Pyk2 Regulates MAMs and Mitochondrial Dynamics in Hippocampal Neurons

Pyk2 is a non-receptor tyrosine kinase enriched in hippocampal neurons, which can be activated by calcium-dependent mechanisms. In neurons, Pyk2 is mostly localised in the cytosol and dendritic shafts but can translocate to spines and/or to the nucleus. Here, we explore the function of a new localisation of Pyk2 in mitochondria-associated membranes (MAMs), a subdomain of ER-mitochondria surface that acts as a signalling hub in calcium regulation. To test the role of Pyk2 in MAMs’ calcium transport, we used full Pyk2 knockout mice (Pyk2−/−) for in vivo and in vitro studies. Here we report that Pyk2−/− hippocampal neurons present increased ER-mitochondrial contacts along with defective calcium homeostasis. We also show how the absence of Pyk2 modulates mitochondrial dynamics and morphology. Taken all together, our results point out that Pyk2 could be highly relevant in the modulation of ER-mitochondria calcium efflux, affecting in turn mitochondrial function.


Introduction
Proline-rich tyrosine kinase 2 (Pyk2) is a Ca 2+ -activated non-receptor tyrosine kinase belonging to the focal adhesion kinase (FAK) family [1]. Pyk2 is enriched in hippocampal adult neurons and plays an important role in neuronal plasticity and hippocampal-related memory [2]. In neurons, Pyk2 is mostly localised in the cytosol and dendritic shafts [3,4]. However, we and others have previously demonstrated that following neuronal activation, Pyk2 translocates to other subcellular compartments, such as the nucleus or the excitatory synapses [5,6]. Indeed, several studies have suggested Pyk2 as a key regulator in synaptic plasticity involved with both LTP and LTD [2,7,8]. In this line, we and others have shown that this role of Pyk2 in synaptic plasticity is compromised in different neurodegenerative disorders, such as Alzheimer's disease and Huntington's disease [2,[7][8][9].
Besides this nuclear and synaptic localisation, following Ca 2+ entry into the cell via the transient receptor potential (TRP) channel, Pyk2 can be phosphorylated and partially translocated to mitochondria [10,11]. So far, Pyk2 translocation to mitochondria has been described in cardiac cells, where Pyk2 can be activated and phosphorylate the mitochondrial calcium uniporter (MCU) [11][12][13] enhancing mitochondrial Ca 2+ uptake and mitochondrial ROS production [12]. However, little is known about the molecular mechanism on how Pyk2 can regulate mitochondrial morphology and its consequent role in neuronal physiology.
Indeed, mitochondria are fundamental regulators in synapses since they can travel from soma to synaptic terminals to supply local ATP production and Ca 2+ buffering [14]. To ensure this mitochondrial distribution and number, mitochondria morphology by means of fusion and fission events is critical and both processes must be coordinated [15]. Hence, alterations in mitochondrial morphology can hamper their transport within the cell and could result in synaptic dysfunction and Ca 2+ mishandling [16]. Indeed, to regulate Ca 2+ homeostasis, mitochondria interact with other subcellular compartments such as endoplasmic reticulum (ER), forming the mitochondria-associated membranes (MAMs), a hotspot for Ca 2+ signalling. In this specialised region, calcium is transported from the ER through the IP3R and enters the mitochondria, crossing the channel VDAC1 (voltage dependent "anion" channel 1, which can be permeable to Ca 2+ ). A proper establishment of ER-mitochondria contacts is crucial not only for Ca 2+ homeostasis but also for the control of mitochondrial dynamics and morphology [17]. Indeed, defects in MAMs compromise neuronal functionality and survival [18,19] and several studies have shown either upregulation or disruption of ER-mitochondrial crosstalk in different neurological disorders such as Alzheimer's disease (AD) [19], Parkinson's disease (PD) [20,21], Huntington's disease (HD) [22] and anxiety and depression [23].
In the current work, we aimed to study whether Pyk2 also localises in MAMs and its relevance in neuronal function. We observed that neuronal Pyk2 interacts with several molecules relevant for mitochondrial and endoplasmic reticulum physiology, while the absence of Pyk2 in hippocampal neurons increases the ER-mitochondria interaction leading to a disrupted Ca 2+ buffering. Moreover, we show that Pyk2 modulates fission-fusion events. With all this evidence, we postulate Pyk2 as an essential player in the function of MAMs.

Animals
For this study, full knock-out mice for Pyk2 (Pyk2 −/− ) were used [24]. Animals were genotyped from a tail biopsy. The housing room was kept at 19-22 • C, under a 12:12 h light/dark cycle and animals had access to food and water ad libitum. All mice used in the present study were males of 5 months of age. All animal procedures were approved by local committees (Universitat de Barcelona, CEEA (10141) and Generalitat de Catalunya (DAAM 315/18)), in accordance with the European Communities Council Directive (86/609/EU).

Mass Spectrometry
Immunoprecipitated protein extracts were diluted in Laemmli sample buffer and loaded into a 0.75 mm thick polyacrylamide gel with a 4% stacking gel casted over a  12.5% resolving gel. The run was stopped as soon as the front entered 3 mm into the resolving gel so that the whole proteome became concentrated in the stacking/resolving gel interface. Bands were stained with Coomassie Brilliant Blue, excised from the gel and protein enzymatic cleavage was carried out with trypsin (Promega, Madison, WI, USA; 1:20, w/w) at 37 • C for 16 h, as previously described [25]. Purification and concentration of peptides was performed using C18 Zip Tip Solid Phase Extraction (Millipore). Peptide mixtures were separated by reverse phase chromatography using an UltiMate 3000 UHLPC System (Thermo Scientific, Waltham, MA, USA) fitted with an Aurora packed emitter column (Ionopticks, 25 cm × 75 µm ID, 1.6 µm C18). Samples were first loaded for desalting and concentration into an Acclaim PepMap column (ThermoFisher, Waltham, MA, USA. 0.5 cm × 300 µm ID, 5 µm C18) packed with the same chemistry as the separating column. Mobile phases were 100% water 0.1% formic acid (FA) (buffer A) and 100% Acetonitrile 0.1% FA (buffer B). The column gradient was developed in a 120 min two step gradient from 5% B to 20% B in 90 min and 20% B to 32% B in 30 min. The column was equilibrated in 95% B for 10 min and 5% B for 20 min. During all processes, the precolumn was in line with the column and the flow maintained all along the gradient at 300 nL/min. The column temperature was maintained at 40 • C using an integrated column oven (PRSO-V2, Sonation, Biberach, Germany) and interfaced online with the Orbitrap Exploris 480 MS. Spray voltage were set to 2 kV, funnel RF level at 40, and heated capillary temperature at 300 • C. For the DDA experiments full MS resolutions were set to 1,200,000 at m/z 200 and the full MS AGC target was set to Standard with an IT mode Auto. The mass range was set to 375-1500. The AGC target value for fragment spectra was set to Standard with a resolution of 15,000 and 3 s for cycle time. Intensity threshold was kept at 8E3. Isolation width was set at 1.4 m/z. Normalised collision energy was set at 30%. All data were acquired in centroid mode using positive polarity, and peptide match was set to off and isotope exclusion was on.
Raw files were processed with MaxQuant [26] v1.6.17.0 using the integrated Andromeda Search engine [27]. All data were searched against a target/decoy version of the mouse Uniprot Reference Proteome without isoforms (55,366 entries) from the March 2021 release. First search peptide tolerance was set to 20 ppm, main search peptide tolerance was set to 4.5 ppm. Fragment mass tolerance was set to 20 ppm. Trypsin was specified as enzyme, cleaving after all lysine and arginine residues and allowing up to two missed cleavages. Carbamidomethylation of cysteine was specified as fixed modification and peptide N-terminal acetylation, oxidation of methionine, deamidation of asparagine and glutamine and pyro-glutamate formation from glutamine and glutamate were considered variable modifications with a total of two variable modifications per peptide. "Maximum peptide mass" was set to 7500 Da, the "modified peptide minimum score" and "unmodified peptide minimum score" were set to 25 and everything else was set to the default values, including the false discovery rate limit of 1% on both the peptide and protein levels. The Perseus software (version 1.6.14.0) was used for statistical analysis and data visualization.

Electron Microscopy
Mice were transcardially perfused with a solution containing 4% paraformaldehyde and 0.1% glutaraldehyde (both weight/vol) made up in 0.1 M phosphate buffer (PB), pH 7.4. After perfusion, the brains were removed from the skull, and immersed in the same fixative 12 h 4 • C. Tissue blocks containing the hippocampus were dissected, washed with PB and cut with a vibratome to obtain samples of 1 mm 3 . Hippocampal neuronal cultures after 21 days in vitro were treated with 40 µM glutamate or vehicle and then fixed 30 min with the same fixative solution used for the tissues and washed with PB. Samples from tissues or cell cultures were post-fixed with 2% osmium tetroxide in PB 0.1 M for 20 min. They were dehydrated in a series of ethanol and flat embedded in epoxy resin (EPON 812 Polysciences, Warrington, PA, USA). After polymerisation, blocks from the CA1 region or blocks from cell cultures were cut at 70 nm thickness using an ultramicrotome (Ultracut E Leica, Wetzlar, Germany). Sections were cut with a diamond knife, picked up on formvar-coated 200 mesh nickel grids. They were then immunostained by indirect immunolabeling using protein gold as immunomarker. Protein A-gold probes (20 nm) were obtained from CMC Utrecht (Utrecht, The Netherlands). After immunolabeling, the sections were double stained with uranyl acetate and lead citrate, prior to observation with a Philips (CM-100) electron microscope. Digital images were obtained with a CCD camera (Gatan Orius. Las Positas Blvd, Pleasanton, CA, USA). EPON embedded sections were ultrathin-sectioned and reacted with anti-Pyk2 (Sigma, Burlington, MA, USA. #P3902) antibody revealed with protein A coupled to 10 nm gold particles. For particle counting in animal tissue and in cell cultures, the number of individual gold particles (10 nm) localised in morphometrically determined areas was hand counted and the labelling density was calculated as the number of gold particles per µm and then we relativised the values to a percentage. Each mitochondrial contact closely associated with ER was counted. We considered contact of the mitochondria to ER to be if their distance was less than 30 nm, as previously described [28]. We counted the number of these contacts per cell and fifty mitochondria from six cells per condition were analysed.

Primary Cultures of Hippocampal Neurons
Hippocampal neurons were obtained from E17.5 Pyk2 +/+ and Pyk2 −/− mice. The hippocampus was dissected and mechanically dissociated with a fir-polished Pasteur pipette. Cells were seeded at a density of 50,000 or 100,000 neurons for immunocytochemistry and transfections, respectively, onto 12 mm coverslips placed in 24-well plates or at a density of 80-90,000 neurons onto 25 mm coverslips placed in 6-well plates for calcium analysis. Plates were previously precoated with 0.1 mg/mL poly-D-lysine (Sigma Chemical Co., St. Louis, MO, USA) and neurons were cultured in Neurobasal medium (Gibco-BRL, Renfrewshire, UK) supplemented with 1% Glutamax and 2% B27 (Gibco-BRL). Cultures were maintained at 37 • C in a humidified atmosphere containing 5% CO 2 . All experiments with neuronal cultures were analysed at DIV21.

Proximity Ligation Assay
ER-mitochondria contact sites were analysed performing a proximity ligation assay (Duolink, DUO92101-1KT) following manufacturer's instructions. After fixation and blocking, neurons were incubated with MAP2 to stain cell area. Then, antibodies IP3R3 (1:500, Merck AB9076) and VDAC1 (1:500, Abcam ab15895) were incubated overnight to label ER and mitochondria, respectively. PLA probes anti-mouse PLUS and anti-rabbit MINUS were incubated for 1 h at 37 • C and hybridisation was amplified by rolling circle amplification. Nuclei were stained with DAPI using mounting media provided in the kit. Confocal images were taken using 3.0 digital zoom and stacks of 0.4 µm. Fluorescent dots were counted using ImageJ and number of particles were relativised to cell area.

Calcium Imaging in Neuronal Cultures
To study calcium handling capacity from ER and mitochondria, calcium imaging was performed in hippocampal neuronal cultures. Pyk2 +/+ and Pyk2 −/− neurons at DIV21 were washed with Krebs buffer and loaded with 5 µM Fluo-4 (Invitrogen F-14201) for 30 min and then with 20 nM TMRM (Invitrogen T668) for 20 min at RT to detect intracellular Ca 2+ changes (Cai 2+ ) and mitochondrial membrane potential (∆Ψm), respectively. Coverslips were assembled in a chamber filled with 500µl of Krebs buffer on the stage of confocal microscope with an incubator system that controls temperature and CO 2 . After 50 s, 0.5 µM thapsigargin (TG, Sigma-Aldrich T9033) was injected to inhibit SERCA ATPase and block Ca 2+ storage at the ER. Then, 2 µM FCCP (Sigma-Aldrich C2920) were added at min 7.5 to depolarise mitochondrial membrane and inducing Ca 2+ release to the cytosol. Images were captured every 2.5 s throughout 750 s of experiment. Fluorescence intensity was quantified using ImageJ software and results are expressed as absolute values of intensity.

Mitochondrial Morphology
For electron microscopy, images were obtained from mouse brain sections. The number of mitochondria, aspect ratio (AR) and form factor (FF) were quantified by counting the total number of visible mitochondria present in a fixed size surface of the stratum radiatum from the CA1 region. AR and FF were determined manually tracing individual mitochondria using ImageJ software (NIH, Bethesda, MD, USA). For cell immunocytochemistry, mitochondrial morphology was quantified as previously described [22]. Images were processed with ImageJ software using a convolve filter and automatically given a threshold to obtain binary images. Using the plug-in "Analyse particles", individual mitochondria were measured to calculate the number of mitochondria per micron, AR (length-to-width ratio), and FF (Pm2/4 Am), where Pm is the perimeter and Am is the area of mitochondrion. AR values of 1 mean a perfect circle, whereas higher values correspond to more elongated mitochondria. FF values of 1 represent unbranched mitochondrion and higher FF values indicate a more complex mitochondria network.

Statistical Analysis
Statistical analyses were carried out using the GraphPad Prism 8.0 software (GraphPad, San Diego, CA. USA). All experiments were blind coded to the experimenter. Two-tailed Student's t-test (95% confidence), one-way ANOVA or two-way ANOVA, with Tukey's or Bonferroni's post hoc multiple comparison test was performed as required if distributions were normal. In case of not normal distributions Mann-Whitney t-test or the Dunn's test were used. A p value <0.05 was considered significant.

Pyk2 Is Localised in Mitochondria and MAMs and Interacts with Specific Partners
The physiological and molecular role of Pyk2 in the brain is not yet completely understood. We considered that the elucidation of the Pyk2 interactome could help to decipher its direct association with specific cellular processes in the brain. For that reason, Pyk2 was immunoprecipitated from the hippocampus of Pyk2 +/+ mice. Pyk2 −/− hippocampal material as well as IgG irrelevant antibodies were included as negative controls to detect non-specific-associated proteins. First, hippocampal Pyk2 immunoprecipitation was confirmed using Western blot ( Figure 1A). Subsequent mass-spectrometry analysis was performed to identify Pyk2-associated proteins ( Figure 1B and Supplementary Table S1). Data were curated by excluding non-specific-associated proteins detected in both Pyk2 −/− and IgG immunoprecipitations. After restrictive curation, 52 proteins were considered as co-immunoprecipitating with Pyk2. This Pyk2 interactome was functionally analysed using Metascape [30]. As shown in Figure 1C, multiple functional categories were significantly enriched (Supplementary Table S2). "NMDA receptor activation events" (LogP: −13.95) were represented by 34 Pyk2 associated-proteins (CALM3, CAMK2D, PRKAR2A, TUBA4A,  TUBA1B, TUBB3, TUBB4A, TUBB4B, TBAL3, TCP1 were also significantly over-represented in our Pyk2-interactome data. In summary, the results confirmed the main function of Pyk2 as a downstream regulator of NMDA receptor activity and synaptic plasticity, but they also highlighted a novel role of Pyk2 in the regulation of calcium homeostasis and mitochondrial function in neurons.  Figure 1B; see Supplementary Table S2 for details).

The Lack of Pyk2 Increases ER-Mitochondria Contact Sites
We next explored Pyk2 localisation in the mitochondria of the CA1 pyramidal neurons from Pyk2 +/+ mice by electron microscopy. Pyk2-positive immunogold particles were found within mitochondria and its interactions with ER ( Figure 2A). To test method specificity of the immunogold staining procedure, the primary antibody was omitted.

The Lack of Pyk2 Increases ER-Mitochondria Contact Sites
We next explored Pyk2 localisation in the mitochondria of the CA1 pyramidal neurons from Pyk2 +/+ mice by electron microscopy. Pyk2-positive immunogold particles were found within mitochondria and its interactions with ER ( Figure 2A). To test method specificity of the immunogold staining procedure, the primary antibody was omitted. Under these conditions, no selective labelling was observed ( Figure 2B). Moreover, we validated these results performing a subcellular fractionation and isolating mitochondria. Pyk2 was detected both in total lysate and the different cellular compartments, including nucleus, cytosol and mitochondria ( Figure 2C). Under these conditions, no selective labelling was observed ( Figure 2B). Moreover, we validated these results performing a subcellular fractionation and isolating mitochondria. Pyk2 was detected both in total lysate and the different cellular compartments, including nucleus, cytosol and mitochondria ( Figure 2C). Since we saw that Pyk2 was present at MAMs, we wondered whether this protein could affect the establishment of contact sites between the two implicated organelles. First, we evaluated levels of MAM-resident proteins in hippocampal total lysates. Pyk2 −/− Since we saw that Pyk2 was present at MAMs, we wondered whether this protein could affect the establishment of contact sites between the two implicated organelles. First, we evaluated levels of MAM-resident proteins in hippocampal total lysates. Pyk2 −/− showed lower IP3R3 and higher VDAC1 protein levels compared to Pyk2 +/+ , suggesting structural alterations of MAMs ( Figure 3A). To further examine these organelles associations, we next measured ER-mitochondria interactions both in vivo and in vitro neurons of Pyk2 +/+ and Pyk2 −/− mice. Hippocampal sections were processed for electron microscopy and analysed. Quantification of the number of MAMs revealed an increased number of contact sites in Pyk2 −/− neurons compared to Pyk2 +/+ ( Figure 3B). We also measured ER-mitochondria association in hippocampal neuronal cultures using a proximity ligation assay (PLA) labelling IP3R3 and VDAC1 as probes for ER and mitochondria, respectively. In line with the in vivo results, Pyk2 −/− neurons presented an increase in the number of interactions between organelles compared to Pyk2 +/+ neurons ( Figure 3C).
showed lower IP3R3 and higher VDAC1 protein levels compared to Pyk2 +/+ , suggesting structural alterations of MAMs ( Figure 3A). To further examine these organelles associations, we next measured ER-mitochondria interactions both in vivo and in vitro neurons of Pyk2 +/+ and Pyk2 −/− mice. Hippocampal sections were processed for electron microscopy and analysed. Quantification of the number of MAMs revealed an increased number of contact sites in Pyk2 −/− neurons compared to Pyk2 +/+ ( Figure 3B). We also measured ER-mitochondria association in hippocampal neuronal cultures using a proximity ligation assay (PLA) labelling IP3R3 and VDAC1 as probes for ER and mitochondria, respectively. In line with the in vivo results, Pyk2 −/− neurons presented an increase in the number of interactions between organelles compared to Pyk2 +/+ neurons ( Figure 3C).

Pyk2 Is Involved in Calcium Homeostasis in Hippocampal Neuronal Cultures
MAMs are specialised regions highly implicated in Ca 2+ homeostasis. Given the observed changes in the number of contact sites in Pyk2 −/− neurons, we aimed to study if the absence of Pyk2 could also be affecting Ca 2+ intracellular levels. Therefore, levels of intracellular Ca 2+ (Cai 2+ ) and mitochondrial membrane potential (∆ψm) were labelled using the dyes Fluo4 and TMRM, respectively, in hippocampal Pyk2 +/+ and Pyk2 −/− neurons ( Figure 4A and Supplementary Videos S1 and S2). In basal conditions, no differences in fluorescence intensity were detected between genotypes. Then, thapsigargin (TG) was injected to inhibit SERCA (sarco-endoplasmic reticulum Ca 2+ -ATPases) activity and to deplete the Ca 2+ store in the ER. As expected, after the treatment, Pyk2 +/+ cells significantly increased Cai 2+ levels and lowered ∆ψm. On the contrary, in Pyk2 −/− neurons thapsigargin did not induce a significant increase in Cai2+ levels, either compared to Pyk2 +/+ or to Pyk2 −/− in basal conditions. Consistent with these results, ∆ψm remained unchanged. Next, cells were incubated with FCCP, an oxidative phosphorylation uncoupler, to induce maximal mitochondria depolarisation. Right after this stress, Cai 2+ levels raised in Pyk2 +/+ neurons and, to a lesser extent, in Pyk2 −/− cells. Accordingly, mitochondria depolarisation was observed in Pyk2 +/+ neurons, manifested as a significant decrease in TMRM fluorescence ( Figure 4B,C). Since the mitochondrial calcium uptake strongly depends on the cytoplasmic calcium concentration [31,32], less reduction in the ∆ψm of Pyk2 −/− neurons was detected. This effect could be attributable to the lower response to thapsigargin observed in Pyk2 −/− cells. Altogether this experiment demonstrates that ER depletion in the absence of Pyk2 leads to smaller cytoplasmic calcium elevation, which could be a sign of altered calcium capacity of ER in this condition.

Pyk2 Modulates Mitochondrial Morphology In Vivo and In Vitro
One of the many functions of MAMs is the regulation of mitochondrial dynamics [17]. Considering that the absence of Pyk2 triggers defects in ER-mitochondrial interaction, we wondered whether this could influence mitochondrial dynamics. Thus, we evaluated mitochondrial morphology in CA1 pyramidal neurons in hippocampal sections from both Pyk2 +/+ and Pyk2 −/− mice ( Figure 5A). Pyk2 −/− mice showed a higher mitochondria density ( Figure 5B), along with a reduction in mitochondrial length ( Figure 5C) and mitochondrial branching ( Figure 5D). Furthermore, Pyk2 −/− mice showed an increase in hippocampal TOM20 levels ( Figure 5E,F), corroborating the increase in the number of mitochondria. Additionally, Drp1 protein levels were increased in Pyk2 −/− mice ( Figure 5E,G) whereas both Opa1 and Mitofusin-2 protein levels presented no differences between genotypes ( Figure 5H-J), suggesting that the excessive mitochondrial fragmentation is due to alterations in fission but not fusion processes.
We then assessed mitochondrial morphology in Pyk2 −/− hippocampal primary neurons. Cells were labelled with TOM20 and MAP2 and mitochondrial morphometric analysis was performed ( Figure 6A). Pyk2 −/− neurons had an increased number of mitochondria, together with lower values of AR and FF, indicating mitochondrial fragmentation and validating previous in vivo results ( Figure 6B-D). These results confirmed the data obtained in vivo.

Pyk2 Modulates Mitochondrial Morphology In Vivo and In Vitro
One of the many functions of MAMs is the regulation of mitochondrial dynamics [17]. Considering that the absence of Pyk2 triggers defects in ER-mitochondrial interaction, we wondered whether this could influence mitochondrial dynamics. Thus, we evaluated mitochondrial morphology in CA1 pyramidal neurons in hippocampal sections from both Pyk2 +/+ and Pyk2 −/− mice ( Figure 5A). Pyk2 −/− mice showed a higher mitochondria density ( Figure 5B), along with a reduction in mitochondrial length ( Figure  5C) and mitochondrial branching ( Figure 5D). Furthermore, Pyk2 −/− mice showed an increase in hippocampal TOM20 levels ( Figure 5E,F), corroborating the increase in the number of mitochondria. Additionally, Drp1 protein levels were increased in Pyk2 −/− mice ( Figure 5E,G) whereas both Opa1 and Mitofusin-2 protein levels presented no differences between genotypes ( Figure 5H-J), suggesting that the excessive mitochondrial fragmentation is due to alterations in fission but not fusion processes.  We then assessed mitochondrial morphology in Pyk2 −/− hippocampal primary neurons. Cells were labelled with TOM20 and MAP2 and mitochondrial morphometric analysis was performed ( Figure 6A). Pyk2 −/− neurons had an increased number of mitochondria, together with lower values of AR and FF, indicating mitochondrial fragmentation and validating previous in vivo results ( Figure 6B-D). These results confirmed the data obtained in vivo.

Nuclear Import/Export Domains of Pyk2 Control Dynamics of Hippocampal Mitochondria
To explore the molecular mechanisms underlying mitochondrial regulation by Pyk2, we transfected Pyk2 −/− hippocampal primary neurons with several truncated forms of GFP-Pyk2: DFAT, YF, and RRST ( Figure 7A). Mutated constructs were compared with GFP transfected cells both from Pyk2 +/+ and Pyk2 −/− cultures. As expected, restoring Pyk2 levels in Pyk2 −/− cultures rescued mitochondria numbers in the hippocampal neurons ( Figure 7B,C). Similarly, Pyk2 −/− neurons transduced with DFAT and YF constructs also rescued the number of mitochondria. Interestingly, the RRST construct was unable to rescue this parameter displaying the same phenotype as the one in Pyk2 −/− neurons only transfected with GFP ( Figure 7B,C). These results suggest that the corresponding motifs in the Pyk2 protein (RR184,185, and/or S747-T749) are directly involved in the role of Pyk2 in mitochondria, for example in the binding of specific partners, or that the effects of Pyk2 on mitochondria indirectly requires the nuclear translocation of Pyk2 that is prevented in the RRST mutant [2,5]. In an attempt to assess this possible translocation of Pyk2 into the

Nuclear Import/Export Domains of Pyk2 Control Dynamics of Hippocampal Mitochondria
To explore the molecular mechanisms underlying mitochondrial regulation by Pyk2, we transfected Pyk2 −/− hippocampal primary neurons with several truncated forms of GFP-Pyk2: DFAT, YF, and RRST ( Figure 7A). Mutated constructs were compared with GFP transfected cells both from Pyk2 +/+ and Pyk2 −/− cultures. As expected, restoring Pyk2 levels in Pyk2 −/− cultures rescued mitochondria numbers in the hippocampal neurons ( Figure 7B,C). Similarly, Pyk2 −/− neurons transduced with DFAT and YF constructs also rescued the number of mitochondria. Interestingly, the RRST construct was unable to rescue this parameter displaying the same phenotype as the one in Pyk2 −/− neurons only transfected with GFP ( Figure 7B,C). These results suggest that the corresponding motifs in the Pyk2 protein (RR 184,185 , and/or S 747 -T 749 ) are directly involved in the role of Pyk2 in mitochondria, for example in the binding of specific partners, or that the effects of Pyk2 on mitochondria indirectly requires the nuclear translocation of Pyk2 that is prevented in the RRST mutant [2,5]. In an attempt to assess this possible translocation of Pyk2 into the mitochondria, wild-type hippocampal neurons were treated with 40 µM glutamate for 15 min. Neurons were immunolabeled with Pyk2 and observed by electron microscopy to analyse particle distribution. Interestingly, after glutamate treatment, density of Pyk2 particles in mitochondria increased, supporting the idea of a translocation of Pyk2 into the mitochondria upon neuronal activation ( Figure 7D,E). mitochondria, wild-type hippocampal neurons were treated with 40 μM glutamate for 15 min. Neurons were immunolabeled with Pyk2 and observed by electron microscopy to analyse particle distribution. Interestingly, after glutamate treatment, density of Pyk2 particles in mitochondria increased, supporting the idea of a translocation of Pyk2 into the mitochondria upon neuronal activation ( Figure 7D,E).

Discussion
Pyk2 is considered to be a scaffolding protein with numerous protein interactions (reviewed in [33]). In this work, we have partially defined the Pyk2 hippocampal interactome using immunoprecipitation coupled to mass-spectrometry. Most of Pyk2-associated proteins are directly involved in NMDA function, calcium regulation and mitochondrial homeostasis. Among all these protein partners, Pyk2 has been described to interact with some MAMs proteins as Grp75 and Grp78/BiP in the glioblastoma cell line [34]. In ERmitochondria contacts, Grp75 strengthens the IP3R3-VDAC1 complex whereas Grp78, together with SIGR1, stabilises IP3R3 and boosts Ca 2+ signalling [35,36]. In line with these results, we found that the absence of Pyk2 promotes alterations in the protein levels of IP3R3 and VDAC1 accompanied with an increase in ER-mitochondria apposition in hippocampal neurons both in in vivo and in vitro murine models, suggesting Pyk2 as an important protein in the formation of MAMs in a physiological context. Conversely, the upregulated function of MAM together with increased levels of resident proteins has been described in AD mouse brain and cell line and human fibroblasts [19,37], as well as in PD in in vitro models [38].
Considering that MAMs are highly implicated in calcium homeostasis and given the structural alterations described in Pyk2 −/− mice, we postulate that Pyk2 is an important player in the calcium homeostasis regulated by ER and mitochondria. Our results suggested that Pyk2 depletion could hamper the calcium buffering capacity of ER. In line with this outcome, previous work in endothelial cells has reported that Pyk2 phosphorylates STIM1, a calcium sensor in the ER, thus modulating Ca 2+ store depletion in the ER [39,40]. Since the methodology used in this work for calcium imaging only traced cytoplasmic calcium, we cannot conclude that Pyk2 directly modulates calcium transfer from ER to mitochondria. However, the observed loss of contact sites between organelles and alterations of MAMs resident proteins involved in calcium efflux suggests that calcium homeostasis can be affected in the absence of Pyk2 due to altered functional crosstalk of these microdomains. Hence, further investigation would be necessary to unravel the role of Pyk2 in ER-mitochondria calcium efflux.
Although the molecular mechanisms by which Pyk2 regulates mitochondrial function are scarcely understood, it has been proposed that Pyk2/MCU pathway modulates the mitochondrial calcium uptake [12,41]. Thus, in cardiomyocytes, adrenergic stimulation induces Pyk2 translocation from cytosol to mitochondrial matrix where it directly phosphorylates the mitochondrial calcium uniporter (MCU), facilitating Ca 2+ entry into the mitochondria [12]. Other studies in cardiac and neuroblastoma cells have shown that Ca 2+ entry via TRPM2 activates Pyk2/MCU signalling and modulates mitochondrial function and cell survival [10,11].
MAMs regulate not only calcium buffering but also mitochondrial dynamics [17]. Mitochondrial dynamics are essential to control mitochondrial morphology and distribution within the cell [14] and need to be balanced for a proper cellular function. In the present work, we show that Pyk2 suppression triggers an exacerbated mitochondrial fission but does not alter fusion events. Moreover, Pyk2 −/− hippocampal neurons present augmented numbers of mitochondria and a disrupted mitochondrial network. This mitochondrial fragmentation could compromise neuronal function and viability. Other studies have reported that ablation of fusion or fission proteins promotes an oxidative stress response, neuroinflammation and neuronal death in the hippocampus [42][43][44].
Nevertheless, although we describe here a new function of Pyk2 in mitochondrial dynamics, the underlying mechanism remains unclear. We tried to address how Pyk2 can control the changes in hippocampal mitochondria density and morphology. We showed that upon neuronal activation Pyk2 is capable of translocating to the neuronal mitochondria. Supporting this result, Pyk2 activation can induce its own translocation to the mitochondria in cell lines and trigger production of mitochondrial ROS [12]. Searching for the main Pyk2 domain in the regulation of mitochondria biogenesis we notably observed that, from the tested domains, the RRST was the most important. The latter result indicates that Pyk2 domains of nuclear location and transport are essential for the regulation of mitochondrial dynamics. These findings could indicate two potential independent mechanisms: on the one hand, Pyk2 needs to translocate into the mitochondria and, once there, it regulates undetermined physiological molecular pathways. Indeed, MAMs' chaperone, HSPA8, is responsible for protein imports into the mitochondria [45] and it has been described as interacting with Pyk2 [34]. On the other hand, Pyk2 nuclear import motifs are necessary to translocate Pyk2 to the nucleus as previously shown [3,5] to regulate the transcription of genes [46] related with the function of the mitochondria. Supporting the later hypothesis, it is well known that the mitochondria and the nucleus are coordinated in several physiological conditions [47,48]. In this context, Pyk2 could be a molecular bridge of such a process. However, future research should work in that direction to disentangle the exact mechanism.
Taken together, our results identify Pyk2 as a key molecule in the calcium regulation at ER-mitochondria contact sites. Such calcium regulation is essential for the expression of several forms of synaptic plasticity. Synaptic plasticity is a complex process modulated upon neuronal activity and implicates functional and structural changes in neurotransmitters' release, cytoskeleton reorganisation, gene transcription and protein synthesis [49]. To complete all these cellular and molecular modifications, synapses require a tight calcium modulation and high energetic demands. Thus, mitochondria arise as a fundamental regulator in synapses since they can travel from soma to synaptic terminals to supply local ATP production and Ca 2+ buffering [14]. To ensure mitochondrial distribution and numbers, fusion and fission events must be coordinated [15]. Alterations in mitochondrial morphology can hamper their transport within the cell and could result in synaptic dysfunction and Ca 2+ mishandling [16]. Accordingly, it is widely accepted that calcium mishandling is a hallmark of neurological diseases. In this line, Pyk2 has been strongly implicated in neurodegenerative and psychiatric disorders as HD [2], chronic stress [50] and AD [8,51]. Indeed, the encoding gene PTK2B has been strongly associated as a genetic risk factor for late-onset AD [52]. Hence, we conclude that further investigation on Pyk2's functions in physiological conditions implicating ER-mitochondria contact sites will contribute to a better understanding on the pathophysiology of several brain diseases.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/cells11050842/s1, Table S1: Mass-spectrometry data associated to immunoprecipitation experiments. Blue light cluster of proteins indicate the potential interactors with Pyk2. Table S2: Functional mapping of Pyk2-associated proteins obtained by Metascape. The analysis was performed with various ontology sources, including KEGG Pathway, GO Biological Process, Reactome Gene Sets, Canonical Pathways, CORUM and Wikipathways. All genes in the genome were used as the enrichment background. Terms with a p-value < 0.01, a minimum count of 3, and an enrichment factor (ratio between the observed counts and the counts expected by chance) >1.5 were grouped into clusters based on their membership similarities. To represent each cluster, the most statistically significant term was chosen. Video S1: Pyk2 +/+ culture. Video S2: Pyk2 −/− culture.