Behavioral and neurological analyses of adult mice carrying null and distinct loss-of-receptor function mutations in protein tyrosine phosphatase receptor type Z (PTPRZ)

Protein tyrosine phosphatase receptor type Z (PTPRZ) is preferentially expressed in the central nervous system as two transmembrane receptor isoforms PTPRZ-A/B and one secretory isoform PTPRZ-S. Ptprz-knockout mice lacking the expression of all three isoforms show behavioral, learning, and neurological abnormalities, including increased exploratory activities to novelty, deficits in spatial and contextual learning, and reduced responses to methamphetamine, relative to wild-type mice. To investigate whether PTPRZ isoforms play distinct physiological roles, we herein performed behavioral studies on two knock-in mouse lines: One expresses the catalytically inactive Cys-1930 to Ser (CS) mutants of PTPRZ-A/B, while the other generated in the present study expresses catalytically active mutants of PTPRZ-A/B lacking the negative regulatory PTP-D2 domain and C-terminal PDZ-binding motif (ΔD2) instead of wild-type PTPRZ-A/-B. In contrast to Ptprz-knockout mice, neither increased responses to novelty in the open field nor memory impairments in the inhibitory-avoidance task were observed in Ptprz-CS or Ptprz-ΔD2 mice. However, the effects of methamphetamine on locomotor activity were significantly weaker in Ptprz-KO mice and CS mutant mice than in wild-type mice, but were normal in ΔD2 mutant mice. Furthermore, microdialysis experiments revealed that methamphetamine-evoked dopamine release in the nucleus accumbens was reduced in Ptprz-KO mice and CS mutant mice. These results suggest that the extracellular region of PTPRZ, including the secretory isoform, is crucial for behavioral responses to novelty and the formation of aversive memories, whereas the PTPase activities of PTPRZ receptor isoforms are involved in regulating the dopaminergic system.


Introduction
Signaling from the ligand to PTPRZ receptors is regarded as the forward signal. Secretory PTPRZ-S, also known as phosphacan/6B4 proteoglycan/DSD-1, is a major chondroitin sulfate-proteoglycan in the CNS [24][25][26]. PTPRZ-S is one of the extracellular matrix (ECM) and perineuronal net (PNN) components, serving as a substratum for multiple cell adhesion molecules, including F3/contactin [27,28]. The binding signal from the extracellular region of PTPRZ isoforms to (unknown) receptors on different cells has been regarded as a reverse signal.
These findings suggest that the receptor and secreted isoforms play distinct, but complementary roles in regulating development and functions; however, the specific contributions of individual PTPRZ isoforms have remained unclear. In the present study, we generated and characterized the neurological phenotypes of knock-in mutant mice carrying targeted loss-ofspecific functions or domains of PTPRZ receptors relative to Ptprz-KO mice.

Generation of Ptprz-ΔD2 knock-in mutant mice
We recently generated a targeted knock-in mouse line carrying a Cys to Ser mutation in PTP-D1 (PTPase-inactive CS mutant) [29]. According to the inactive "head-to-toe" dimerization model [2], D2-truncated mutants of PTPRZ-A/B are considered to be constitutively active PTPase. In the present study, we newly generated a Ptprz-ΔD2 knock-in mouse line in which Pro (1998) at the carboxy-terminal end of PTP-D1 was substituted to a termination codon ( Fig 1A). Ptprz-CS mice and Ptprz-ΔD2 mice were backcrossed over ten generations with the C57BL/6J (WT) strain as well as Ptprz-KO mice. Homozygous ΔD2 mice were healthy and exhibited apparently normal growth and reproduction. DNA sequencing of cDNA clones prepared from homozygous Ptprz-ΔD2 mouse brains validated the successful knock-in of the desired point mutation (S1A Fig). Quantitative reverse transcription-PCR (qRT-PCR) analyses revealed that the expression levels of mRNAs for the ΔD2 mutants of PTPRZ-A and -B were significantly lower than those of their normal forms in wild-type mice (Fig 1B, left and center; wild-type > heterozygous > homozygous), whereas the level of Ptprz-S mRNA was not significantly different (Fig 1B, right).
Western blotting showed that the expression amount of the ΔD2 mutant form of PTPRZ-B was lower than that of wild-type PTPRZ-B in brain extracts (Fig 1C), consistent with the results of qRT-PCR. In the brain, large amounts of PTPRZ proteins are proteolytically processed by metalloproteinase [17] or plasmin [30]. The core protein of PTPRZ-A at 380 kDa was only negligibly detected because of proteolytic cleavage into the whole extracellular fragment, Z A -ECF [17,30]. PTPRZ-A/B receptor isoforms undergo metalloproteinase-or plasmin-mediated cleavage, releasing various extracellular fragments in the normal brain [17,30]. Since Z A -ECF is indistinguishable from PTPRZ-S based on their mobilities in SDS gel electrophoresis, it was not possible to quantify its amount at the protein level. Consistent with the reduction of the receptor isoforms with the ΔD2 mutation, the amount of its whole extracellular fragment, Z B -ECF at 180 kDa, was also reduced by the ΔD2 mutation. No apparent changes were observed in Z-ECF-100, -90, or -70, which corresponded to the fragments of the extracellular portion of PTPRZ-A/-B produced by plasmin cleavage. To discriminate two close bands of PTPRZ-B and PTPRZ-B-ΔD2, we also analyzed the blot with another antibody (anti-RPTPß), which recognizes an epitope on the C-terminal part of the PTP-D2 domain of PTPRZ-A/-B receptor isoforms [17], but does not detect the ΔD2 mutant. The Western blotting with anti-RPTPß indicated no protein expression of the native full length form of PTPRZ-B products in the homozygous, and verified that the band of a slightly lower size than  PTPRZ-B corresponds to the ΔD2 mutant form (Fig 1D). Here, we should note that the ΔD2 mutant is an constitutively active PTPase in living cells [31].

Ptprz-ΔD2 knock-in mice
Adult (4 to 6 months old) male mice were used in subsequent experiments. In the open field test, Ptprz-KO mice showed significantly stronger horizontal activity than wild-type mice on day 1, but not thereafter (Fig 2), indicating an enhanced response to a novel environment, as reported previously [32]. However, Ptprz-CS and Ptprz-ΔD2 mice showed similar responses to wild-type mice in the open-field test under the same conditions. In the novel object exploratory test, exploratory activity to a novel object was greater by Ptprz-KO mice than by wildtype mice on day 1 (Fig 3), whereas Ptprz-CS and Ptprz-ΔD2 mice showed no alterations, indicating that this activity is independent of PTPase activity.
We then performed a step-through inhibitory avoidance task that depends on the functions of the hippocampus and amygdala [33,34]. Mice placed in the bright chamber exhibit behavior to immediately escape to the dark compartment. Escape latency did not significantly differ among the genotypes ("Pre" in Fig 4). Mice then received footshocks in the dark compartment, and were left there for 2 min to learn the footshock-context association. This learning session was repeated until each mouse showed a predetermined latency of 3 min. The number of repetitions required to reach the predetermined time did not significantly differ among the genotypes (Mean ± standard deviation; WT = 3.1 ± 0.99, KO = 2.8 ± 0.79; CS = 2.8 ± 1.03, ΔD2 = 3.0 ± 0.67; UNIANOVA: F (3,36) = 0.288, P = 0.834), indicating the normal acquisition of aversive conditioning in all genotypes (groups). Memory retention was then tested 24 h later. Only Ptprz-KO mice showed a shorter latency than wild-type, CS, and ΔD2 knock-in mice ("24 hr" in Fig 4). These results indicated that the catalytic activity of PTPRZ was not essential for a normal novelty response or aversive learning.

Reduced responses to METH in CS mutant mice
Methamphetamine (METH) is a drug of abuse that induces an increase in locomotor activity in rodents, and repeated METH exposure causes adaptive changes in central dopaminergic systems, which may underlie the mechanism of locomotor sensitization to METH [35][36][37]. We previously reported that Ptprz-KO mice showed significantly weaker locomotor responses to a single and repeated injections of METH at a dose of 1 mg/kg than wild-type mice (whereas the locomotor response to 3 mg/kg of METH did not significantly differ) [32]. Under the same experimental conditions, Ptprz-CS mice and Ptprz-KO mice similarly showed weaker responses to METH than wild-type mice ( Fig 5A). Furthermore, METH-induced locomotor sensitization was significantly reduced in Ptprz-CS and Ptprz-KO mice. However, Ptprz-ΔD2 Filled triangles show loxP sites. (B) Quantitative RT-PCR. The mRNA expression levels of Ptprz-A and Ptprz-B (total of both wild-type and ΔD2 mutant forms) and Ptprz-S in adult brain tissues were measured using the respective primer sets shown in S1 Fig. They were normalized to Gapdh expression, and plotted as relative values to wt/wt mice (n = 3 individual mice per group). wt/wt, homozygous for the wild-type allele; wt/ΔD2, heterozygous; and ΔD2/ΔD2, homozygous for the ΔD2 mutant allele. There were significant effects of genotype on the expression levels of Ptprz-A (F (2, 6) = 55.663, P = 0.000) and Ptprz-B (F (2, 6) = 37.680, P = 0.000), but not Ptprz-S (F (2, 6) = 0.006, P = 0.874) by a univariate analysis of variance (UNIANOVA). �� , P < 0.01, significant difference between the indicated groups by Tukey's post-hoc HSD test. (C, D) Western blot analyses of brain extracts treated with (+) or without (-) chondroitinase ABC (chABC). In brain tissues, PTPRZ-A, -B, and -S and their processed derivatives (Z A -ECF and Z B -ECF) were highly modified with chondroitin sulfate chains, and, therefore, the chABC treatment beforehand was necessary for resolving their core proteins by SDS-PAGE [17,30]. Anti-PTPRZ-S (C) and anti-RPTPß (D) recognize the extracellular region of all three PTPRZ isoforms, and the epitope on the C-terminal part of the PTP-D2 domain, respectively [17]. (E) CBB staining of samples to verify their protein amounts applied to each lane.
https://doi.org/10.1371/journal.pone.0217880.g001 . Two values that were obtained from one CS-knock-mouse were regarded as outliners by Grubb's test (pointed by arrowheads). Therefore, the results from the CS knock-in mouse were eliminated for the reliable verification, and the trimmed data set was shown as a bar graph with the mean and stander error (SE) mice showed no alterations ( Fig 5B). These results indicated that these reductions in response are due to PTPase activity.
METH-induced hyper-locomotion is caused by increases in extracellular dopamine levels in the terminal field of the mesolimbic dopamine pathways, particularly in the nucleus accumbens [35]. Ptprz-KO mice showed a reduction in METH-induced dopamine efflux in the nucleus accumbens [32]. Microdialysis measurements revealed that METH-evoked dopamine release in the nucleus accumbens was also lower in Ptprz-CS mice than in wild-type mice ( Fig  6). METH-induced dopamine release was slightly lower in Ptprz-ΔD2 mice than in wild-type mice (p = 0.063 vs wild-type mice by Tukey's post-hoc test), but was significantly higher than that in Ptprz-CS mice. These results indicated that the loss of PTPase activity is associated with reduced responses to METH.

Discussion
The present results revealed that Ptprz-CS mice and Ptprz-KO mice showed attenuated METH-induced locomotor activations with reductions in dopamine efflux in the nucleus accumbens, whereas increased responses to novelty and impaired memory retention in the aversive learning task were only observed in Ptprz-KO mice. To the best of our knowledge, the present study is the first to suggest that PTPRZ isoforms may be involved in distinct physiological functions in the brain; therefore, we considered that the enhanced response to low METH responsiveness should be due to the loss of the forward signal by the PTPase activity of PTPRZ receptors, whereas novelty and impaired memory retention appeared to be due to the loss of the reverse signal by the extracellular region, including the secretory isoform, PTPRZ-S (see Fig 7).
Ptprz-KO mice exhibit maturation-dependently enhanced LTP in the CA1 region in hippocampal slices and impaired hippocampal-dependent learning [11,14]. This phenotype is slightly inconsistent with the finding that mice lacking PTN, an inhibitory ligand of PTPRZ receptors, exhibit enhanced hippocampal LTP with learning impairments [38]. However, CS and ΔD2 mutant mice both displayed normal memory in the aversive learning task, indicating that the loss of PTPRZ activity or interactions with synaptic PDZ domain-containing proteins do not confer the hippocampal phenotypes found in Ptprz-KO mice. PTN may modulate hippocampal synaptic plasticity through receptors other than PTPRZ, including syndecan3 (SDC3), a transmembrane heparan sulfate proteoglycan predominantly expressed in the hippocampus of adults. Sdc3-knockout mice exhibit impaired performance in tasks assessing hippocampal functioning together with enhanced LTP in the hippocampal CA1 region [39]. Enhanced LTP in Sdc3-knockout mice is not responsive to PTN, which inhibits LTP in wildtype animals [39].
Normal aversive learning in CS and ΔD2 knock-in mutant mice may emphasize the crucial role of secretory PTPRZ-S in the regulation of hippocampal function. PTPRZ-S acts as a ligand for F3/contactin1 (CNTN1) [27,28]. Whole-cell recordings from the CA1 pyramidal cells of Cntn1-knockout mice at P15 to P17 revealed their impaired paired-pulse facilitation (PPF) and long-term depression (LTD) with normal LTP and synaptic morphology [40]. Notably, in Cntn1 knockout mice, the abnormal distribution of anti-phosphacan (3F8) immunoreactivity was detected in synapse-rich areas, the stratum radiatum and stratum lacunosum moleculare, in the hippocampus [40]; however, the effects of the knockout of Cntn1 on hippocampal (E). There were significant effects of day (F (2, 68) = 30.665, P = 0.000), and interactions of between day and genotype (F (6, 68) = 2.363, P = 0.049), but no significant effects of genotype (F (3, 34) = 1.889, P = 0.170) by two-way mixed design ANOVA. �� , P < 0.01, significantly different between the indicated groups by posthoc Tukey HSD test.
https://doi.org/10.1371/journal.pone.0217880.g002 learning ability in adult mice has not been examined because of early postnatal lethality (by P18) [41]. Transgenic mice overexpressing CNTN1 are functionally normal at 5 months of age, but show enhanced LTP in the CA1 and improved spatial and object recognition memory at 12 months of age [42], which appears to be consistent with the maturation-dependent (>13 weeks old) hippocampal alteration in Ptprz-KO mice [14]. PTPRZ-S may function as a ligand of the CNTN1 receptor (or receptor complex) in the hippocampus. Future studies on mice selectively lacking the Ptprz-S isoform are needed to clarify this point and elucidate the signaling pathways involved in the regulation of hippocampal functions as the reverse signal.
PTPRZ receptor isoforms are expressed in the majority of midbrain dopamine neurons, and Ptprz-KO mice showed reduced responses to METH [32]. The present study revealed that CS knock-in mice also showed weaker locomotor responses and lower mesolimbic dopamine release to METH. Furthermore, no significant changes were observed in the expression level of tyrosine hydroxylase (TH), the rate-limiting enzyme for DA biosynthesis, in striatal extracts in the mouse groups (S2 Fig). These results strongly suggest that the PTPase activities of PTPRZ receptors positively affects the stimulant effects of METH as the forward signal. Mice deficient in RET tyrosine kinase show reduced dopamine transporter (DAT) activity and diminished behavioral responses when exposed to cocaine [43], suggesting that the loss of PTPs that counteract RET kinase activity reduces METH responsiveness. It is important to note that mice carrying a heterozygous knockout mutation in the Ptprd gene encoding PTPRD take longer to establish the self-administration of cocaine than wild-type littermate controls [44]. The same study also revealed that 7-butoxy illudalic acid analog (7-BIA), which inhibited the phosphatase activity of PTPRD in vitro, reduced cocaine rewards in self-administration and conditioned place-preference tests [44]. Although the downstream substrate proteins for PTPRD involved in cocaine rewards or their relevance to the inhibition potency of 7-BIA for PTPRD activity in vivo remain unclear, these findings suggest the loss of PTPs counteracting GDNF-RET signaling is also connected to reductions in METH responsiveness. It will be of interest to elucidate the relationship and crosstalk between RPTPs, including PTPRZ, and RPTKs, such as RET.
The inhibitory PTPRZ ligand, PTN is up-regulated after a single administration of amphetamine in the nucleus accumbens [45], and it (a relatively high dose of 10 mg/kg) induces conditioned place preference in Ptn-KO and wild-type mice at similar levels, whereas, 5 days after the last amphetamine injection, Ptn-KO mice, but not wild-type mice, still show a preference for the amphetamine-paired side [46]. However, this study identified enhanced amphetamineinduced astrocytosis in the striatum of Ptn-KO mice [46], suggesting that PTN limits amphetamine-induced reward effects by ameliorating neurotoxic effects. Besides endogenous PTPRZ ligands, it was recently reported that mice treated with MY10, a small-molecule inhibitor of PTPRZ, drank less ethanol than controls, and the MY10 treatment blocked ethanol-conditioned place preference; however, the effect, efficacy, and specificity of MY10 on PTPRZ phosphatase activity have not yet been demonstrated [47]. Ligand (PTN)-induced PTPase inactivation through "head-to-toe" dimerization [48] is expected to be abolished in ΔD2 mutant mice. Future studies using these mutant mice will be valuable for clarifying the

Ethics statement and experimental animals
All experimental animal protocols used in the present study were approved by the Institutional Animal Care and Use Committee of the National Institutes of Natural Sciences (approval numbers: 12A078, 13A172, 14A150, 15A095, 16A147, and 17A023), and the RIKEN Kobe Branch (approval number: A2001-03-72), Japan.
Ptprz-KO mice [4] and Ptprz-CS mice (ref) were backcrossed with the inbred C57BL/6J strain (CLEA Japan) for more than ten generations. Mice were housed under specific pathogen-free (SPF) conditions at a constant room temperature (23˚C) and 50-55% humidity with a 8:00 to 20:00 light cycle. Four to 5 weeks after birth, 3 to 4 sex-matched mice were housed in a plastic cage (cage size: 12 × 21 × 12.5 cm) with paper-chip bedding, and food and water were provided ad libitum.
Adult male mice (4 to 6 months old) were used in the present study. Mice were handled gently to minimize stress. Surgeries for implanting a guide cannula for brain microdialysis were performed under isoflurane anesthesia, 2% lidocaine cream was applied to the incision site after surgery for acute pain relief, and all efforts were made to minimize suffering. Behavioral experiments were performed during the light period (between 9:00 and 17:00) by at least two different individuals under blind conditions. The method of sacrifice used for the experimental mice is decapitation with surgical scissors.
https://doi.org/10.1371/journal.pone.0217880.g004 Bio), and used as a template for real-time PCR using a commercial kit (TaKaRa One Step SYBR, Takara Bio) on a real-time PCR system (StepOnePlus Real-Time PCR System). The relative mRNA expression levels of the Ptprz-A, -B, and -S isoforms and ΔD2 mutant isoforms were estimated and normalized to that of glyceraldehyde-3-phosphate dehydrogenase (Gapdh). The PCR primers used and sizes of the amplified products are shown in S1 Fig.

Open field and novel object exploration tests
Mice were placed in the center circle of a round field (internal diameter, 75 cm and wall height, 40 cm) divided by a grid into twenty-five equal segments under normal lighting conditions (approximately 240 lux at the center area of the open field), and their behaviors were observed for 5 min per day over 3 consecutive days. Horizontal locomotor activity was examined by counting the number of crossings over boundary lines. On the day after the open-field test, the novel object test was started in the same open field. Mice stayed in the open field without objects for 9 min, a white plastic cube (5×5×5 cm) was then placed in the central area, and after the METH injection with the mean of sensor counts with SE on days 1 to 15 (n = 13 per group). There were significant effects of genotype (F (3,48) = 7.151, P = 0.000) by a two-way mixed design ANOVA. � , P < 0.05; �� , P < 0.01, significant difference between the indicated groups by the Bonferroni post-hoc test.
https://doi.org/10.1371/journal.pone.0217880.g005 their exploratory behavior (the number of crossings over the center circle) was recorded for a further 9 min.

Inhibitory avoidance test
Training and testing were performed in an apparatus with two compartments (15×15×15 cm each): a transparent plexiglas box and a black plexiglas box, separated by a black-colored guillotine door, which was set in a behavioral analyzing system (SCANET, Melquest). The floor was a stainless-steel rod grid comprised of twenty bars with a diameter of 2 mm, which were electrified with a shock generator scrambler (NS-SG01, Neuroscience). The conditioning apparatus was located in a separate room with low levels of background noise produced by ventilation. Illumination was provided by a 36 W light bulb 30 cm above the apparatus.
On the day before the test, mice were allowed to move freely between the two compartments for 30 min for habituation to the conditioning apparatus. Mice were initially placed in the light compartment and this was followed by opening the door. When mice entered the dark compartment, the door was immediately closed, a 1-sec, 0.5-mA footshock was applied after 10 sec, and mice were left there for 2 min. This sequence was repeated until each mouse remained in the light compartment for more than 3 min (cut-off time) when newly placed there. The escape latency of the first trial before receiving the footshock was recorded as the pre-conditioning value. After 24 hrs, the retention of the aversive memory was assessed by measuring escape latency from the light to dark compartment (two trials per session).

METH-induced locomotor response
METH (Methamphetamine hydrochloride, Dainippon Pharmaceutical) was dissolved to 0.1 mg/ml (for 1 mg per kg per injection of METH) with sterilized 0.9% NaCl (normal saline, Otsuka Pharmaceutical) and injected subcutaneously into the backs of mice. Locomotor The phenotype is evident (Up/increase or down/decrease with arrows) or not (horizontal arrows). Ligand (PTN) sensitivity means the ligand-induced PTPase inactivation of PTPRZ receptors. Distinct physiological roles of PTPRZ isoforms might be a reason to explain the discrepancy in the myelination phenotype between our Ptprz-KO mice [7,8,29] and another KO mouse line [13], which is mentioned in the introduction section. Because the latter KO line was generated by the replacement of an exon encoding a portion of the extracellular CAH domain with a pgk-neo cassette [5], which may result in unexpected expression of an aberrant extracellular fragment of PTPRZ.

In vivo microdialysis
Mice were anesthetized with 2% vaporized isoflurane during surgery. An intracerebral guide cannula (CMA 11, CMA/Microdialysis AB) was then stereotaxically implanted on the left nucleus accumbens at a depth of 4.0 mm (coordinates with respect to the bregma: 1.4 mm anterior and 0.8 mm lateral). The guide cannula was secured with dental cement (GC Fuji I, GC Corporation) and then closed with a dummy cannula. Lidocaine cream was applied to the incision site after surgery for acute pain relief. After allowing at least 3 days of recovery from surgery, the dummy cannula was removed and a dialysis probe (membrane length of 1 mm, 6 kDa cut-off, CUP 11, CMA/Microdialysis AB) was set through the guide cannula. The probe was perfused with Ringer solution (147 mM Na + , 4 mM K + , and 155.6 mM Ca 2+ ) at a flow rate of 2.0 μl/min. Microdialysis was performed in awake and unrestrained mice, and the amounts of dopamine in dialysates were measured using online HPLC coupled to an electrochemical detector system (BMA-300, Eicom) according to the manufacturer's instructions.

Statistical analyses
Statistical analyses were performed using IBM SPSS Statistics 25 software (SPSS) together with Microsoft EXCEL (Excel for Mac version 16.16.2, Microsoft). Since ANOVA assumes the homogeneity of variance across all conditions, Mauchly's test of Sphericity was adopted to examine the sphericity assumption within repeatedly measured data; when this test was significant, a Greenhouse-Geisser correction was applied. In Figs 2 and 4, statistical outliers were determined using GraphPad's QuickCalc Grubb's test (http://graphpad.com/quickcalcs/ grubbs1) with significance set to α = 0.05, and the trimmed data sets were then analyzed by two-way mixed design ANOVA.

S2 Fig. Expression levels of TH proteins.
Striatal tissue extracts were analyzed by Western blotting using anti-TH antibody (AB152, Merck). The scatter plot shows the signal intensity of TH staining relative to that from the wild-type mice, in which each circle corresponds to an independent experiment (n = 6 per group). There were no significant effects of genotype (F (3,20) = 0.703, P = 0.561) by UNIANOVA. (TIF)