The cell adhesion molecule neuroplastin (Np) is a novel candidate to influence human intelligence. Np-deficient mice display complex cognitive deficits and reduced levels of Plasma Membrane Ca2+ ATPases (PMCAs), an essential regulator of the intracellular Ca2+ concentration ([iCa2+]) and neuronal activity. We show abundant expression and conserved cellular and molecular features of Np in glutamatergic neurons in human hippocampal-cortical pathways as characterized for the rodent brain. In Nptn lox/loxEmx1Cre mice, glutamatergic neuron-selective Np ablation resulted in behavioral deficits indicating hippocampal, striatal, and sensorimotor dysfunction paralleled by highly altered activities in hippocampal CA1 area, sensorimotor cortex layers I-III/IV, and the striatal sensorimotor domain detected by single-photon emission computed tomography. Altered hippocampal and cortical activities correlated with reduction of distinct PMCA paralogs in Nptn lox/loxEmx1Cre mice and increased [iCa2+] in cultured mutant neurons. Human and rodent Np enhanced the post-transcriptional expression of and co-localized with PMCA paralogs in the plasma membrane of transfected cells. Our results indicate Np as essential for PMCA expression in glutamatergic neurons allowing proper [iCa2+] regulation and normal circuit activity. Neuron-type-specific Np ablation empowers the investigation of circuit-coded learning and memory and identification of causal mechanisms leading to cognitive deterioration.
Learning and memory depend on balanced excitation and inhibition in neuronal networks interconnected via synapses1,2,3,4,5. Although the underlying molecular mechanisms are not fully understood, it is clear that cell adhesion molecules (CAMs) regulate decisive mechanisms to fine-tune neuronal communication6,7,8. The CAM neuroplastin9 (Np) regulates synaptic plasticity i.e. long-term potentiation10,11,12 and formation/stabilization of excitatory synapses13, 14 and balances the ratio of excitatory/inhibitory synapses13.
Among 50,000 gene variants analyzed, a promoter polymorphism in the human Np gene (NPTN) correlated with cortical thickness and cognitive capabilities in adolescents15. Abnormally increased Np expression is linked to schizophrenia16. Recently, new mouse models for the study of Np functions have been reported. Mice with loss of Np65 expression (one of the two Np isoforms, Np55 and Np65) display altered spatial performance and anxiety-like behaviors17. Other groups have reported that mice with polymorphic or truncated Np gene variants are deaf14, 18. In contrast Nptn −/− mice, completely devoid of Np expression, are not deaf but display loss of fear-conditioned associative learning, altered sensorimotor capabilities, complex swimming and diving behaviors as well as further deficits related to autism, depression, and schizophrenia10. Moreover, neuron-specific Nptn gene inactivation in adult Nptn lox/loxPrpCreERT mice triggered hippocampal and cortical alterations and induced retrograde amnesia of associative memories assigning a unique role to Np in cognitive functions10.
Although the distribution of neural CAMs such as N-CAM or L1 are similar in human, rat, and mouse hippocampus19, differences in hippocampal Np distribution between these species were reported20. Detailed expression, abundance, subcellular distribution or synaptic localization of the Np isoforms in the human brain have not been clarified. Thus, important information necessary to relate behavioral deficiencies in Np mouse models to specific human brain functions is missing.
In Nptn −/− and Nptn lox/loxPrpCreERT mice, PMCA expression levels are severely reduced10. These high-affinity Ca2+-pumps, but no other mechanisms, are capable to fully restore basal [iCa2+] in all eukaryotic cells21, 22. In neurons, PMCAs are key players in iCa2+-homeostasis essential for proper neuronal activity. Alternative splicing confers diversity in expression and localization to four PMCA paralogs (PMCA1–4)23,24,25,26,27,28. PMCA1 and 4 display ubiquitous expression in the mammalian brain with high levels in the hippocampal CA1 area and neocortical and piriform neurons24,25,26. PMCA2 is expressed in the hippocampus and specific PMCA2 isoforms are restricted to either GABAergic presynapses27 or glutamatergic spines28. PMCA3 is expressed in cortical regions with lower levels in the hippocampus23. Information on the specific contribution of PMCA paralogs to normal brain capabilities is scarce. Furthermore, despite alterations in the expression of PMCA paralogs and their association to impaired [iCa2+] handling in autism29, 30, schizophrenia31, Alzheimer’s disease32, 33, Niemann-Pick disease34, deafness35, and spinocerebellar ataxia36, 37, mechanisms leading to altered levels of PMCA paralogs remain largely unsolved.
Here we show that, similarly as reported for rodent, essential cellular and molecular features of Np are conserved in the human brain. Furthermore, human Np is also mainly expressed in glutamatergic neurons of hippocampal-cortical circuits. In mice, Np ablation in glutamatergic neurons impaired specific behaviors and activities in CA1 hippocampal area and in the cortex-striatum sensorimotor circuit. The differential loss of selective PMCA paralogs as direct consequence of Np ablation suggests the Np-PMCA interplay as a novel mechanism associated with brain area-specific cognitive functions.
Recently, Np has received much interest because of its correlation with cortical thickness and intelligence in adolescents15, 38. As Np expression has not been sufficiently characterized in human brain, we investigated the expression and molecular characteristics of human Np (hNp) further. First, preservation of hNp immunoreactivity, specificity of pan-Np55/65 and anti-Np65 antibodies and reliability of our immunohistochemistry procedures were demonstrated with hippocampal slices from 4 or 8 days post-mortem wild-type and Nptn −/− mice prepared under similar conditions as the human samples (Supplementary Fig. S1). Then, using Nissl staining, DAB-based detection and large-scale and high-resolution scanning (Fig. 1a), we analyzed hNp65 isoform immunoreactivity within the main areas/ layers of human hippocampal formation. hNp65 was mainly segregated to all major neuron-containing layers of the hippocampus –stratum granulosum and stratum moleculare of the dentate gyrus, the hilar part of the CA3 region, the pyramidal cell layer of CA1 and CA2-3 and the cellular layers of subiculum (Fig. 1a) as demonstrated by semi-automated analysis (Fig. 1b,c). Prominent hNp65 immunoreactivity is detected in densely packed apical dendrites of granular neurons and in stratum moleculare of DG. Presence of hNp65 is low in stratum plexiforme and within the hilar neuropil of DG, and we did not detect hNp65 in mossy fibers traversing the polymorphic layer of the DG (Fig. 1a,d). Also, it seems probable that some interneurons in the molecular layer of DG are hNp-positive as reported20. hNp65 expression was low and scattered throughout the hilar part of the CA3 region, but strong in cell bodies of glutamatergic pyramidal neurons and dendrites of stratum radiatum in CA1 while moderate in strata pyramidale of CA2-3 and subiculum (Fig. 1a–e). At the cellular level, hNp65 is very abundant at membranes of the cell body of granular and pyramidal neurons (Fig. 1d,e), dendrites, and in punctate structures within the neuropil (Fig. 1d,e). In the hilus, many hNp65-positive cells (Fig. 1e) largely resemble big multipolar calretinin-positive interneurons described previously39. In the entorhinal cortex, a prominent hNp65 localization is observed in glutamatergic pyramidal neurons in the layers II, IV and V (Fig. 1a,e) strikingly similar to the rat Np65 (rNp65) immunoreactivity12. In conclusion, unequivocally identified hNp65-positive glutamatergic neurons are granular neurons of DG, pyramidal neurons of CA1, CA2–3, subiculum, and layers II, IV, and V of the entorhinal cortex. Further direct visual inspection using a bright-field microscope confirmed that the hNp65 expression in major areas/ layers of the glutamatergic pathways within the entorhinal cortex and hippocampus (Supplementary Table 1) is very similar to its reported expression in mouse and rat9, 12, 13, 20, 40.
The NPTN gene is highly expressed in human striatum (see Supplementary Fig. S2 for Np protein expression), cortex and hippocampus41 (http://hbatlas.org/hbtd/images/wholeBrain/NPTN.pdf). Correlation analysis shows that hNp55/65 and hNp65 transcripts are tightly co-expressed in the prefrontal cortex of 0–80 year-old subjects (Fig. 1f). The highest hNp mRNA level occurred in 10–20 year-old subjects to decline afterwards with aging (P = 5.8E-4, two samples t tests, Fig. 1g). On the protein level, as described for rodent brain40, we detected two protein backbones of 28 and 40 KDa with a pan-Np55/65 antibody and only one of 40 KDa using an Np65-specific antibody in human brain homogenates after deglycosylation (Fig. 1h). Also, Np55/65- and Np65-associated immunoreactivity was co-localized and substantially similar in plasma membranes of human and mouse glutamatergic hippocampal neurons as observed using confocal microscopy (Supplementary Fig. S3). The hNp55/65 content in synaptic membranes, synaptic junctions, and postsynaptic densities (PSDs) was higher than in other subcellular fractions from human cortex (Fig. 1i) indicating relative enrichment of the human isoforms in glutamatergic synapses.
As human (Fig. 1 and Supplementary Fig. S1) and mouse9, 12, 13, 20, 40 Np55/65 share common molecular features and are expressed abundantly by hippocampal and cortical glutamatergic neurons, we generated Nptn lox/loxEmx1Cre mice to test whether Np deletion in these neuronal types affects selective behaviors. Importantly, hippocampal-cortical Np expression was eliminated specifically in glutamatergic but not in GABAergic neurons (Supplementary Fig. S4). Nptn lox/loxEmx1Cre mice displayed normal basic movement abilities and grip strength (Supplementary Fig. S5).
In the Morris water maze, Nptn −/− mice show an aberrant diving behavior10 and pause swimming very frequently (data not shown) compromising the evaluation of spatial memory and strategies. Notably, Nptn lox/loxEmx1Cre mice did not dive but displayed frequent pausing during swimming as evidenced by the appearance of “knots” in the track recordings (Fig. 2a). Consistently, the average speed of Nptn lox/loxEmx1Cre mice was lower (Fig. 2b) (day 1, D1: F(1,21) = 12.413, P = 0.002; D2: F(1,21) = 16.683, P = 0.0005; D3: F(1,21) = 8.674, P = 0.0077; D4: F(1,21) = 7.181, P = 0.014; D5: F(1,21) = 5.885, P = 0.0244) caused by longer lasting immobility (Fig. 2c) (D1: F(1,21) = 4.968, P = 0.0369; D2: F(1,21) = 8.783 P = 0.0074) resulting in longer escape latencies during acquisition (D2: F(1,21) = 6.645, P = 0.0175; Fig. 2d) compared to control mice. When the escape latencies were corrected for times of immobility (Fig. 2e), Nptn lox/loxEmx1Cre and Nptn lox/lox mice showed the typical reduction in escape latency during acquisition at days 1–3 (Fig. 2e). In agreement, Nptn lox/loxEmx1Cre and Nptn lox/lox control mice showed similar reduction of the pathlength during acquisition (D1-D3; Fig. 2f). Furthermore, Nptn lox/loxEmx1Cre and Nptn lox/lox mice develop spatial memory as revealed by similar preference for the quadrant where the platform was located during acquisition (old goal) vs. the quadrant with the new platform position (new goal) or other quadrants (QA, QB) during the first two trials after platform reversal (Fig. 2g). However, reversal of the platform at D4-D5 expectedly prolonged only in Nptn lox/lox mice escape latency (Fig. 2e) and pathlength (Fig. 2f) (D4: F(1,21) = 11.216, P = 0.003; D5: F(1,21) = 6.456, P = 0.019), whereas Nptn lox/loxEmx1Cre mice were not perturbed at all. This lack of delay by platform reversal may indicate alternative non-spatial navigation by Nptn lox/loxEmx1Cre mice. Supportingly, the track recordings for day 4 (Fig. 2h) immediately after platform reversal indicate for Nptn lox/lox an initial persistence to the old platform position (trial 1) that changes into a more spatial strategy with further training for the new position (trial 6). In contrast, Nptn lox/loxEmx1Cre display after reversal a circular non-spatial strategy swimming equidistant to the wall indicating a hippocampal dysfunction related to building an allocentric representation of their environment required for a coordinated navigation. In conclusion, Np ablation in glutamatergic neurons did not interfere with formation of spatial memory but affected (a) the continuity of task execution with repetitive pause-swimming cycles indicative for altered striatum-dependent decision making and (b) the swimming strategy after platform reversal suggesting hippocampal deficits42.
Nptn −/− and induced Nptn lox/loxPrpCreERT mice are unable to learn associative fear conditioned tasks10. Here, we investigated whether Np deletion only in glutamatergic neurons is sufficient to impair this associative capacity. Using the two-way active avoidance shuttle-box paradigm, Nptn lox/loxEmx1Cre mice performed similar to Nptn lox/lox controls during training from D1 to D4, but slightly better at D5 (Fig. 2i) (F(1,21) = 4.748, P = 0.0409), indicating that Np expression in glutamatergic neurons is not essential for this associative learning task. Using the rota-rod test (Fig. 2j), Nptn lox/loxEmx1Cre mice showed significantly better sensorimotor-coordination at 16 rpm (F(1,21) = 5.149, P = 0.0339) and 24 rpm (F(1,21) = 4.343, P = 0.0496) compared to Nptn lox/lox and consistent with the improved performance of Nptn lox/loxPrpCreERT mice10. Thus, Np ablation in glutamatergic neurons is sufficient to produce behavioral alterations related to cerebellar and/or motor cortex function.
The observed behavioral alterations may result from abnormal activities in the respective Nptn lox/loxEmx1Cre brain structures. Therefore, we analyzed the patterns of regional cerebral blood flow (rCBF) in awake mice as reported by functional99mTcHMPAO-SPECT-imaging10, 43 (Fig. 3). Although whole-brain topographies were performed, we identified specific differences with decreased 99mTc-contents in layers I-III/ IV in both left and right hemispheres of the frontal and middle sensorimotor cortex and in the ventral part of the left hippocampal CA1 region of Nptn lox/loxEmx1Cre compared to Nptn lox/lox control mice. Interestingly, the largest and highly significant decreases in 99mTc-content occurred in the left somatosensory and motor cortex I and II which are contrasted by increased99mTc-content in anterior and lateral hypothalamic and striatal regions in Nptn loxloxEmx1Cre mice. Notably, the highest increase of 99mTc-content in the striatum occurred within its lateral sensorimotor domain a known target of glutamatergic projection neurons from the sensorimotor cortex44. Cerebellar activity was not different between Nptn lox/loxEmx1Cre and Nptn lox/lox control mice (data not shown).
Previously, we observed that Np-deficiency strongly reduced total PMCA expression in brain extracts and was associated with altered brain activity10 linking alterations in neuronal activity to [iCa2+] and Ca2+-signaling in Np animals. We further characterized PMCA expression levels by immunofluorescence staining and confocal microscopy and observed homogeneously reduced PMCA immunoreactivity in hippocampus (not shown) and cortex of Nptn −/− mice (Fig. 4a). Indeed, fluorescence intensity measurements revealed a reduction from 100 ± 15.8% in wild-type to 42.2 ± 19.1% in Nptn −/− slices (P < 0.05, six animals per genotype, unpaired two-tailed t-test). Furthermore, using an additional paralog-specific PMCA2 antibody and Western blot analysis, decreased protein levels of this neuron-specific pump in Nptn −/− brain homogenates were confirmed (Fig. 4b,c). Importantly, PMCA1, 2, and 4 mRNA transcript levels remained unchanged (Fig. 4d) indicating normal transcription of PMCA paralogous genes in the complete absence of Np in Nptn −/− mice.
Guided by the SPECT-imaging results (Fig. 3), we analyzed Np and PMCA immunoreactivities throughout the sensorimotor cortex of Nptn lox/loxEmx1Cre mice using confocal microscopy (Fig. 4e). PMCA immunoreactivity decreased in all cortical layers with the largest and significant decrease in layers II/III and IV/V (Fig. 4f, P < 0.05, four animals per genotype, three sections each, unpaired two-tailed t-test), whereas in interneurons of Nptn lox/loxEmx1Cre mice Np and PMCA immunoreactivities were preserved (Supplementary Figs S4 and S6) as expected. Protein levels of each PMCA paralog were assessed in homogenates of Nptn lox/loxEmx1Cre cortex and hippocampus using paralog-specific antibodies and Western blotting (Fig. 4g). Densitometric analysis confirmed severely reduced Np55/65 and pan-PMCA immunoreactivities in Nptn lox/loxEmx1Cre cortical and hippocampal homogenates. Compared to control Nptn lox/lox, PMCA1, 3, and 4 were clearly reduced in cortex of Nptn lox/loxEmx1Cre mice (Fig. 4g,h). In the Nptn lox/loxEmx1Cre hippocampus, PMCA1, 2, and 3 were only slightly reduced, while PMCA4 was strongly reduced (Fig. 4g,h). Additionally, distinct reduction of PMCA1 immunoreactivity in cortex and hippocampus was confirmed by confocal microscopy (Supplementary Fig. S6b).
To gain insight into the molecular mechanisms underlying the Np-PMCA association, we tested in cultured cells the hypothesis that Np modulates PMCA levels directly. In HEK cells, a human cell line not expressing Nps nor PMCA240, 45, all three GFP-tagged versions of Nps, namely rNp55, rNp65, and hNp65 increased endogenous hPMCA protein levels as evaluated by Western blot in total extracts (Fig. 5a upper panel) and confirmed by quantification of the optical density of the bands (Fig. 5a lower graph). rNp65-GFP was effective to increase consistently the basal over-expression of exogenous hPMCA2 (Fig. 5b) in transiently transfected HEK cells. Furthermore, the expression and plasma membrane localization of exogenous hPMCA1-GFP and hPMCA2-GFP were promoted robustly by rNp55-RFP as evaluated by confocal microscopy (Fig. 5c). Immature rat hippocampal neurons cultured for 5 days express low levels of endogenous Np65 and PMCA223, 46. As in HEK cells, hNp65-GFP cDNA constructs promoted the expression of exogenous hPMCA2 (Fig. 5d) in co-transfected immature neurons with large and pyramidal soma. Therefore, both rat and human Np can promote PMCA protein levels in both human and rat cultured cells.
Because PMCA is the only high affinity extrusion mechanism able to fully restore basal [iCa2+], reduced levels or compromised activity of PMCA result in elevated basal [iCa2+] and/or reduced clearance of the ion21, 22, 34, 47. In agreement with our data showing that PMCA expression in glutamatergic neurons depends on Np in vivo, mature pyramidal Np-deficient neurons cultured for 16 days displayed reduced PMCA levels in comparison to wild-type (Fig. 5e). As expected, we obtained stable measurements of [iCa2+] using single-cell Fluo-4/Fura-Red ratiometric Ca2+ and confocal imaging in the absence of any potentially interfering intrinsic neuronal network activity (1 µM TTX) (Fig. 5f). [iCa2+] was elevated by 51.9 ± 2.1% in pyramidal glutamatergic neurons lacking Np compared to wild-type neurons (Fig. 5g, P < 0.05, unpaired two-tailed t-test). Elevated basal [iCa2+] was also observed in Fluo-4-loaded mutant neurons after fluorescence intensity normalization to wild-type neurons (Fig. 5h). Consistent with the literature34, basal [iCa2+] was restored after washed-out of KCl (30 mM, 15 sec) with a recovery delay of 23.4 ± 6.4 sec in wild-type neurons (Fig. 5h,i). In contrast, slow (Fig. 5i, P < 0.01, unpaired two-tailed t-test) and inefficient restoring of basal [iCa2+] occurred in Np-deficient neurons further indicating insufficient PMCA activity after KCl stimulation. Differences in rise time or maximal response were not observed between the phenotypes (Fig. 5h).
Here, we show that Np ablation in cortical and hippocampal glutamatergic neurons – the predominant neuron type expressing Np in human and rodent brain – results in strong and specific behavioral deficits correlating well with decreased neural activities and reduction of PMCAs in the mutant brain areas. Although Nps are capable of regulating directly the protein levels of PMCAs in cultured cells in vitro, the Np-dependent post-transcriptional expression of each specific PMCA differs between neuron-types in cortex and hippocampus in vivo. These results provide new molecular grounds to link the causality of Np ablation to alterations in [iCa2+] and neuronal activities in cognitive circuits. As the expression, molecular features, and synaptic localization of Np in human and rodent hippocampus and cortex are very similar, the revealed mechanisms of learning and memory are translatable to conditions underlying cognition and its deterioration.
The behavioral analysis of Nptn lox/loxEmx1Cre mice allowed us to identify functions mediated by Np expressed in glutamatergic neurons, which appear to be molecularly distinct from other Np functions e.g. in GABAergic neurons13, 48. Using the water maze, we observed that Nptn lox/loxEmx1Cre mice formed a reliable spatial memory, indicated by their memory for the goal quadrant. However, Nptn lox/loxEmx1Cre mice were not handicapped by moving the platform position during reversal but displayed a more circular alternative strategy indicative for a shift from use of hippocampus-dependent distal to striatum-dependent proximal cues42. In Np65-deficent mice, a role of Np in hippocampus-dependent use of spatial information has been suggested17 and is supported functionally by insufficient mEPSC and synaptic plasticity in the CA1 region10 as well as anatomically by the reduced number of excitatory synapses13. Additionally, the prolonged immobility and frequent re-initiation of swimming by Nptn lox/loxEmx1Cre mice in the water maze caused a delay to reach the platform during acquisition and suggests further alterations in the cortical-striatal pathways related to decision processing49, 50.
In the two-way active avoidance (shuttle-box) paradigm, Nptn lox/loxEmx1Cre mice acquired the associative fear conditioning task normally contrasting strongly with Nptn −/− mice and Nptn lox/loxPrpCreERT induced when adult which both do not learn this task10. Strong evidence indicates that disinhibition –“a transient break in the balance of excitation and inhibition”- of glutamatergic projection neurons mastered by GABAergic interneurons is crucial for acquisition and expression of associative fear conditioning5. In Nptn −/− and induced Nptn lox/loxPrpCreERT mice, Np is lacking in GABAergic neurons causing altered GABAergic transmission10, 13 and potentially affecting disinhibition resulting in inability to acquire this associative task. In contrast, GABAergic interneurons express Np and may be able to regulate inhibition in the glutamatergic Nptn lox/loxEmx1Cre mutant. In favor of this assumption, our results show that expression of Np in glutamatergic neurons is not essential for the acquisition. Strikingly, we identified unequivocally abundantly Np expressing parvalbumin-containing GABAergic interneurons within the cortical layer II/III (Supplementary Results), which are known as a fundamental neuron type mediating disinhibition5. The hypothesis of Np participating in GABAergic disinhibition of glutamatergic neurons fits to previous reports relating Np with the balance of the ratio of excitatory/inhibitory synapses10, 13, 17 via fundamentally independent synaptic mechanisms13.
The capacity of Nptn lox/loxEmx1Cre mice to form associative memories argues against a role of Np expressed by glutamatergic neurons in this process. In contrast, in trained Nptn lox/loxPrpCreERT mice loss of Np in all neuron types elicits retrograde amnesia of associative memories10 and suggests the requirement of Np expressed by other neurons for the retrieval of memories.
In resting Nptn lox/loxEmx1Cre mice, we observed altered patterns of 99mTc-content in those key neuronal circuits known to be directly involved in the processing of the analyzed behaviors. As disturbances in neuronal circuit activities are reflected in detectable changes of 99mTc-content10, 43, we interpret that decreased activities in the mutant hippocampal CA1 and sensorimotor cortex layers II/III are immediate consequences of Np ablation in glutamatergic neurons. Increased activity in the sensorimotor domain of the striatum might result from cortex-striatum circuit adaptation to decreased activity of the mutant sensorimotor glutamatergic afferents44. This striking correlation of permanently altered resting brain activity with compromised specific behavioral performance indicates that the loss of Np in glutamatergic neurons affects basal properties of these neuronal circuits.
The characterization of PMCA isoforms, whose expression is depending decisively on Np, establishes a mechanistic ground linking causally Np ablation to alterations of activities and functions of the mutant cortical and hippocampal areas. PMCA reduction cannot be explained as an adaptive compensatory phenomenon during development in Nptn −/− or Nptn lox/loxEmx1Cre mice as it can be induced efficiently and similarly in adult Nptn lox/loxPrpCreERT mice10. Also, a drastic reduction on PMCA1, 3, and 4 levels occurred in the cortical layers II/III in Nptn lox/loxEmx1Cre mice - coincidently the layers with the highest expression of Np and PMCA in wild-type cortex - arguing for a direct dependency of these paralogs on Np expression in cortical glutamatergic neurons. Very interestingly, PMCA2 levels remained grossly unchanged in Nptn lox/loxEmx1Cre brain but were clearly reduced in Nptn −/− brain. Thus, neuron types other than glutamatergic neurons may express most of the PMCA2 sensitive to Np ablation. Parvalbumin-positive interneurons are good candidates to play this role as we identified them expressing Np and PMCA abundantly and they are known to express particular PMCA2 splicing variants27. Considering PMCAs as key regulators of neuronal activity and synaptic plasticity45, 51, 52, it is very likely that the reduction of cortical PMCA1, 3, and 4 contributed directly to decreased cortical activity in Nptn lox/loxEmx1Cre mice. Similarly, loss of hippocampal PMCA4 may be related to decreased activity in the mutant CA1 hippocampal area. Although hippocampal PMCA1, 2, and 3 were only slightly reduced it is not possible to fully exclude that the collective decrease of these three paralogs could further contribute to [iCa2+] alterations in hippocampal circuit activity.
Reduction of PMCA protein was not accompanied by changes of mRNA paralog levels in vivo. Using cDNA transfection of human or rat Np55 or Np65 in human or rat cultured cells, we demonstrated that either exogenous Np55 or Np65 is sufficient to promote PMCA expression independently of the cellular transcriptional program. Thus, we conclude that Nps have the capacity to modulate PMCA levels post-transcriptionally. Indeed, human and rat Np55/65 increased levels of and co-localized in the cell plasma membrane with hPMCAs suggesting the conserved function between species. Furthermore, our recent findings indicate that PMCAs encoded by all four paralogs form complexes with Np55/65 with a defined stoichiometry (Dr. Thilo Kähne, unpublished data) in the neuronal plasma membrane to maintain low basal [iCa2+]. It is tempting to speculate, that Np may stabilize PMCA or withdraw it from degradation by forming complexes. In the light of our data confirming Np as necessary for diverse and/or specialized PMCA expression in vivo for example PMCA1, 3 in cortical glutamatergic neurons, we propose that specific Np-PMCA complexes regulate differentially [iCa2+]-homeostasis and neuronal activity of particular neuron types in brain areas related to cognitive processing.
Recently a genome-wide association (GWA) study of intelligence differences on 1,500 adolescents concluded that Np gene expression accounted for 0.5% of the variance of cognitive capacities15. Although, GWA studies have contributed to the understanding that many genes exert small effects on the heritability of intelligence, the association of Np with intelligence has been diminished based on a larger analysis of 18,000 subjects38. Therefore, the exact relation of Np expression and high cognitive functions as well as the contribution to this trait in a given genetic background or social environment remain to be determined. Evident similarities in expression and molecular features between human and rodent Np further support the translational value of Np animal models for learning and memory. We described prominent inter-species similarities in Np distribution in the principal excitatory pathways across the glutamatergic circuits within human hippocampus and entorhinal cortex, namely strong hNp expression in the cell-containing layers, in the somatic and neuropil layers of the CA1, in the inner molecular layer of the DG and in pyramidal neurons from the layer II and IV/V in the entorhinal cortex. Despite the highly specific and efficient hNp immunodetection, we did not observe relative enrichment of hNp immunoreactivity in the CA1 area compared to CA3 or DG contrasting with the rodent CA1 region. This difference can be explained by anatomical variations in position in the used human sections as well as known histological differences such as larger CA1 and CA2 areas and thinner DG polymorphic layer in humans vs. rodents53, 54. As in rodents, hNp mRNA levels increase rapidly after birth, are high during adolescence-young adulthood, and decline with age. After translation, hNp mRNAs yield the two main glycosylated isoforms hNp55 and hNp65, which co-exist in close proximity in the neuronal plasma membrane of glutamatergic neurons and in glutamatergic synapses. Interestingly, Np55 and Np65 are abundant in human PSDs contrasting rodent PSDs where only Np65 is detected suggesting a greater demand of Np in human glutamatergic transmission, for example, to regulate PMCA levels. Deregulated expressions of PMCA431 and Np16 have been reported independently in schizophrenic patients. Interestingly, the severely impaired startle response and prepulse inhibition revealed a strong schizophrenic component in the behavior of Nptn −/− and Nptn +/- mice10. Very recently, autism has been associated to low expression of ATP2B229, 30, the human gene encoding PMCA2, which may be compared with decreased PMCA2 in Nptn −/− mice displaying altered social behavior10.
The experimental approach using Np ablation in mice is a promising unique tool for studying mechanisms related to learning and memory and cognitive dysfunctions. As we show here, brain area- and neuron type-selective Np ablation may not only allow the delineation of causal molecular mechanisms in specific neuronal circuits influencing higher brain processes, but may also model important aspects of iCa2+-homeostasis dysregulation linked to cognitive deterioration and potentially to neuropathological conditions.
Human mRNA expression and brain material
All procedures involving human subjects were done in accordance with the approval of the Central Ethical Committee of the University of Zagreb Medical School case no. 380-59/11-500-77/90, class 641-01/11-02 signed on 19th May 2011. We used http://BrainCloud.jhmi.edu 55 to evaluate gene expression (mRNA levels) in dorsolateral prefrontal cortex from 215 subjects (0 to 80 years old) with RNA integrity number ≥8 as reported15. Probes targeting translated mRNA regions were selected. Human material from six subjects with no history of neuropsychiatric diseases and cause of death unrelated to neurological disorders or head trauma (mean age: 69.5; range: 59–81 years old) was obtained from the Zagreb brain collection56 and the Brain Bank at Karolinska Institute (Stockholm, Sweden) with permission of the ethical committees. For staining, brains were fixed (10% pH neutral formalin for 21 days) and hippocampal-entorhinal cortex was cut in rostro-caudal direction. We used five 12 μm thick sections for Nissl staining to delineate cytoarchitectonic boundaries. Frozen brain samples derived from one additional subject were processed for biochemical analysis, as described below.
After dewaxing and rehydration, human brain sections were incubated in citrate buffer (pH 6.0, 20 min, 95 °C), treated with 0.3:25:74.7 hydrogen peroxide:methanol:water (20 min), immersed in blocking solution (5% horse serum, 0.5% Triton X-100 in PBS, Sigma) for two hours, and incubated with goat anti-Np65 in blocking solution (1:500, R&D Systems, overnight, 4 °C). Incubation with anti-goat secondary antibody conjugated with horse-radish peroxidase (1:1000, Jackson Immunolaboratories) in blocking solution followed (one hour, room temperature). Negative controls included pre-immune serum replacing primary antibody, omission of secondary antibody, or use of non-related secondary antibody. Sections were visualized using diaminobenzidine (DAB) and scanned (453 nm/pixel) using a Hamamatsu NanoZoomer C10730-12. Signal intensities were scored using a bright-field microscope by three independent researchers as none (0), low (+), moderate (++), or strong (+++). Additional quantification of immunoreactivity used ImageJ densitometric analysis.
Neuroplastin mutant mice, behavior, and SPECT-imaging
Nptn −/− and floxed Nptn loxlox mice were described10, 13. To obtain Nptn loxloxEmx1Cre mice, B6.129S2-Emx1tm1(Cre) (The Jackson Laboratory) were crossed with Nptn loxlox mice and maintained on a Nptn loxlox background. Mice were kept with a 12-hour light/dark cycle and food and water ad libitum. All procedures were in accordance with institutional, state, and government regulations (Landesverwaltungsamt Sachsen-Anhalt, permit 42502-2-1366 LIN). Associative learning was assessed by two-way active avoidance in a shuttle-box, spatial learning in the hidden platform Morris Water Maze, and motor coordination using a rota-rod as described10, 57 and detailed in Supplementary Material and Methods. Small animal single photon emission computed tomography (SPECT) imaging of regional cerebral blood flow (rCBF) was performed in vivo mapping spatial patterns of neuronal activity as published10, 43 and described in Supplementary Material and Methods.
Subcellular fractionation, protein deglycosylation, Western blot analysis, and Immunofluorescence
Samples were stored at −80 °C until fractionation experiments, unfrozen in ice-cold buffer (320 mM sucrose, 5 mM HEPES, pH 7.4), and homogenized. Subcellular fractions were obtained as described58. Deglycosylation, Western blot analysis, and Immunofluorescence are described in Supplementary Material and Methods.
Cell cultures, cDNA transfections, and iCa2+ measurements
Np mutant hippocampal neurons were prepared and cultured for 12–16 DIV as described13. Human Embryonic Kidney (HEK) cells were maintained as recommended by ATCC. Cells were transfected using Lipofectamine 2000 as recommended and analyzed after 24 hours. cDNAs: Human PMCA2 constructs (ID: 47581 and 47584) and PMCA1 (ID:47758) were purchased from Addgene (Cambridge, MA), rat Np constructs were described40, and human Np65 (OHu06230D, Genscript, Piscataway, NJ) was inserted into pFUGW-EGFP (Addgene ID:14883). For single cell ratiometric iCa2+ imaging, cell somata were loaded with Fluo-4 (1.3 μg/ml) and Fura-Red (2.7 μg/ml) fluorescent probes for 30 minutes. A single wavelength laser (488 nm) was used for stimulation, and fluorescence emission was collected dually at 520 and 680 nm. Ratio of fluorescent intensities was normalized against the basal ratio of wild-type neurons as required. Measurements were performed with a confocal microscope under controlled conditions of temperature and pH. To prevent interference of spontaneous action potentials, intrinsic neuronal activity was silenced using 1 μM TTX in the recording bath (145 mM NaCl, 2.5 mM KCl, 2 mM MgCl2, 2 mM CaCl2, 10 mM HEPES, and 10 mM D-glucose, pH 7.4). Perfusion (2 ml/min) of isosmotic 30 mM KCl for 15 sec was performed according published protocols34 and using a RC-37R chamber from Warner Instruments (Hamden, CT, USA).
Statview (SAS Institute Inc., Cary, NC) was used for Analysis of Variance (ANOVA), post hoc analysis (Scheffe’s or Fisher PLSD), and repeated measures ANOVA. In the voxelwise analysis unpaired t-tests were made to compare brain 99mTc distributions using the MagnAn-software (version 2.4, BioCom, Germany) and uncorrected p-values59,60,61.
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Bird, C. M. & Burgess, N. The hippocampus and memory: insights from spatial processing. Nat. Rev. Neurosci. 9, 182–194 (2008).
Wang, S. H. & Morris, R. G. Hippocampal-neocortical interactions in memory formation, consolidation, and reconsolidation. Ann. Rev. Psychol. 61, 49–79 (2010).
Preston, A. R. & Eichenbaum, H. Interplay of hippocampus and prefrontal cortex in memory. Curr. Biol. 23, R764–773 (2013).
Bannerman, D. M. et al. Hippocampal synaptic plasticity, spatial memory and anxiety. Nat. Rev. Neurosci. 15, 181–192 (2014).
Letzkus, J. J., Wolff, S. B. E. & Lüthi, A. Disinhibition, a Circuit Mechanism for Associative Learning and Memory. Neuron 88, 264–276 (2015).
Dalva, M. B., McClelland, A. C. & Kayser, M. S. Cell adhesion molecules: signalling functions at the synapse. Nat. Rev. Neurosci. 8, 206–220 (2007).
Dean, C. & Dresbach, T. Neuroligins and neurexins: linking cell adhesion, synapse formation and cognitive function. Trends Neurosci. 29, 21–29 (2006).
Sandi, C. Stress, cognitive impairment and cell adhesion molecules. Nat. Rev. Neurosci. 5, 917–930 (2004).
Beesley, P. W., Herrera-Molina, R., Smalla, K. H. & Seidenbecher, C. The Neuroplastin adhesion molecules: key regulators of neuronal plasticity and synaptic function. J. Neurochem. 131, 268–283 (2014).
Bhattacharya, S. et al. Genetically Induced Retrograde Amnesia of Associative Memories After Neuroplastin Ablation. Biol. Psychiatry 81, 124–135 (2017).
Empson, R. M. et al. The cell adhesion molecule neuroplastin-65 inhibits hippocampal long-term potentiation via a mitogen-activated protein kinase p38-dependent reduction in surface expression of GluR1-containing glutamate receptors. J. Neurochem. 99, 850–860 (2006).
Smalla, K. H. et al. The synaptic glycoprotein neuroplastin is involved in long-term potentiation at hippocampal CA1 synapses. Proc. Natl. Acad. Sci. USA 97, 4327–4332 (2000).
Herrera-Molina, R. et al. Structure of excitatory synapses and GABAA receptor localization at inhibitory synapses are regulated by neuroplastin-65. J. Biol. Chem. 289, 8973–8988 (2014).
Carrott, L. et al. Absence of Neuroplastin-65 Affects Synaptogenesis in Mouse Inner Hair Cells and Causes Profound Hearing Loss. J. Neurosci. 36, 222–234 (2016).
Desrivières, S. et al. Single nucleotide polymorphism in the neuroplastin locus associates with cortical thickness and intellectual ability in adolescents. Mol. Psychiatry. 20, 263–274 (2015).
Saito, A. et al. Association study of putative promoter polymorphisms in the neuroplastin gene and schizophrenia. Neurosci. Lett. 411, 168–173 (2007).
Amuti, S. et al. Neuroplastin 65 mediates cognitive functions via excitatory/inhibitory synapse imbalance and ERK signal pathway. Neurobiol. Learn. Mem. 127, 72–83 (2016).
Zeng, W. Z. et al. Neuroplastin Isoform Np55 Is Expressed in the Stereocilia of Outer Hair Cells and Required for Normal Outer Hair Cell Function. J. Neurosci. 36, 9201–9216 (2016).
Miller, P. D., Chung, W. W., Lagenaur, C. F. & DeKosky, S. T. Regional distribution of neural cell adhesion molecule (N-CAM) and L1 in human and rodent hippocampus. J. Comp. Neurol. 327, 341–349 (1993).
Bernstein, H. G. et al. The immunolocalization of the synaptic glycoprotein neuroplastin differs substantially between the human and the rodent brain. Brain Res. 1134, 107–112 (2007).
Strehler, E. E. Plasma membrane calcium ATPases: From generic Ca(2+) sump pumps to versatile systems for fine-tuning cellular Ca(2+). Biochem. Biophys. Res. Comm. 460, 26–33 (2015).
Strehler, E. E. Plasma membrane calcium ATPases as novel candidates for therapeutic agent development. J. Pharmacy Pharmaceut. Sci. 16, 190–206 (2013).
Kip, S. N. et al. Changes in the expression of plasma membrane calcium extrusion systems during the maturation of hippocampal neurons. Hippocampus 16, 20–34 (2006).
Stahl, W. L., Keeton, T. P. & Eakin, T. J. The plasma membrane Ca(2+)-ATPase mRNA isoform PMCA 4 is expressed at high levels in neurons of rat piriform cortex and neocortex. Neurosci. Lett. 178, 267–270 (1994).
Burette, A., Rockwood, J. M., Strehler, E. E. & Weinberg, R. J. Isoform-specific distribution of the plasma membrane Ca2+ ATPase in the rat brain. J. Comp. Neurol. 467, 464–476 (2003).
Kenyon, K. A. et al. Cellular and subcellular localization of the neuron-specific plasma membrane calcium ATPase PMCA1a in the rat brain. J. Comp. Neurol. 518, 3169–3183 (2010).
Burette, A. C., Strehler, E. E. & Weinberg, R. J. “Fast” plasma membrane calcium pump PMCA2a concentrates in GABAergic terminals in the adult rat brain. J. Comp. Neurol. 512, 500–513 (2009).
Burette, A. C., Strehler, E. E. & Weinberg, R. J. A plasma membrane Ca2+ ATPase isoform at the postsynaptic density. Neurosci. 169, 987–993 (2010).
Carayol, J. et al. Converging Evidence for an Association of ATP2B2 Allelic Variants with Autism in Male Subjects. Biol. Psychiatry 70, 880–887 (2011).
Oksenberg, N. et al. Genome-wide distribution of Auts2 binding localizes with active neurodevelopmental genes. Transl. Psychiatry 4, e431, doi:10.1038/tp.2014.78 (2014).
Martins-de-Souza, D. Alterations in oligodendrocyte proteins, calcium homeostasis and new potential markers in schizophrenia anterior temporal lobe are revealed by shotgun proteome analysis. J. Neural Transmission 116, 275–289 (2009).
Berrocal, M. et al. Inhibition of PMCA activity by tau as a function of aging and Alzheimer’s neuropathology. Biochim. Biophys. Acta 1852, 1465–1476 (2015).
Mata, A. M., Berrocal, M. & Sepulveda, M. R. Impairment of the activity of the plasma membrane Ca(2)(+)-ATPase in Alzheimer’s disease. Biochem. Soc. Trans. 39, 819–822 (2011).
Pérez-Cañamás, A. et al. Sphingomyelin-induced inhibition of the plasma membrane calcium ATPase causes neurodegeneration in type A Niemann-Pick disease. Mol. Psychiatry Epub ahead of print 13 Sep 2016, doi:10.1038/mp.2016.148 (2016).
Schultz, J. M. et al. Modification of human hearing loss by plasma-membrane calcium pump PMCA2. New Engl. J. Med. 352, 1557–1564 (2005).
Calì, T. et al. A Novel Mutation in Isoform 3 of the Plasma Membrane Ca2+ Pump Impairs Cellular Ca2+ Homeostasis in a Patient with Cerebellar Ataxia and Laminin Subunit 1α Mutations. J. Biol. Chem. 290, 16132–16141 (2015).
Zanni, G. et al. Mutation of plasma membrane Ca2+ ATPase isoform 3 in a family with X-linked congenital cerebellar ataxia impairs Ca2+ homeostasis. Proc. Natl. Acad. Sci. USA 109, 14514–14519 (2012).
Plomin, R. & Deary, I. J. Genetics and intelligence differences: five special findings. Mol. Psychiatry 20, 98–108 (2015).
Seress, L., Abraham, H., Czeh, B., Fuchs, E. & Leranth, C. Calretinin expression in hilar mossy cells of the hippocampal dentate gyrus of nonhuman primates and humans. Hippocampus 18, 425–434 (2008).
Langnaese, K., Beesley, P. W. & Gundelfinger, E. D. Synaptic membrane glycoproteins gp65 and gp55 are new members of the immunoglobulin superfamily. J. Biol. Chem. 272, 821–827 (1997).
Kang, H. J. et al. Spatio-temporal transcriptome of the human brain. Nature 478, 483–489 (2011).
Vorhees, C. V. & Williams, M. T. Assessing spatial learning and memory in rodents. ILAR J. 55, 310–332 (2014).
Kolodziej, A. et al. SPECT-imaging of activity-dependent changes in regional cerebral blood flow induced by electrical and optogenetic self-stimulation in mice. NeuroImage 103, 171–180 (2014).
Van Waes, V., Beverley, J. A., Siman, H., Tseng, K. Y. & Steiner, H. CB1 Cannabinoid Receptor Expression in the Striatum: Association with Corticostriatal Circuits and Developmental Regulation. Front Pharmacol. 3, 21, doi:10.3389/fphar.2012.00021 (2012).
Sinkins, W. G. et al. Maitotoxin converts the plasmalemmal Ca(2+) pump into a Ca(2+)-permeable nonselective cation channel. Am. J. Physiol. Cell Physiol. 297, C1533–1543 (2009).
Owczarek, S. et al. Neuroplastin-55 binds to and signals through the fibroblast growth factor receptor. FASEB J. 24, 1139–1150 (2010).
Simons, S. B., Escobedo, Y., Yasuda, R. & Dudek, S. M. Regional differences in hippocampal calcium handling provide a cellular mechanism for limiting plasticity. Proc. Natl. Acad. Sci. USA 106, 14080–14084 (2009).
Sarto-Jackson, I. et al. The cell adhesion molecule neuroplastin-65 is a novel interaction partner of gamma-aminobutyric acid type A receptors. J. Biol. Chem. 287, 14201–14214 (2012).
Amemori, K. I., Gibb, L. G. & Graybiel, A. M. Shifting responsibly: the importance of striatal modularity to reinforcement learning in uncertain environments. Front. Hum. Neurosci. 5, 47, doi:10.3389/fnhum.2011.00047 (2011).
Znamenskiy, P. & Zador, A. M. Corticostriatal neurons in auditory cortex drive decisions during auditory discrimination. Nature 497, 482–485 (2013).
Empson, R. M., Garside, M. L. & Knopfel, T. Plasma membrane Ca2+ ATPase 2 contributes to short-term synapse plasticity at the parallel fiber to Purkinje neuron synapse. J. Neurosci 27, 3753–3758 (2007).
Jensen, T. P., Buckby, L. E. & Empson, R. M. Reduced expression of the “fast” calcium transporter PMCA2a during homeostatic plasticity. Mol. Cell. Neurosci. 41, 364–372 (2009).
West, M. J. Stereological studies of the hippocampus: a comparison of the hippocampal subdivisions of diverse species including hedgehogs, laboratory rodents, wild mice and men. Prog. Brain Res. 83, 13–36 (1990).
Amaral, D. G., Scharfman, H. E. & Lavenex, P. The dentate gyrus: fundamental neuroanatomical organization (dentate gyrus for dummies). Prog Brain Res. 163, 3–22 (2007).
Colantuoni, C. et al. Temporal dynamics and genetic control of transcription in the human prefrontal cortex. Nature 478, 519–523 (2011).
Judaš, M. et al. The Zagreb Collection of human brains: a unique, versatile, but underexploited resource for the neuroscience community. Annal. N. Y. Acad. Sci. 1225(Suppl 1), E105-130, doi:10.1111/j.1749-6632.2011.05993.x. (2011).
Montag-Sallaz, M., Schachner, M. & Montag, D. Misguided axonal projections, NCAM180 mRNA upregulation, and altered behavior in mice deficient for the Close Homolog of L1 (CHL1). Mol. Cell. Biol. 22, 7967–7981 (2002).
Smalla, K. H., Klemmer, P. & Wyneken, U. Isolation of the Postsynaptic Density: A Specialization of the Subsynaptic Cytoskeleton in Springer Protocols Neuromethods: The Cytoskeleton - Imaging, Isolation, and Interaction (ed. Dermietzel, R.) 265–282 (Humana Press: New York, 2012).
Endepols, H. et al. Effort-based decision making in the rat: an [18F]fluorodeoxyglucose micro positron emission tomography study. J. Neurosci. 30, 9708–9714 (2010).
Thanos, P. K. et al. Mapping brain metabolic connectivity in awake rats with μPET and optogenetic stimulation. J. Neurosci. 33, 6343–6349 (2013).
Wyckhuys, T. et al. Hippocampal deep brain stimulation induces decreased rCBF in the hippocampal formation of the rat. Neuroimage 52, 55–61 (2010).
RH-M and KM-J gratefully acknowledge the German Academic Exchange Service (DAAD). Grants obtained from Deutsche Forschungsgemeinschaft supported RH-M (SFB854-B08) and SKV (GRK1167). GS thanks the Croatian Science Foundation (CSF-IP-2014-09-9730). The publication of this article was funded by the Open Access Fund of the Leibniz Association. We thankfully acknowledge the technical assistance of Karla Sowa, Daniela Hill, Kristin Böttger, Ting Jiao, Dr. Angela Breme and Dr. Anja Oelschlegel.
The authors declare that they have no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
About this article
Cite this article
Herrera-Molina, R., Mlinac-Jerkovic, K., Ilic, K. et al. Neuroplastin deletion in glutamatergic neurons impairs selective brain functions and calcium regulation: implication for cognitive deterioration. Sci Rep 7, 7273 (2017) doi:10.1038/s41598-017-07839-9
Journal of Cellular and Molecular Medicine (2019)
Cold Spring Harbor Perspectives in Medicine (2019)
Neuronal Signaling by Thy-1 in Nanodomains With Specific Ganglioside Composition: Shall We Open the Door to a New Complexity?
Frontiers in Cell and Developmental Biology (2019)
Neuroplastin 65 modulates anxiety‐ and depression‐like behavior likely through adult hippocampal neurogenesis and central 5‐ HT activity
The FEBS Journal (2019)