935-specific antibody (Fig. 2). None of the inhibitors interfered with the initial IFN-γ stimulation, as confirmed by STAT1 Y701 phosphorylation, except for K-252a (Covy and Giasson 2009) at the highest concentration (1 μM) that also reduced the level of LRRK2 up-regulation.
Inhibition of IN-1 cross-reactive kinase ERK5 suppresses LRRK2 induction
Considering the inhibitory profiles of the used inhibitors, IN-1 stands out as having cross-reactivity with ERK5 (Luerman et al. 2013). Therefore, we tested a selective ERK5 inhibitor, namely, XMD8-92 (Yang et al. 2010). Similar to IN-1, XMD8-92 inhibited the induction of LRRK2 in IFN-γ activated macrophage-like cells over a 1-day time course (Fig. 3a). Like IN-1, XMD8-92 did not interfere with proximal IFN-γ signaling, as confirmed by unaltered STAT1 phosphorylation at Y701 and S727 throughout the time course (Fig. 3a). We also used another, unrelated inhibitor of the ERK5 module, BIX02189 (Tatake et al. 2008). Like XMD8-92, BIX02189 suppressed the IFN-γ mediated induction of LRRK2 protein, in a concentration-dependent manner (Fig. 3b).
The suppressive effects on LRRK2 up-regulation were confirmed by LRRK2 immunostaining (Fig. 4). IFN-γ treatment caused morphological alterations in the macrophage-differentiated THP-1 cells, which were highly immunoreactive for LRRK2. These morphological changes were suppressed along with the return to basal LRRK2 levels in the presence of both IN-1 and XMD8-92 (Fig. 4).
To validate the pharmacological findings, we performed RNAi to knock down ERK5. Acute transient transfections with several small interfering RNAs against ERK5 yielded inconclusive results. Thus, we transduced THP-1 cells with lentiviral vectors harboring ERK5-directed shRNA and selected stable ERK5 KD clones. These cells were pre-treated with PMA and then stimulated with IFN-γ. Control cells showed the IFN-γ concentration-dependent augmentation of LRRK2 protein, whereas in ERK5 KD cells LRRK2 induction was attenuated (Fig. 5a). Clone #5 with strong ERK5 KD showed a greater reduction of IFN-γ-induced LRRK2 protein than clone #10 with partial ERK5 KD, in which LRRK2 induction was only mildly affected (Fig. 5a). Thus, residual quantities of ERK5 are sufficient to mediate the effects on LRRK2 induction.
Finally, we found that the PMA pre-treatment induced ERK5 expression in THP-1 cells, which was a pre-requisite for subsequent IFN-γ-mediated LRRK2 induction (Fig. 5b). PMA differentiation alone led to a modest elevation of LRRK2 protein levels, which was boosted by IFN-γ but not IFN-β or IFN-α. Taken together, these experiments establish that ERK5 mediates IFN-γ induction of LRRK2, as evidenced by (i) pharmacological inhibition of ERK5 (ii) ERK5 RNAi, and (iii) correlation between ERK5 levels and LRRK2 induction strength.
IFN-γ induces LRRK2 via ERK5 in human macrophages
Next, we sought to extend our findings in primary human MDMs. Peripheral blood adherent monocytes were differentiated in vitro to macrophages with GM-CSF. On day 7, the MDMs were challenged with IFN-γ. Western blot analysis confirmed that IFN-γ increased LRRK2 protein levels (Fig. 6). Additional stimulation with tumor necrosis factor-α did not lead to further increases in LRRK2 levels. Importantly, the ERK5 inhibitor XMD8-92 suppressed the IFN-γ-mediated induction of LRRKs in human MDMs, as did the kinase inhibitor IN-1 (Fig. 6).
Since the discovery of LRRK2 being a major risk factor for the development of PD, most research has focused on its biological function and role in PD pathogenesis in neuronal models. Most studies were dedicated to identify cell-autonomous mechanisms that could explain the observed neurodegeneration in the affected nigrostriatal system. In spite of tremendous efforts, the role of LRRK2 in neurons is still enigmatic. Although very useful to study focused aspects of LRRK2 (patho)physiology, genetically engineered mice with neuronal over-expression of wild-type or mutant LRRK2 or Lrrk2 knocked out partially modeled PD pathology at best (Sloan et al. 2012; Yue 2012). In fact, overall LRRK2 expression in neuronal cells is very low compared to expression in immune cells. Therefore, it is appealing to speculate that LRRK2 plays a major role in the immune system (Greggio et al. 2012) and contributes to PD pathology in a non-cell-autonomous manner by a yet unknown mechanism.
In macrophage-like cells, IFN-γ but not IFN-α or IFN-β strongly up-regulated LRRK2. Therefore, the LRRK2-inducing pathway is specific for signaling via the IFN-γ receptor and not the IFN-α/β receptor. The type-II interferon IFN-γ classically signals through the JAK/STAT pathway. Specifically, binding of an IFN-γ homodimer induces activation of the distinct IFN-γ receptor system and subsequent activation of receptor-associated JAKs, which in turn phosphorylate the STAT1 homodimer leading to the expression of genes with IFN-γ-activated sequence promoter elements (de Weerd and Nguyen 2012).
Interestingly, repression of ERK5 greatly diminished the IFN-γ-mediated induction of LRRK2. Because phosphorylation of STAT1 Y701 was not affected by ERK5 inhibition and JAK1 inhibition itself was also reducing LRRK2 up-regulation, ERK5 likely exerts its function downstream of JAK1, but independent of STAT1 and general IFN-γ response. Whether JAK1 is signaling directly or indirectly through ERK5 remains to be elucidated. The related p38MAPK has been suggested to mediate STAT1 phosphorylation at S727 (Goh et al. 1999), which is required for full transcriptional activity of STAT1. Yet, neither IN-1 nor XMD 8-92 treatment altered phosphorylation at S727 indicating that ERK5 modulates the IFN-γ response at a more downstream level.
Other MAPKs have been associated with IFN signaling in macrophages (Katsoulidis et al. 2005; Valledor et al. 2008), however, ERK5 dependency has not yet been studied this context. Classically, oxidative stress, hyperosmolarity, and several growth factors including epidermal growth factor induce ERK5 activation by phosphorylation of the threonine-x-tyrosine motif within its activation loop by signaling through the upstream MAPK/ERK kinase 5 (Hayashi and Lee 2004). Surprisingly, two recent studies proposed additional non-canonical mechanisms in which catalytically inactive ERK5 is mediating a transcriptional response independent of MAPK/ERK kinase 5 (Díaz-Rodríguez and Pandiella 2010; Erazo et al. 2013). In contrast to the other MAPKs, ERK5 has a unique, extended C-terminal tail containing a nuclear localization signal and a potent transcriptional activation domain (Nithianandarajah-Jones et al. 2012). Therefore, ERK5 can activate transcription indirectly by phosphorylating transcription factors but also itself directly by serving as a transcriptional stimulator for the activator protein-1 (Morimoto et al. 2007). Another major target of ERK5 indirect transcriptional modulation is a group of transcription factors termed myocyte enhancer factor 2, which have been suggested to play a role in PD (Yin et al. 2012). The IFN-γ–ERK5 pathway could also indirectly influence the LRRK2 promotor by inducing transcription factors that in turn activate LRRK2 expression. Beyond transcriptional regulation, it should be noted that the ERK5 inhibition effects on LRRK2 mRNA levels are somewhat less pronounced than at the protein level. Thus, in addition to transcriptional regulation, the IFN-γ-stimulated ERK5 pathway might also influence LRRK2 expression at the post-transcriptional level.
The IFN-γ-induced ERK5-dependent up-regulation of LRRK2 is not restricted to PMA-differentiated monocytic leukemia THP-1 cells but also occurs in human macrophages. Upon IFN-γ treatment, THP-1 cells as well as MDMs are activated and undergo morphological changes from a roundish toward a more elongated (spindle-like) and ‘active’ appearance. Pharmacological inhibition of ERK5 did not only reduce the LRRK2 induction but also had a repressive effect on these morphological changes. This might indicate that LRRK2 itself or other coregulated ERK5-dependent IFN-γ targets are involved in these processes. Lipopolysaccharide stimulation of toll-like receptor 4 in mouse brain microglial cells also led to an induction of LRRK2, and IN-1 treatment as well as LRRK2 RNAi had similar repressive effects on microglial morphology (Moehle et al. 2012) as found here for IFN-γ-treated macrophages. Taken together, this could indicate that LRRK2 could be a regulator of such morphological responses.
In conclusion, we provide evidence for an ERK5-dependent IFN-γ-induced pathway, which up-regulates LRRK2 in human macrophages. This study suggests that ERK5 and LRRK2 may have novel functions regulating macrophage morphology and functions.
Acknowledgments and conflict of interest disclosure
We thank Thomas Gasser and Hans-Georg Rammensee for help and discussions. This work was supported by a DZNE PhD stipend (to M.K.), the German Center for Neurodegenerative Diseases within the Helmholtz Association, and the Hertie Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
The authors have no conflict of interest to declare.