SEARCH

SEARCH BY CITATION

Keywords:

  • ataxia;
  • cerebellum;
  • extracellular-signal regulated kinase;
  • gene targeting;
  • protein tyrosine phosphatase;
  • signal transduction

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The neuronal protein tyrosine phosphatases encoded by mouse gene Ptprr (PTPBR7, PTP-SL, PTPPBSγ-42 and PTPPBSγ-37) have been implicated in mitogen-activated protein (MAP) kinase deactivation on the basis of transfection experiments. To determine their physiological role in vivo, we generated mice that lack all PTPRR isoforms. Ptprr−/− mice were viable and fertile, and not different from wildtype littermates regarding general physiology or explorative behaviour. Highest PTPRR protein levels are in cerebellum Purkinje cells, but no overt effects of PTPRR deficiency on brain morphology, Purkinje cell number or dendritic branching were detected. However, MAP kinase phosphorylation levels were significantly altered in the PTPRR-deficient cerebellum and cerebrum homogenates. Most notably, increased phospho-ERK1/2 immunostaining density was observed in the basal portion and axon hillock of Ptprr−/− Purkinje cells. Concomitantly, Ptprr−/− mice displayed ataxia characterized by defects in fine motor coordination and balance skills. Collectively, these results establish the PTPRR proteins as physiological regulators of MAP kinase signalling cascades in neuronal tissue and demonstrate their involvement in cerebellum motor function.

Abbreviations used
DUSP

dual-specificity phosphatase

ERK

extracellular signal-regulated kinase

KIM

kinase interaction motif

MAPK

mitogen-activated protein kinase

PTP

protein tyrosine phosphatase

The cerebellum is the major centre of fine motor coordination in the central nervous system, and in addition serves in cognitive processing and sensory discrimination. These functions require proper cellular signalling through reversible phosphorylation of proteins. Neurotrophic factors, for example, exploit phosphorylation cascades to exert their effect on the developing nervous system (Kaplan and Miller 2000) and synaptic plasticity and memory in the adult brain relies heavily on the mitogen-activated protein kinase (MAPK) cascade (Sweatt 2004; Thomas and Huganir 2004). Concomitantly, abnormal protein phosphorylation is associated with neurological pathologies involving malformations, excitotoxicity and neuron loss (Colucci-D’Amato et al. 2003; Gee and Mansuy 2005). In the cerebellum, this may cause autism, mental retardation, palsy, epilepsy and, most notably, ataxia (Chizhikov and Millen 2003). Thus, signalling relays such as the MAPK signalling pathway must be tightly controlled to warrant proper functioning.

Spatiotemporal control of MAPK activity is exerted through docking and scaffolding interactions with activating upstream kinases and inactivating phosphatases (Pouyssegur and Lenormand 2003; Tanoue and Nishida 2003). MAPK activation involves the dual phosphorylation of a threonine and a tyrosine residue that are close together in the activation lip. Inactivation of MAPKs thus can be achieved by serine/threonine phosphatases, by tyrosine-specific phosphatases (PTPs), or by so-called dual-specificity phosphatases (DUSPs). Some phosphatases are specifically tailored for this task. They contain a so-called kinase interaction motif (KIM) that mediates selective binding to the MAPK common docking domain (Tanoue and Nishida 2003), thereby enabling retention and inactivation of the MAPK in the cytosol.

Three human genes encode KIM-containing tyrosine-specific phosphatases: PTPN5, PTPN7 and PTPRR. When over-expressed in cells they display pronounced effects on MAPK signalling cascades, most notably the extracellular signal-regulated kinases ERK1/2 relay (Pulido et al. 1998; Ogata et al. 1999; Saxena et al. 1999; Zúñiga et al. 1999; Buschbeck et al. 2002; Munoz et al. 2003; Paul et al. 2003). Physiological relevance has been provided for PTPN5 (STEP) by pharmacological studies in vitro and in vivo (Paul et al. 2003; Valjent et al. 2005), for PTPN7 (HePTP/LC-PTP) exploiting a knock-out mouse model (Gronda et al. 2001) and for a homologous protein in Drosophila, PTP-ER, through mutational analyses (Karim and Rubin 1999). Here, we provide evidence that the protein products of the neuronal mouse gene Ptprr (i.e. PTPBR7, PTP-SL, PTPPBSγ-42 and PTPPBSγ-37; Chirivi et al. 2004) also represent physiological regulators of MAPK activities. We have generated mice that lack all four PTPRR isoforms and found that ERK1/2 phosphorylation levels in Ptprr−/− brain tissue were increased when compared with wildtype material. Furthermore, Ptprr−/− mice displayed fine motor coordination and balance skill defects in multiple tests, establishing a role for PTPRR proteins in cerebellum motor function and the regulation of MAPK signalling cascades in neuronal tissue.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Generation of Ptprr−/− mice

A replacement-type targeting vector was constructed that contained a 1.7 kbp BglII-BamHI fragment from intron 5 and a 3 kbp HindIII-XhoI fragment from intron 6 of the mouse gene Ptprr to facilitate homologous recombination. Both genomic fragments where amplified by PCR from mouse 129/Ola genomic DNA using high-fidelity polymerase and digested with the appropriate restriction enzymes. In between the Ptprr homology regions a neomycin phosphotransferase selection cassette was inserted in the opposite transcriptional orientation, and at the upstream end of the targeting vector a diphtheria toxin expression cassette was included to allow selection against random insertions. The linearized targeting construct was introduced into mouse 129/Ola ES cells and resulting G418-resistant colonies were scored for proper recombination by Southern blot analysis, exploiting 5′ and 3′ diagnostic probes derived from flanking intronic sequences, or by allele-specific PCR analysis. Details, including primer sequences, are provided in Table S1. Successfully targeted ES clones were karyotyped and only purely diploid cell lines were used for injection into blastocysts and embryo transfer. Resulting chimaeric male mice were used to produce heterozygous F1 offspring with C57BL/6 females. F1 animals were backcrossed onto the C57BL/6 genetic background for one more generation and resulting heterozygotes were then intercrossed to produce a mouse colony containing Ptprr−/− animals and wildtype littermate controls. Mice were kept at the Central Animal Facility of the Radboud University Nijmegen in a standard room with a day/night rhythm of 06:00/18:00 h at a temperature of 21°C and a humidity of 50–60%. Mice were housed in Macrolon cages and fed ad libitum. All procedures involving animals were approved by the Animal Care Committee of the Radboud University Nijmegen Medical Centre, The Netherlands, and conformed to the guidelines of Dutch Council for Animal Care and the NIH.

RNA analysis

Total RNA from cerebrum and cerebellum of wildtype, heterozygous, and homozygous Ptprr−/− mice was purified using RNazol B (Campro Scientific, Veenendaal, the Netherlands). As much as 10 μg RNA samples were size-fractionated on a 1% formamide agarose gel, transferred to Hybond-N membranes (Amersham Pharmacia Biotech, Amersham, UK) and probed with a 32P-labelled cDNA fragment (nucleotide positions 980–3430; accession no D31898) comprising exons 5 until 14 of PTPBR7 that is present in all Ptprr RNA isoforms. Overnight hybridization, subsequent washes in 40 mmol/L phosphate buffer containing 0.2% SDS, and exposure to Kodak X-Omat S1 films was as described (Church and Gilbert 1984).

Antibodies

The rabbit polyclonal antiserum α-SL (1 : 5000) and the monoclonal antibodies 6A6 (1 : 16) and 5E4 (1 : 16), all directed against the phosphatase moiety in PTPRR isoforms, have been described previously (van den Maagdenberg et al. 1999; Chirivi et al. 2004). The rabbit antisera raised against STEP and the extracellular domain of PTPBR7 (α-BR7; Dilaver et al. 2007) were a kind gift of Drs R. Pulido and P. Ríos (Valencia, Spain). Polyclonal antisera against total ERK1 (1 : 8000), p38 (1 : 2000), and JNK1 (1 : 500) were obtained from Santa Cruz Biotechnology (sc-93, sc-535 and sc-474, respectively; Santa Cruz, CA, USA), antiserum against total MEK1 (1 : 1000) was from Biosource (AH01102), monoclonal antibody ERK-PT115 (1 : 500) against threonine-phosphorylated ERK1/2 was from Sigma-Aldrich (St Louis, MO, USA) (M7802), antiserum against Elk-1 (1 : 1000) and monoclonal phospho-specific antibodies against ERK1/2 (1 : 2000), p38 (1 : 2000), JNK1 (1 : 500), MEK1/2 (1 : 1000), Elk-1 (1 : 1000) and S6 ribosomal protein (1 : 500) were from Cell Signalling Technology (Beverly, MA, USA) (nos 9182, 9106, 9216, 9255, 9121, 9186 and 4856, respectively). The rabbit polyclonal 28 kD-calbindin antiserum (1 : 500) was from Sigma-Aldrich (C7354). Goat anti-Rabbit and anti-Mouse HRP-conjugated secondary antibodies (1 : 20 000; Pierce Biotechnology Inc., Rockford, IL, USA) were used for chemiluminescent detection on immunoblots. Alternatively, 1 : 2000 dilutions of Goat anti-Mouse and Goat anti-rabbit secondary antibodies conjugated to Alexa Fluor 680 (Molecular Probes, Eugene, OR, USA) or IRDye800 (Rockland Immunochemicals, Gilbertsville, PA, USA) fluorescent dyes were used for detection and quantification with the Odyssey infrared imaging system (LI-COR). Donkey anti-Mouse and anti-Rabbit biotin-conjugated IgG (1 : 250; Pierce) secondary antibodies in conjunction with Vector ABC Elite staining protocols (Vector Laboratories, Burlingame, CA, USA) served in immunohistochemical analyses.

Immunoblot analyses

Protein lysates of whole brain or separated cerebrum and cerebellum from individual wild-type, heterozygous, and knockout mice were prepared by homogenization in lysis buffer: 100 mmol/L Na2PO4 pH 8.1, 1% Triton X-100, 2 mmol/L EDTA, 1 mmol/L phenylmethylsulphonyl fluoride (PMSF), 0.1 mmol/L Na3VO4, 25 mmol/L NaF, 1 mmol/L Na4P2O7 and 1 protease inhibitor cocktail tablet (Roche Diagnostics, Mannheim, Germany) per 50 mL. Homogenates were then incubated at 4°C for 60 min while rotating, followed by a 10 min centrifugation at 16 000 g. Protein concentration in the resulting supernatant fraction was determined spectrophotometrically (Bradford 1976). As much as 50 $\mu$g of protein per sample was subjected to SDS-PAGE and electroblotted onto PVDF membranes. Subsequent blocking and chemiluminescent immunodetection was as described previously (Chirivi et al. 2004). In some experiments, first immunoprecipitations with monoclonal antibody 6A6 (Chirivi et al. 2004) or with rabbit antiserum α-BR7 (2 µL per sample; Dilaver et al. 2007) from pre-cleared lysates (500 µL) were performed before the immunoblot procedure. For the ratio-imaging of relative phosphorylation levels, brain lysates derived from individual mice (30 µg of protein per sample) were subjected to 10% SDS-PAGE and size-separated proteins were blotted onto PVDF membranes. Blots, containing six wildtype and six PTPRR-deficient samples for a given tissue, were blocked for 30 min using Odyssey blockbuffer (LI-COR) and incubated overnight at 4°C with pairs of rabbit and mouse antibodies directed against total and phosphorylated proteins using standard procedures. Immunodetection with Alexa Fluor 680 or IRDye800 conjugated secondary antibodies enabled quantifiable data acquisition on the Odyssey infrared imaging system (LI-COR). Results are expressed as the ratio of the signal intensities detected by the phospho-specific antibody and the antibody that is immunoreactive also to the unphosphorylated version of the protein, respectively.

Immunohistochemistry

Mice, 5–7 months of age, were anaesthetized and perfused, and brain tissue cryosections were prepared and processed for immunohistochemistry essentially as described (van der Zee et al. 2003). In brief, 8 μm coronal cryosections were incubated with monoclonal antibody 5E4 (Chirivi et al. 2004) and donkey-anti-mouse-biotin-conjugated IgG, and immunoreactivity was visualized using Vector ABC Elite with AEC as a substrate. Subsequently, sections were rinsed with water, incubated in 0.9% NaCl and 0.5% CuSO4 for 5 min, and counterstained with haematoxylin for 3 min. Finally, sections were dehydrated with ethanol and xylene, embedded in Eukitt (Electron Microscopy Sciences) and examined by light microscopy. For calbindin and biphosphorylated ERK1/2 immunostaining, paraformaldehyde-fixed brain slices were prepared and used essentially as described (van der Zee et al. 2003). In brief, sagittal vibratome sections (40 μm thick) of perfusion-fixed mouse cerebellum were stained free-floating using 28 kD calbindin or phospho-ERK1/2 rabbit antibodies and biotin-conjugated donkey-anti-rabbit secondary antibody. Visualization was performed with Elite Vectastain and DAB-Ni as substrate, resulting in a black precipitate. Haematoxilin/eosin staining was performed on the alternate sagittal 40 μm brain sections according to standard procedures. Haematoxilin/eosin and calbindin-antibody stained images were collected using a Dialux 20 microscope (Leitz) with a video camera attached to a computer, and analysed using the software package PC-Image. Cerebellum Purkinje cell numbers were counted along 5–6 different 1.7 mm long lobule stretches per section and in five sections for each mouse. Numbers were averaged and expressed as the number of Purkinje cells per mm lobule stretch per mouse, and per genotype. Digitised images of phospho-ERK1/2-antibody stained brain sections were analysed by outlining the clearly phospho-ERK1/2-positive areas of the cerebellum Purkinje cells (mostly at the basal side) and recording the corresponding optical density and area values. A total of 50–70 Purkinje cells were measured for each mouse with density values ranging between 100 and 243 (in 1–256 grey level mode). Phospho-ERK1/2-positive neurons in the motor cortex, striatum, nucleus accumbens shell and lateral globus pallidus were analysed similarly, using 35 cells per region per mouse, with three mice per genotype. The average immunodensity value per genotype group was expressed as phospho-ERK1/2 density/100 μm2.

Motor coordination and balance tests

All tests were performed on age-, sex- and weight-matched animals. Detailed protocols are provided in Appendix S1. The rotarod and the rope grip tests were performed as described in Kolkman et al. (2004) and Peled-Kamar et al. (1997), respectively, with minor modifications (see Appendix S1). In the single bar and parallel bars tests (Ding et al. 2002) mice transversed horizontally suspended rods towards their home cage. Two trials per test were performed and ‘traverse time’ (in second) and number of times a hindpaw slipped from the rod were registered. In addition, for the single bar test also the number of mice falling off was recorded per genotype group. Gait assessment (Hemsley and Hopwood 2005) was performed by analysing the hind-limb walking pattern of individual mice that walked through a small alley, using the animal’s hind feet print pattern as recorded on photographic paper. From at least three footprint pairs the following parameters were measured (in mm): distance length from left to right foot (LTRF) and from right to left foot (RTLF) (van der Zee et al. 2003). Aberrations from the normally equal distances between the footprints were calculated [100 − (LTRF/RTLF) × 100%) and expressed as ‘% foot mismatch’.

Basal activity and explorative behaviour

Basal nocturnal activity (i.e. from 15:00 h until 9:00 h the next morning) of mice was monitored individually using an activity cage (36 × 24 × 25 cm) equipped with three photoelectric cells 2 cm above the grid floor (Kolkman et al. 2004). The activity data of the mice were recorded per min, summed per hour, and presented as total activity counts per hour for each genotype group. Explorative behaviour of individual mice in a square open field (48 × 48 × 40 cm, with white Plexiglas walls) was videotaped for 30 min and subsequently analysed in three blocks of 10 min for the duration (in seconds) of walking, wall leaning, rearing, sitting and grooming (Kolkman et al. 2004). In addition, the accumulative total walking distance in the 30 min period was calculated by applying a computer-assisted walking pattern analysis of the video-taped sessions (Kolkman et al. 2004).

Statistical analysis

All data are presented as means ± SEM per genotype group, and the statistical analyses involved Student’s t-test except for data presented in Figs 6c and f (chi-square test) and Fig. 6h (anova repeated measures). All tests were performed using the SPSS 11.0 statistical software package. Statistical significance was set at p < 0.05.

image

Figure 6.  Motor coordination and balance defects in Ptprr−/− mice. (a) Rotarod. The time wildtype (+/+, white bars) and Ptprr−/− (−/−, black bars) mice remained on the accelerating rotarod, as measured during three trials, is depicted as mean ± SEM (*p < 0.015). (b) Rope grip test. The time wildtype and Ptprr−/− mice (n = 9 per group) were holding onto a horizontally tightly suspended rope, measured during three trials, is depicted as mean ± SEM (*p < 0.031). (c) The percentage of mice falling off during the 3 Rope grip trials is shown for both genotypes (chi-square test, *p < 0.01). Parallel bars. The mean ± SEM number of slips that mice show while traversing parallel bars (d, *p < 0.04) or a single bar (e, *p < 0.05) during two trials is indicated for both genotypes. (f) The percentage of mice falling off the single bar during the two trials is shown for each group (chi-square test, *p < 0.02). (g) Hindlimb gait. Aberrations in hindfoot placing of Ptprr−/− (−/−) mice and wildtype controls (+/+) that were walking through an alley are depicted as mean ‘% foot mismatch’ ± SEM (*p < 0.002). (h) Activity cage. General basal activity of wildtype (+/+, open circles) and Ptprr−/− (−/−, filled squares) mice (n = 8 per group) was measured overnight for 15 h continuously, and is expressed as mean total counts per 60 min ± SEM [anova repeated measures, F(1,14) = 13.831; *p < 0.002]. For a, d, e, f, g: wildtype, n = 12; Ptprr−/−, n = 11.

Download figure to PowerPoint

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Generation of Ptprr−/− mice

The structure of mouse gene Ptprr, having three distinct transcription start sites that are some 150 kilobasepairs (kbp) apart (Chirivi et al. 2004), impelled us to devise a replacement-type gene targeting strategy involving deletion of exon 6 (Fig. 1a). Following homologous recombination in mouse embryonic stem cells the exon 6 deletion will shift the reading frame in any residual transcript coming from the Ptprr promoters despite the presence of the antiparallel selection marker transcription unit, thus precluding the formation of catalytically active or MAPK-binding PTPRR mutant proteins from the targeted locus. Presence of the wildtype and mutant Ptprr alleles in resulting animals and their offspring was monitored by Southern blotting (Fig. 1b) and PCR analysis of genomic DNA. Allele distribution in the offspring (n = 100) of the heterozygous crosses was according to Mendelian law (data not shown) and size and gross morphology of heterozygous and Ptprr−/− animals was indistinguishable from that of their wildtype littermates. Subsequent breeding steps also demonstrated no overt effects on fertility.

image

Figure 1. Ptprr targeted disruption. (a) The structure of mouse gene Ptprr (upper) and the targeting strategy (lower) is depicted. Genomic DNA is represented by a grey horizontal bar and exons are depicted as vertical bars with exon number or name indicated above. The three distinct transcription start sites (Chirivi et al. 2004) are represented by hooked arrows. The PTPBR7 transmembrane domain is encoded within exons 5 and 6, the KIM domain codons originate from exons 6 and 7, and the catalytic PTP domain coding region spans exons 8–14. DNA stretches used to obtain homologues recombination are boxed and named 5′- and 3′-arm, respectively. A neomycin resistance cassette (Neo) in the opposite transcriptional orientation replaces exon 6 in the targeting construct, which at the 5′ end also harbours the diphtheria toxin cassette (DT) for negative selection purposes. Two-headed arrows indicate BamHI restriction fragments detected by diagnostic probes 5′p and 3′p on Southern blots. (b) Southern analysis of BamHI-digested genomic DNA from liver of wildtype (+/+), heterozygous (+/−), and Ptprr−/− (−/−) mice. Hybridisation with the 5′ (left) and 3′ (right) diagnostic probes revealed the wildtype 8.6-kbp fragment and a mutant 2.0- or 6.9-kbp fragment, respectively. (c) Northern analysis of total RNA extracted from cerebrum (cbr) or cerebellum (cbl) of wildtype, heterozygous, or Ptprr−/− mice using a probe that detects exons 5–14. (d) Western analysis of proteins, immunoprecipitated by monoclonal antibody 6A6 from brain extracts of wildtype, heterozygous, and Ptprr−/− mice, using rabbit α-SL antiserum. Upper bands of around 75, 70 and 65 kDa in size represent PTPBR7-derived processed proteins; the 65 kDa species may also contain PTP-SL; and the smaller immunoreactive proteins correspond to PTPPBS-γ isoforms (Chirivi et al. 2004).

Download figure to PowerPoint

PTPRR messenger and protein isoforms are absent in Ptprr−/− mice

To assess the effects of the targeted mutation at the RNA level, total RNA was isolated from cerebral and cerebellar brain parts, and analysed by Northern blotting (Fig. 1c). Use of a cDNA probe that spans exon 5 through 14 revealed the presence of the 4.1 kb PTPBR7 transcript in both brain areas of wildtype and heterozygous animals. The shorter c. 3 kb transcript that is detected in cerebellum represents PTP-SL or PTPPBSγ mRNAs. These transcripts are all absent in RNA samples obtained from Ptprr−/− mice. The somewhat smaller RNA product observed in knockout cerebellar RNA was found to represent a Ptprr-derived transcript that lacks exon 6-derived sequences (data not shown). If translated at all, this messenger would at best result in a truncated PTPRR mutant protein lacking the KIM and PTP domain.

To investigate the phenotypic consequences on the protein level, PTPRR isoforms were immunoprecipitated from mouse brain lysates to ensure maximum sensitivity, and analysed on blot using immunotools that are reactive towards all PTPRR proteins. As expected, no immunoreactivity was detected in Ptprr-/–derived material (Fig. 1d). We also exploited a polyclonal antibody raised against the extracellular domain of PTPBR7 (Dilaver et al. 2007; a kind gift of Drs R. Pulido and P. Ríos, Valencia, Spain) for immunoprecipitation and immunoblot detection of any truncated PTPRR mutant but, apart from immunoglobulin chain background signals, we did not detect immunoreactivity in Ptprr−/− samples (Figure S1) firmly establishing that these mice lack PTPRR phosphatase activity and MAP kinase-binding potency. Absence of PTPRR protein had no effect on the expression levels of the closest homologous PTP family, STEP, as determined by immunoblot analyses of brain lysates (data not shown) using a STEP-specific antiserum (kindly provided by Dr R. Pulido, Valencia, Spain).

Unaltered brain and Purkinje cell morphology in Ptprr−/−mice

Haematoxylin-eosin stained brain sections demonstrated a normal cortical and hippocampal layering in the cerebrum of wildtype and Ptprr−/− mice, and cerebellum granule and molecular layers also had a normal appearance (Fig. 2a–d). Immunohistochemical staining with antibody 5E4 (Chirivi et al. 2004) confirmed that PTPRR is predominantly present in the cerebellum Purkinje cells of wildtype mice and absent in the Ptprr−/− Purkinje cell layer (Fig. 2e,f). Calbindin immunostaining did not reveal any difference in Purkinje cell body and dendritic tree details in wildtype and Ptprr−/− cerebellum (Fig. 2g–j). Numbers of calbindin-positive or haematoxilin-eosin stained Purkinje cells per mm lobule stretch in the cerebellum were found to be 52.5 ± 6 in wildtype (n = 4) and 49.0 ± 9 in Ptprr−/− (n = 4) mice, respectively. Thus, both morphology and number of Purkinje cells is unaffected in mice deficient for PTPRR proteins.

image

Figure 2.  Normal brain and cerebellum Purkinje cell morphology in Ptprr−/− mice. Haematoxylin/eosin stained sagittal brain sections displaying similar cortex, striatum, corpus callosum, fimbria fornix and hippocampus layering and morphology in wildtype (a) and Ptprr−/− (b) mice are shown. Cerebellum cortex granular, Purkinje and molecular layers in wildtype (c) and Ptprr−/− (d) mice are similar as well. Immunostaining for PTPRR revealed specific staining of Purkinje cells in wildtype (e), but not Ptprr−/− (f), mice. Purkinje cell staining with calbindin (28 kD) antibody demonstrated very similar cell numbers and branching morphology in both wildtype (g, i) and Ptprr−/− (h, j) mice. Bars represent 500 (a, b), 425 (c, d), 50 (e–h) and 40 (i, j) μm, respectively.

Download figure to PowerPoint

PTPRR deficiency results in ERK1/2 hyperphosphorylation

The reported interaction of PTPRR isoforms with several MAPKs (Pulido et al. 1998; Ogata et al. 1999; Buschbeck et al. 2002; Munoz et al. 2003) prompted us to investigate MAPK activity in mouse brain lysates through the monitoring of MAPK biphosphorylation levels. Equal amounts of protein, from tissue lysates obtained from Ptprr−/− animals or wildtype controls, were size-separated using SDS-PAGE and blotted onto PVDF membranes. Subsequently, pairs of phosphospecific MAPK antibodies and total MAPK antisera were used in combination with fluorescent second antibodies for quantitative detection. Ratio-imaging of biphosphorylated and total ERK1/2 levels (Fig. 3a) disclosed a highly significant hyperphosphorylation (p < 0.0001) of these MAP kinases in both cerebellum and cerebrum tissue lysates of Ptprr−/− animals (n = 7) when compared with wildtypes (n = 7). No differences were apparent in ERK1/2 total protein levels between wildtype and PTPRR deficient samples. Importantly, the use of an antibody immunoreactive towards phospho-threonine-containing ERK1/2 did not reveal any significant differences between the two genotype groups (data not shown), demonstrating that it is the level of tyrosine phosphorylation that is altered and is causing ERK hyperactivation in PTPRR deficient brain material. It is mainly ERK2 that is hyperactivated in cerebellum of PTPRR deficient mice, whereas ERK2 and to a lesser extend ERK1 both display increased phosphorylation ratios in cerebrum (data not shown). Although PTPRR proteins can bind to p38 (Zúñiga et al. 1999) no alteration in phosphorylation levels was discernible for this MAPK (Fig. 3b). Surprisingly, a slight but significant reduction in phosphorylation levels for the stress-related c-Jun kinases JNK1/2 was obtained in lysates from Ptprr−/− cerebellum (Fig. 3c), despite the lack of a direct interaction with PTPRR isoforms (Zúñiga et al. 1999).

image

Figure 3.  Altered MAPK phosphorylation levels in Ptprr−/− mouse brain. Degree of ERK1/2 (a), p38 (b) and JNK1/2 (c) biphosphorylation in cerebellum and cerebrum of wildtype (+/+, white bars) and Ptprr−/− (−/−, black bars) mice as determined by Western analyses of tissue lysates from individual mice (n = 6 for each tissue and genotype group) using phospho-specific antibodies. Representative examples (three out of the six for each tissue and genotype group) of phosphorylated ERK1/p44 and ERK2/p42 (pERK) and total ERK1/2 (totERK) immunofluorescent signals are depicted above the corresponding bars. Phosphorylated ERK1/2, p38 and JNK1/2 signals were normalized to total ERK1/2, p38 and JNK1/2 levels, respectively, and plotted as relative phosporylation levels. Results are the mean ± SEM of six samples per group and are representative of at least three independent experiments (*p < 0.0001, **p < 0.01).

Download figure to PowerPoint

To pinpoint the increased levels of phospho-ERK1/2 at the histological level, we analysed brain sections immunostained for phospho-ERK1/2 (Fig. 4 and data not shown). In cerebrum, phospho-ERK1/2-positive neurons were the most notable in motor cortex, striatum, nucleus accumbens shell and lateral globus pallidus areas, but densitometric analyses did not reveal significant differences between wildtype and PTPRR deficient animals. In cerebellum, phospho-ERK1/2-positive staining was visible in wildtype Purkinje cells and mostly confined to the basal portion of the cytoplasm including the axon hillock region (Fig. 4a), which is in line with published data (Zsarnovszky and Belcher 2004). The phospho-ERK1/2 immunostaining of Purkinje cells in Ptprr−/− mice demonstrated a similar localization but at a higher intensity (Fig. 4b). Measurements of the phospho-ERK1/2 densities revealed values of Purkinje cell staining in wildtypes (n = 3) ranging between 148 and 176, and in Ptprr−/− mice (n = 3) between 195 and 243. Thus, the absence of PTPRR isoforms results in a significant, 37% increase in phospho-ERK1/2 density levels in Purkinje cells (Fig. 4c; p < 0.023), firmly corroborating a physiological link between PTPRR and ERK1/2 signalling.

image

Figure 4.  ERK1/2 hyperphosphorylation in Ptprr−/− cerebellum Purkinje cells. Immunostaining for biphosphorylated, active ERK1/2, confined to the basal portion and the axon hillock of wildtype (a) and Ptprr−/− (b) Purkinje cells. Bars represent 50 μm in a1 and b1, and 20 μm in a2 and b2 (for a2–a5 and b2–b5). (c) Quantification of phospho-ERK1/2 staining density demonstrated a significant increase of 37% in Ptprr−/− Purkinje cells (−/−) when compared with wildtype (+/+) specimen (n = 3; *p < 0.023).

Download figure to PowerPoint

PTPRR affects ERK1/2 signalling downstream of MEK

The observed hyperphosphorylated state of ERK1/2 in Ptprr−/− brain parts triggered us to investigate the phosphorylation levels of the upstream MAPK kinases MEK1/2 and of potential downstream ERK1/2 targets, such as the transcription factors Elk-1 and c-myc. Antibodies directed against phosphorylated and total MEK1/2 protein demonstrated equal MEK1/2 phosphorylation levels upon immunoblot analysis of wildtype and knockout brain lysates (Fig. 5a). An Elk-1 phospho/total antibody combination revealed higher phospho-Elk-1 levels in Ptprr−/− cerebellum when compared with wildtype samples, and in PTPRR deficient cerebrum this Elk-1 hyperphosphorylation reached significance (Fig. 5b; p < 0.02). In contrast, c-myc phosphorylation levels were not affected in Ptprr−/− brain lysates (data not shown). We also screened for effects of PTPRR depletion on the PI(3)K-Akt signalling pathway. Phospho-specific Akt and p90RSK antibodies did not provide reliable signals on blot (data not shown) but S6 ribosomal protein phosphorylation levels were found to be unaffected in PTPRR deficient animals (Fig. 5c). Thus, alterations in phosphosignalling relays in our PTPRR deficient mice appear specific for the MAPK signalling circuitry downstream of MEK1/2, and warrant further studies towards possible functional repercussions.

image

Figure 5.  PTPRR effects on ERK1/2 signalling pathways are downstream of MEK. MEK1/2 (a), Elk-1 (b) and S6 ribosomal protein (c) phosphorylation levels were analysed using phospho-specific antibodies on blots containing cerebellum or cerebrum tissue lysates of individual wildtype (+/+, white bars) and Ptprr−/− mice (−/−, black bars). Results, representative for three independent experiments, were normalized to total MEK, Elk, or tubulin protein levels, respectively, and are presented as the mean ± SEM of six samples per group (*p < 0.02).

Download figure to PowerPoint

Dysfunction in balance and motor coordination in Ptprr−/− mice

Given the predominant expression of PTPRR in Purkinje cells (Chirivi et al. 2004) and the observed hyperphosphorylation of ERK1/2 in PTPRR deficient cerebellum (Figs 3 and 4), we performed several different motor function tests to assess whether fine motor movement and balance function are impaired in Ptprr−/− mice. On the accelerating rotarod Ptprr−/− mice were falling off much earlier, at much lower rod turning speeds, as their littermate controls (Fig. 6a) indicating reduced motor coordination and balance skills. A similar trend was obvious in the rope grip test (Fig. 6b,c). While most wildtype animals could hang onto the rope for almost 120 s (with only five falls out of a total of 27 trials), Ptprr−/− mice lost their grip already after some 90 s (15 falls in 27 trials).

Fine motor movement and balance were also tested by letting mice walk towards their home cage over suspended 60 cm long parallel bars, or over the more difficult 60 cm long single bar. The time to traverse the parallel bars, or the single bar, was not significantly different between wildtype (14 ± 2 and 16 ± 3 s, respectively) and Ptprr−/− (20 ± 4 and 20 ± 3 s, respectively) mice. However, Ptprr−/− mice made significantly more slips (Fig. 6d,e), and seven of 11 Ptprr−/− mice were falling off the single bar as compared to two of 12 wildtype mice (Fig. 6f). We next performed hindlimb gait analyses to picture any possible uneven walking. Wildtype animals plant both their feet at a regular and equal distance while walking through a narrow alley, with only a small percentage of foot mismatch (6 ± 0.6%). Ptprr−/−mice, however, demonstrated a clear aberration (25 ± 6% foot mismatch) indicating that they move somewhat asymmetrically as if they are limping with the right or left foot (Fig. 6g).

Novel-environment exploration behaviour is not affected in Ptprr−/− mice

Two additional tests were performed to monitor possible effects of PTPRR deficiency on behaviour. General basal locomotion activity of the animals was measured overnight using a so-called Activity Cage that contains motility sensors. In the period from 18:00 h until 09:00 h the Ptprr−/− mice were far less active than wildtype littermates, which showed a normal increased activity in the nocturnal period (Fig. 6h). Specific exploration behaviour was observed in detail during a 30 min-period using an Open Field set-up. Walking, rearing, wall leaning, grooming and sitting activities were scored for individual mice (n = 9 per genotype group, Table 1), but no significant differences in any of the parameters were observed between wildtype and Ptprr−/− mice. Moreover, the total cumulative distance walked in 30 min was not significantly different between wildtype and Ptprr−/− animals (59.4 ± 7 vs. 44.3 ± 8 m, respectively). Thus, although the PTPRR deficient animals are less active than wildtype mice in a familiar surrounding, novel-environment induced exploration behaviour appears unaltered in Ptprr−/− mice.

Table 1.   Open field exploration behaviour of Ptprr−/− mice is normal. The average time (in sec per 10 min period or for the total of 30 min) mice spent on the five different open field parameters was not significantly different between wildtype and Ptprr−/− mice
 0–10 min10–20 min20–30 min0–30 min
Wildtype (n = 9)
 Walking252 ± 37202 ± 18189 ± 27642 ± 76
 Rearing11 ± 423 ± 633 ± 1167 ± 21
 Wall leaning33 ± 837 ± 938 ± 10108 ± 26
 Grooming46 ± 9106 ± 11112 ± 19264 ± 34
 Sitting256 ± 38232 ± 24229 ± 31717 ± 82
Ptprr−/− (n = 9)
 Walking220 ± 35158 ± 26133 ± 29511 ± 83
 Rearing5 ± 29 ± 516 ± 830 ± 15
 Wall leaning14 ± 423 ± 724 ± 961 ± 19
 Grooming50 ± 1681 ± 1789 ± 29220 ± 35
 Sitting310 ± 37329 ± 35338 ± 53977 ± 116

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

In this study, we demonstrate that adult mice that lack PTPRR proteins (PTPBR7, PTP-SL, PTPPBSγ-42 and PTPPBSγ-37; Chirivi et al. 2004) display altered MAPK phosphorylation levels in neuronal tissues, reduced basal locomotive activity and specific motor coordination deficits. These impairments include fine motor movement dysfunction, reduced neuromuscular grip strength, and diminished balance skills, all features that are reminiscent of those seen in ataxic syndrome mouse models. Novel-environment induced exploration behaviour is not affected in Ptprr−/− mice. This suggests that mostly cerebellar functions are impaired, and is in line with the predominant expression of PTPRR in adult cerebellum Purkinje cells. The total number of Purkinje cells or their extent of dendritic branching; however, is unaltered in Ptprr−/− mice. In addition, 20 months old Ptprr−/− mice still showed the ataxia features but no histological signs of neurodegeneration (CEEMvdZ, personal observations).

Multiple ataxia mouse models display histological aberrations (such as loss of cerebellar granule cells or Purkinje cells) that provide an explanation for cerebellum dysfunction. Mutant mice that display ataxic behaviour but have apparently normal cerebellar layering, Purkinje cell numbers and dendritic spine morphology, however, are limited to junctophilin 3 (Nishi et al. 2002), carbonic anhydrase-related protein VIII (Jiao et al. 2005), calbindin and calretinin (Cheron et al. 2004) deficient animals. The cause of ataxia in these mice may converge on calcium homeostasis related with Purkinje cell physiology. In line with this, the human conditions episodic ataxia 2, familial hemiplegic migraine 1 and spinocerebellar ataxia 6, all result from mutations in the pore-forming subunit of P/Q-type voltage-gated calcium channels (Pietrobon 2005). Several other ataxia syndromes are directly linked to alterations in the neuronal phosphoproteome. Familial spinocerebellar ataxia types SCA12 and SCA14, for example, are caused by mutations in genes encoding the phosphatase PP2A regulatory subunit PR55/Bβ (Holmes et al. 1999) and the kinase PKCγ (van de Warrenburg et al. 2003; Yabe et al. 2003), respectively. Perhaps these mutations lead to altered phosphorylation levels of cerebellum calcium channels and receptors that function in synaptic connectivity and plasticity, thereby changing their membrane localization and activity (Catterall 2000; Nakanishi 2005). Although this mostly concerns serine/threonine phosphorylation, the phosphotyrosine-specific PTPRR isoforms may indirectly contribute to this via the action of their substrates ERK1 and ERK2. MAPK signalling cascades are indeed acknowledged as key processes in synaptic plasticity (Sweatt 2004; Thomas and Huganir 2004), and aberrant ERK1/2 signalling activity has been witnessed in several neuropathologies (Colucci-D’Amato et al. 2003).

The elevated phospho-ERK1/2 levels in Ptprr−/− mice may in turn also have a more indirect effect on locomotion, through changes in the transcriptome of affected cells. Indeed, we observed hyperphosphorylation of the transcription factor Elk-1 in cerebrum and it seems reasonable to assume that in cerebellar neurons altered phosphorylation levels of as yet undefined nuclear ERK1/2 substrates may lead to multiple changes at the protein level as well. Perhaps such secondary effects should even be held responsible for the unexpected down-tuning of JNK1/2 activity in Ptprr−/− cerebellum and perhaps neutralizes any effects on p38 activity. Alternatively, the absence of changes in p38 phosphorylation levels in our knockout mice could be taken as evidence that under physiological conditions PTPRR isoforms do not interact with nor dephosphorylate this MAPK.

In the cerebellum of Ptprr−/− mice, increased phospho-ERK1/2 immunostaining was detected predominantly at the basal portions of the Purkinje cell soma, including the axon hillock region. This is the region where inhibitory basket cell afferents regulate Purkinje cell firing via synaptic transmission, and subtle irregularities in Purkinje cell pacemaking are sufficient to cause cerebellar dysfunction and ataxia (Walter et al. 2006). Our immunoblot analyses indicate ERK1/2 hyperphosphorylation in other brain regions as well, in line with the broad expression pattern of the Ptprr gene (van den Maagdenberg et al. 1999; Chirivi et al. 2004), which coins the question whether phenotypic consequences in addition to locomotion disturbance may be uncovered. One should take into account, however, that not only PTPRR proteins, but also multiple other KIM-domain-containing phosphatases, including the phosphotyrosine-specific STEP and HePTP/LC-PTP and a large collection of DUSPs, are offering ways to regulate MAPK activity levels (Tanoue and Nishida 2003; Tárrega et al. 2005). Moreover, genetic analyses in Drosophila demonstrated that DSPs and PTPs may cooperate and perform partially redundant functions as ERK phosphatases in a cell type specific way (Rintelen et al. 2003). The closely homologous striatum-enriched phosphatase (STEP) family of PTPs may well be capable of taking over part of the signalling functions of PTPRR isoforms in certain brain areas, although the binding preferences of these two MAP kinase phosphatase families towards ERK1/2 differ to some extent (Pulido et al. 1998; Munoz et al. 2003; Tárrega et al. 2005). It is unlikely that HePTP/LC-PTP is compensating in some cells for PTPRR as this PTP is not expressed in neuronal tissue but rather restricted to the haematopoietic system (Adachi et al. 1992; Zanke et al. 1994).

The current study firmly establishes the ERK1/2 MAP kinases as physiologically relevant PTPRR substrates. To what extent the four PTPRR isoforms individually contribute to this remains to be established. Its broad neuronal expression, including during early mouse development (van den Maagdenberg et al. 1999), and its potential to receive extracellular signals make the receptor-type isoform PTPBR7 a tempting candidate to account for many of the effects. However, highest PTPRR protein levels are found in Purkinje cells (Chirivi et al. 2004) that are intimately connected with fine motor coordination, but here the predominant transcript encodes the vesicle-associated PTP-SL isoform (van den Maagdenberg et al. 1999). Clearly, additional experiments are needed to disclose further the signalling specificity within this PTP family.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We like to thank C. Verbaas and colleagues at the Central Animal Facility of the Radboud University Nijmegen for performing blastocyst injections and embryo transfers, and for housing the mice. L. Lubbers and M. Verheij are thanked for their assistance with the activity cage and walking pattern analyses, I. Otte-Holler for assisting in the use of the LICOR Odyssey Fluorescence Imager, and R. Bindels for providing the anti-28 kD-calbindin antiserum. Special thanks go to the members of the EU Research Training Network on Neuronal Protein Tyrosine Phosphatases, especially J. den Hertog (Utrecht, The Netherlands) and R. Pulido and P. Ríos (Valencia, Spain) for valuable discussions and provision of immunotools, and A.W. Stoker (London, UK) for critical reading of the manuscript. This work was supported in part by grants from the Radboud University Nijmegen Medical Centre (III.120201) and the EU (HPRN-CT-2000-00085) to W.H.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • Adachi M., Sekiya M., Isobe M., Kumura Y., Ogita Z., Hinoda Y., Imai K. and Yachi A. (1992) Molecular cloning and chromosomal mapping of a human protein-tyrosine phosphatase LC-PTP. Biochem. Biophys. Res. Commun. 186, 16071615.
  • Bradford M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248254.
  • Buschbeck M., Eickhoff J., Sommer M. N. and Ullrich A. (2002) Phosphotyrosine-specific phosphatase PTP-SL regulates the ERK5 signaling pathway. J. Biol. Chem. 277, 29 50329 509.
  • Catterall W. A. (2000) Structure and regulation of voltage-gated Ca2+ channels. Annu. Rev. Cell Dev. Biol. 16, 521555.
  • Cheron G., Gall D., Servais L., Dan B., Maex R. and Schiffmann S. N. (2004) Inactivation of calcium-binding protein genes induces 160 Hz oscillations in the cerebellar cortex of alert mice. J. Neurosci. 24, 434441.
  • Chirivi R. G. S., Dilaver G., Van De Vorstenbosch R., Wanschers B., Schepens J., Croes H., Fransen J. and Hendriks W. (2004) Characterization of multiple transcripts and isoforms derived from the mouse protein tyrosine phosphatase gene Ptprr. Genes Cells 9, 919933.
  • Chizhikov V. and Millen K. J. (2003) Development and malformations of the cerebellum in mice. Mol. Genet. Metab. 80, 5465.
  • Church G. M. and Gilbert W. (1984) Genomic sequencing. Proc. Natl Acad. Sci. U.S.A. 81, 19911995.
  • Colucci-D’Amato L., Perrone-Capano C. and Di Porzio U. (2003) Chronic activation of ERK and neurodegenerative diseases. Bioessays 25, 10851095.
  • Dilaver G., Van De Vorstenbosch R., Tárrega C., Ríos P., Pulido R., Van Aerde K., Fransen J. and Hendriks W. (2007) Proteolytic processing of the receptor-type protein tyrosine phosphatase PTPBR7. FEBS J. 274, 96108.
  • Ding Y., Zhou Y., Lai Q., Li J., Park H. and Diaz F. G. (2002) Impaired motor activity and motor learning function in rat with middle cerebral artery occlusion. Behav. Brain Res. 132, 2936.
  • Gee C. E. and Mansuy I. M. (2005) Protein phosphatases and their potential implications in neuroprotective processes. Cell. Mol. Life Sci. 62, 11201130.
  • Gronda M., Arab S., Iafrate B., Suzuki H. and Zanke B. W. (2001) Hematopoietic protein tyrosine phosphatase suppresses extracellular stimulus-regulated kinase activation. Mol. Cell. Biol. 21, 68516858.
  • Hemsley K. M. and Hopwood J. J. (2005) Development of motor deficits in a murine model of mucopolysaccharidosis type IIIA (MPS-IIIA). Behav. Brain Res. 158, 191199.
  • Holmes S. E., O’Hearn E. E., McInnis M. G., Gorelick-Feldman D. A., Kleiderlein J. J., Callahan C., Kwak N. G., Ingersoll-Ashworth R. G., Sherr M., Sumner A. J., Sharp A. H., Ananth U., Seltzer W. K., Boss M. A., Vieria-Saecker A. M., Epplen J. T., Riess O., Ross C. A. and Margolis R. L. (1999) Expansion of a novel CAG trinucleotide repeat in the 5’ region of PPP2R2B is associated with SCA12. Nat. Genet. 23, 391392.
  • Jiao Y., Yan J., Zhao Y., Donahue L. R., Beamer W. G., Li X., Roe B. A., Ledoux M. S. and Gu W. (2005) Carbonic anhydrase-related protein VIII deficiency is associated with a distinctive lifelong gait disorder in waddles mice. Genetics 171, 12391246.
  • Kaplan D. R. and Miller F. D. (2000) Neurotrophin signal transduction in the nervous system. Curr. Opin. Neurobiol. 10, 381391.
  • Karim F. D. and Rubin G. M. (1999) PTP-ER, a novel tyrosine phosphatase, functions downstream of Ras1 to downregulate MAP kinase during Drosophila eye development. Mol. Cell 3, 741750.
  • Kolkman M. J. M., Streijger F., Linkels M., Bloemen M., Heeren D., Hendriks W. J. A. J. and Van Der Zee C. E. E. M. (2004) Mice lacking Leukocyte common Antigen-Related (LAR) protein tyrosine phosphatase domains demonstrate spatial learning impairment in the two-trial water maze and hyperactivity in multiple behavioural tests. Behav. Brain Res. 154, 171182.
  • Munoz J. J., Tárrega C., Blanco-Aparicio C. and Pulido R. (2003) Differential interaction of the tyrosine phosphatases PTP-SL, STEP and HePTP with the mitogen-activated protein kinases ERK1/2 and p38alpha is determined by a kinase specificity sequence and influenced by reducing agents. Biochem. J. 372, 193201.
  • Nakanishi S. (2005) Synaptic mechanisms of the cerebellar cortical network. Trends Neurosci. 28, 93100.
  • Nishi M., Hashimoto K., Kuriyama K., Komazaki S., Kano M., Shibata S. and Takeshima H. (2002) Motor discoordination in mutant mice lacking junctophilin type 3. Biochem. Biophys. Res. Commun. 292, 318324.
  • Ogata M., Oh-hora M., Kosugi A. and Hamaoka T. (1999) Inactivation of mitogen-activated protein kinases by a mammalian tyrosine-specific phosphatase, PTPBR7. Biochem. Biophys. Res. Commun. 256, 5256.
  • Paul S., Nairn A. C., Wang P. and Lombroso P. J. (2003) NMDA-mediated activation of the tyrosine phosphatase STEP regulates the duration of ERK signaling. Nat. Neurosci. 6, 3442.
  • Peled-Kamar M., Lotem J., Wirguin I., Weiner L., Hermalin A. and Groner Y. (1997) Oxidative stress mediates impairment of muscle function in transgenic mice with elevated level of wild-type Cu/Zn superoxide dismutase. Proc. Natl Acad. Sci. USA 94, 38833887.
  • Pietrobon D. (2005) Function and dysfunction of synaptic calcium channels: insights from mouse models. Curr. Opin. Neurobiol. 15, 257265.
  • Pouyssegur J. and Lenormand P. (2003) Fidelity and spatio-temporal control in MAP kinase (ERKs) signalling. Eur. J. Biochem. 270, 32913299.
  • Pulido R., Zúñiga A. and Ullrich A. (1998) PTP-SL and STEP protein tyrosine phosphatases regulate the activation of the extracellular signal-regulated kinases ERK1 and ERK2 by association through a kinase interaction motif. EMBO J. 17, 73377350.
  • Rintelen F., Hafen E. and Nairz K. (2003) The Drosophila dual-specificity ERK phosphatase DMKP3 cooperates with the ERK tyrosine phosphatase PTP-ER. Development 130, 34793490.
  • Saxena M., Williams S., Tasken K. and Mustelin T. (1999) Crosstalk between cAMP-dependent kinase and MAP kinase through a protein tyrosine phosphatase. Nature Cell Biol. 1, 305311.
  • Sweatt J. D. (2004) Mitogen-activated protein kinases in synaptic plasticity and memory. Curr. Opin. Neurobiol. 14, 311317.
  • Tanoue T. and Nishida E. (2003) Molecular recognitions in the MAP kinase cascades. Cell. Signal. 15, 455462.
  • Tárrega C., Ríos P., Cejudo-Marín R., Blanco-Aparicio C., Van Den Berk L., Schepens J., Hendriks W., Tabernero L. and Pulido R. (2005) ERK2 shows a restrictive and locally selective mechanism of recognition by its tyrosine phosphatase inactivators not shared by its activator MEK1. J. Biol. Chem. 280, 37 88537 894.
  • Thomas G. M. and Huganir R. L. (2004) MAPK cascade signalling and synaptic plasticity. Nat. Rev. Neurosci. 5, 173183.
  • Valjent E., Pascoli V., Svenningsson P., Paul S., Enslen H., Corvol J. C., Stipanovich A., Caboche J., Lombroso P. J., Nairn A. C., Greengard P., Herve D. and Girault J. A. (2005) Regulation of a protein phosphatase cascade allows convergent dopamine and glutamate signals to activate ERK in the striatum. Proc. Natl Acad. Sci. U.S.A. 102, 491496.
  • Van De Warrenburg B. P., Verbeek D. S., Piersma S. J., Hennekam F. A., Pearson P. L., Knoers N. V., Kremer H. P. and Sinke R. J. (2003) Identification of a novel SCA14 mutation in a Dutch autosomal dominant cerebellar ataxia family. Neurology 61, 17601765.
  • Van Den Maagdenberg A. M. J. M., Bächner D., Schepens J. T. G., Peters W., Fransen J. A. M., Wieringa B. and Hendriks W. J. A. J. (1999) The mouse Ptprr gene encodes two protein tyrosine phosphatases, PTP-SL and PTPBR7, that display distinct patterns of expression during neural development. Eur. J. Neurosci. 11, 38323844.
  • Van Der Zee C. E. E. M., Man T. Y., Van Lieshout E. M. M., Van Der Heijden I., Van Bree M. and Hendriks W. J. A. J. (2003) Delayed peripheral nerve regeneration and central nervous system collateral sprouting in leukocyte common antigen-related protein tyrosine phosphatase deficient mice. Eur. J. Neurosci. 17, 9911005.
  • Walter J. T., Alvina K., Womack M. D., Chevez C. and Khodakhah K. (2006) Decreases in the precision of Purkinje cell pacemaking cause cerebellar dysfunction and ataxia. Nat. Neurosci. 9, 389397.
  • Yabe I., Sasaki H., Chen D. H., Raskind W. H., Bird T. D., Yamashita I., Tsuji S., Kikuchi S. and Tashiro K. (2003) Spinocerebellar ataxia type 14 caused by a mutation in protein kinase C gamma. Arch. Neurol. 60, 17491751.
  • Zanke B., Squire J., Griesser H., Henry M., Suzuki H., Patterson B., Minden M. and Mak T. W. (1994) A hematopoietic protein tyrosine phosphatase (HePTP) gene that is amplified and overexpressed in myeloid malignancies maps to chromosome 1q32.1. Leukemia 8, 236244.
  • Zsarnovszky A. and Belcher S. M. (2004) Spatial, temporal, and cellular distribution of the activated extracellular signal regulated kinases 1 and 2 in the developing and mature rat cerebellum. Brain Res. Dev. Brain Res. 150, 199209.
  • Zúñiga A., Torres J., Úbeda J. and Pulido R. (1999) Interaction of mitogen-activated protein kinases with the kinase interaction motif of the tyrosine phosphatase PTP-SL provides substrate specificity and retains ERK2 in the cytoplasm. J. Biol. Chem. 274, 21 90021 907.