Prion protein is essential for the RE1 silencing transcription factor (REST)-dependent developmental switch in synaptic NMDA receptors

Article metrics


It is important that the correct amounts of GluN2 subunits are maintained, as they determine NMDAR functional properties, which are crucial to neuronal communication, synaptogenesis and cognitive function. The transcriptional repressor RE1 silencing transcription factor (REST) is critical for the postnatal developmental switch in NMDARs. However, the mechanisms triggering REST and the link between NMDARs and REST are unclear. Here we show a new physiological essential role for cellular prion protein (PrPC) in REST-dependent homeostasis and the developmental switch of NMDARs. REST and REST-associated proteins were overactivated in the hippocampi of Prnp knockout mice (Prnp0/0) compared with wild-type Prnp (Prnp+/+) mice. This coincided with the disruption of the normal developmental switch from GluN2B-to-GluN2A in vivo. PrPC co-located with REST under physiological environments and mediated the translocation of REST in conditioners of NMDARs in vitro in Prnp+/+ hippocampal neurons. Regardless of whether REST was knocked down or overexpressed, deletion of PrPC not only disrupted REST-mediated distribution of mitochondria, but also prevented REST-regulated expression of GluN2B and GluN2A in Prnp0/0. Importantly, these effects were rescued after overexpression of full-length PrPC through restoration of NMDAR2 subunits and their distributions in dendritic processes in Prnp0/0. Consistently, knockdown of PrPC in Prnp+/+ had a similar effect on Prnp0/0. Furthermore, PrPC colocalized with both GluN2B and GluN2A in Prnp+/+. For the first time, we demonstrate that PrPC is essential for REST-regulated NMDARs. Confirming the regulation of NMDAR-modulating mechanisms could provide novel therapeutic targets against dysfunctions of glutamatergic transmission in the nervous system.


N-methyl-d-aspartate receptors (NMDARs) are glutamate-gated ion channels critical for synaptogenesis and neuronal communication1. These heterotetrameric channels are formed by the assembly of two obligatory GluN1 and two GluN2/GluN3 subunits2,3. NMDARs subunit composition is plastic and diverse, leading to abundant receptor subtypes, each with its own biophysical and pharmacological properties1. The subunit isoform of NR2 (GluN2) is a key determinant of the functional capabilities of NMDARs, including activation, deactivation and desensitization kinetics4. A developmental switch from containing primarily GluN2B-to-GluN2A occurs during postnatal development in NMDARs5. This switch, as well as the correct GluN2A/CluN2B ratio, is critical for neural circuitry6, hippocampus-dependent learning7, plasticity-induced alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors8 and spine growth9. A previous report has shown that the transcriptional repressor element 1-silencing transcription factor (REST), also known as neuron-restrictive silencer factor10, participates in the postnatal switch in synaptic NMDAR subunit by decreasing GluN2B expression through epigenetic remodeling of Grin2b at rat hippocampal synapses. During postnatal development, REST shows both target and temporal characteristics in differentiated neurons for GluN2B11. REST acts as a multiple hub in coordination with other factors regulating the multiple aspects of neurogenesis and preserving the distinct neural phenotype10. However, as a switch from primarily GluN2B-to-GluN2A, the mechanisms that trigger REST expression in differentiated neurons, and the following long-term increase in GluN2A expression during postnatal development, are still unclear.

Although under normal conditions NMDA receptors mediate important physiological functions such as learning and memory, they also take part in glutamate excitotoxicity, which can occur in response to ischemia and is related to many neurodegenerative diseases, including Alzheimer's Disease (AD)12. Hence, strict regulation of NMDAR activity plays a key role in homeostasis by preventing excitotoxicity13. It is well established that misfolded forms of cellular prion protein (PrPC) transform into the β-sheet-rich, aggregate-prone scrapie conformation (PrPSc), resulting in several progressive fatal diseases, known as prion diseases. These transmissible spongiform encephalopathies include bovine spongiform encephalopathy, scrapie, Creutzfeld–Jakob disease, Gerstmann–Straussler–Scheinker syndrome in humans14,15. Although research has been conducted exploring the harmful effects of misfolded or aggregated prion proteins, the physiological roles of PrPC remain only partly understood.

Recent studies have demonstrated that PrPC also communicates with NMDARs; PrPC-deficient mice exhibit enhanced NMDAR-dependent neuronal excitability and are more susceptible to NMDA-induced excitotoxicity16. Research has cumulatively demonstrated that the absence of PrPC increased NMDAR glycine affinity, resulting in persistent NMDAR activity after prolonged agonist treatment17. PrPC and GluN2D are found in the same protein complex and colocalize in hippocampal neurons16. However, in the absence of PrPC, it is still not clear whether other subunits of NMDARs are altered. Therefore, more details of PrPC-mediated regulation of NMDARs need to be revealed.

Here, we explore the relationship between PrPC and REST-dependent developmental switch from GluN2B-to-GluN2A. Our data show that a lack of PrPC gives rise to REST overexpression and disorder of REST-associated proteins in the hippocampus of neonatal mice. REST-dependent epigenetic remodeling of the developmental switch of NMDARs is also suppressed in vivo. Comparing wild-type (WT; Prnp+/+) and Prnp knockout (Prnp0/0) mice primary cultured hippocampal neurons, we found that PrPC not only affects the translocation of REST, but also partially mediates REST-regulated mitochondria distribution and the developmental switch in synaptic NMDARs. The adverse effects induced by overactivated REST in Prnp0/0 neurons are recapitulated by overexpression of exogenous PrPC. Thus, we demonstrate a novel functional role for native PrPC as an essential modulator of REST-dependent NMDARs activity, and provide more evidence to support the hypothesis that PrPC is a regulator of NMDARs.


Overexpression of REST and alterations of REST-associated proteins in the hippocampi of postnatal Prnp 0/0 mice

In almost all tissues, including the brain, epigenetic modification of chromatin is an important modulator of gene expression18,19. First, we confirmed the developmental expression of PrPC in the hippocampus of Prnp+/+ neonatal mice. Consistent with previous reports20,21, PrPC levels increased during the first two postnatal weeks, reached a peak at P13 and were slightly diminished in the adult (Fig. 1a, b). Second, we tested and compared postnatal expression of REST, postsynaptic density 95 kDa protein (PSD-95), synaptophysin (SYP), GluN2B, GluN2A and NMDAR1 in the hippocampal homogenates of Prnp+/+ and Prnp0/0 neonatal mice.

Fig. 1: Western blot (WB) analyses of PrPC (in wild-type (WT) only), REST, PSD-95, SYP, GluN2B, GluN2A, NMDAR1, total and phosphorylated β-catenin and GSK3β during WT Prnp (Prnp+/+) and PrP-null (Prnp0/0) mice hippocampal postnatal development.

a, e, m WB results in whole hippocampal lysates of WT mice (n = 6). c, i, n WB results in whole hippocampal lysates of Prnp0/0 mice (n = 6). b Quantitative analyses of (a). d Quantitative analyses of (f). fh Quantitative analyses of (e). j-l Quantitative analyses of i. Immunoblot density in b and d normalized to β-actin. Immunoblot density in fh and jl normalized to GAPDH. All values were normalized (dashed lines) relative to corresponding data at P3 in each group. Data are presented as means ± SD (n = 6). *P < 0.05; **P < 0.01; ***P < 0.001 vs corresponding data at P3

In the WT group, contrary to the tendency of PrPC, a transient, but marked increase in REST abundance occurred at postnatal day 3 (P3; Fig. 1a, b), with a subsequent, and constant, abundance of GluN2B protein from P7 until P13. REST abundance then declined ~2.25-fold (relative to P3) to a level that persisted until adulthood (Fig. 1e, f). GluN2A protein was barely detectable at P3, and its levels progressively increased ~6.43-fold by P30 relative to that at P3, and continued to increase until P60 (nearly ninefold) (Fig. 1e, g). NMDAR1 levels gradually increased in a similar way to GluN2A (Fig. 1e, h).

In the Prnp0/0 group, REST expression significantly increased after P5 (~9.32-fold) and the highest levels were seen at P13 (~28.32-fold) relative to P3 (Fig. 1c, d). Although GluN2B levels progressively declined after P3 (Fig. 1i, j), GluN2B showed higher expression from P3 to P30 compared with the WT group (Fig. 2a, d). GluN2A was barely detectable before P5, but its expression markedly increased from P7 to P13 (Fig. 1i, k). NMDAR1 was strongly expressed from P3 to P7 and decreased to undetectable levels at P60 (Fig. 1i, l).

Fig. 2: WB analyses of REST and associated proteins to compare Prnp+/+ (n  =  6) with Prnp0/0 (n  =  6) mice hippocampal postnatal development.

a Immunoblot of REST, GluN2B, GluN2A, and NMDAR1. b Immunoblot of total and phosphorylated β-catenin and GSK3β proteins. cf Quantitative analyses of (a). gj Quantitative analyses of (b). Immunoblot density in cg and i normalized to GAPDH and expressed as the ratio to GAPDH density. Immunoblot density in h and j showing the quantification of β-catenin (Ser33)/total β-catenin protein, GSK3β (Ser9)/total GSK3β protein, total β-catenin and GSK3β protein normalized to GAPDH. All values were normalized (dashed lines) relative to corresponding data in the Prnp+/+ group. Data are presented as means ± SD (n = 6). *P < 0.05; **P < 0.01; ***P < 0.001 vs Prnp+/+

Compared with WT controls, Prnp0/0 mice showed significantly higher expression of REST (Fig. 2c), GluN2B (Fig. 2d) and GluN2A (Fig. 2e) during postnatal development (Fig. 2a). NMDAR1 expression was higher from P3 to P9 but decreased remarkably after P30 (Fig. 2f). Levels of SYP and PSD-95, the pre- and postsynaptic markers, showed a consistent upward trend and were not significantly different between the two groups.

We recently demonstrated that β-catenin were expressed in parallel with REST under physiological22 and pathological conditions in LRP6-Wnt-β-catenin signaling pathways23,24. Conversely, total GSK3β protein has an inverse relationship with REST24. Our results reveal that: (1) total β-catenin and p-β-catenin (Ser552)/total β-catenin protein had different expression patterns between WT and PrP-null mice: the former remained relatively high after P7 (Fig. 1m, Fig. S1E, and Fig. S1F), whereas the latter was more highly expressed at the beginning of P3 (Fig. 1n, Fig. S1I, and Fig. S1J). (2) Comparing WT and Prnp0/0 mice, total β-catenin was still more highly expressed in the Prnp0/0 group, and p-β-catenin (Ser552)/total β-catenin protein was generally more highly expressed in the latter, except at P9, but significantly decreased at P60 (Fig. 2b, g, h). (3) Although total GSK3β protein and p-GSK3β (Ser9)/total GSK3β protein had similar expression patterns in both groups (Fig. 1m, Fig. S1G, and Fig. S1H; Fig. 1n, Fig. S1K, and Fig. S1L), GSK3β protein levels were lower in Prnp0/0 mice except at P11, and p-GSK3β (Ser9)/total GSK3β protein was higher in the latter until P60. These findings are consistent with the state of REST in different groups.

Overall, Prnp0/0 mice exhibited REST overexpression and disordered postnatal developmental switching from GluN2B-to-GluN2A, together with overactivation of β-catenin and suppression of GSK3β.

PrPC is essential for REST functional translocation

Previous reports have shown that activation and translocation of REST is a universal feature in response to stressors25. Nuclear REST is a key factor, not only for developmental switch of NMDARs under physiological conditions, but also for neuroprotection in neurodegenerative diseases, such as prion diseases23,24 and AD22. Thus, to further study the potential effect of PrPC on REST, we compared the expression and distribution of REST in primary cultured WT and Prnp0/0 hippocampal neurons by immunofluorescence (IF) and western blot (WB) analysis. In the WT group, PrPC colocalized with REST in the cytoplasm under normal conditions. N-methyl-d-aspartic acid (NMDA), a selective NMDAR agonist (10 μM, 24 h), stimulated the accumulation of REST partly in synapses, and a larger increase of PrPC in the soma. Lithium chloride (LiCl), a selective REST agonist (10 mM, 48 h), remarkably induced the nuclear translocation of REST (Fig. 3a, b) (Fig. S2), a result consistent with our previous report. However, in the absence of PrPC, REST had very little response to the agonist and most of it located in the cytoplasm (Fig. 3a, c). Quantitative WB analysis revealed results consistent with IF (Fig. 3d–i). Although total REST in the PrP-null group was 1.31-fold higher than in the WT group under normal conditions, total REST significantly decreased by 40.41% and 32.22% after exposure to NMDA and LiCl, respectively (Fig. 3d, g). What’s more, in the Prnp0/0 groups, compared with the control, total REST decreased significantly more, by 45.44% and 51.62% in the NMDA and LiCl, respectively (Fig. S3). Nuclear REST was markedly lost in each group when PrPC was absent (Fig. 3e, h). This demonstrates that the functional translocation of REST in response to NMDA and LiCl treatment depends on the presence of PrPC.

Fig. 3: PrPC regulates the functional translocation of REST.

a Representative double-staining confocal immunofluorescent images of Prnp+/+ and Prnp0/0 mice primary hippocampal neurons for REST (green) and PrPC (red) in each group without treatment or treated with NMDA or LiCl. Nuclei (blue) are stained with DAPI. Scale bars = 10 μm. b, c Fluorescence quantitative analyses of the ratio of REST in the nucleus to REST in the cytoplasm in Prnp+/+ and Prnp0/0 (a). Fluorescence intensity was normalized to each control (dashed lines) and ***P < 0.001 vs the control. d-f Immunoblotting confirms the total amount, nuclear amount and cytoplasmic amount of REST in Prnp+/+ and Prnp0/0 groups, as indicated. Nuclear and cytoplasmic fractions were collected separately and the fractions immunoblotted for REST. The nucleus-localized proteins Lamin B and GAPDH demonstrate separation of the nuclear and cytoplasmic fractions. g Quantitative analyses of (d); h Quantitative analyses of (e); i Quantitative analyses of (f). Total REST normalized to β-actin. Nuclear and cytoplasmic REST are normalized to Lamin B and GAPDH, respectively, and expressed as a ratio to the corresponding data in Prnp+/+. Data are presented as means ± SD of triplicate experiments. *P < 0.05; **P < 0.01; ***P < 0.001 vs Prnp+/+

PrPC is essential for REST-dependent NMDAR expression

REST provides a regulatory hub that coordinately regulates multiple physiological and pathological pathways of neuronal development and neurological diseases in vitro and in vivo10. Although REST-dependent epigenetic remodeling is critical to ischemia-induced neuronal death26, other reports reveal that overexpression of REST plays a critical neuroprotective role22,24. Therefore, we examined whether overexpression or knockdown of REST could functionally recover NMDARs expression in the absence of PrPC. Thus, we examined the expression of GluN2B, GluN2A and NMDAR1 when REST was blocked or overexpressed in WT or PrP-null neurons. Importantly, overexpression of REST significantly increased the level of GluN2B by 7.47-fold compared with the HA vector control in the WT group. Conversely, knockdown of REST markedly suppressed the level of GluN2B by 32.16% but increased the expression of Glu2A 1.57-fold compared with WT control (Fig. 4a, b). By contrast, neither knockdown nor overexpression of REST had any significant effect on NMDAR expression (Fig. 4a, b). This strongly suggests that PrPC is essential for REST-dependent NMDAR expression, especially for GluN2B. Additionally, the number of normal mitochondria, mediated by the neuroprotective function of REST24, decreased (Fig. 4e, f). Even though REST was overexpressed, compared with the WT group, the mitochondrial number decreased by 7.92% (Fig. 4e, f) in the Prnp0/0 group, implying that the presence of PrPC might have an effect on the REST-mediated density of mitochondria.

Fig. 4: PrPC is essential for REST-regulated expression of NMDARs and mitochondrial numbers.

Prnp+/+ a and Prnp0/0 c mice primary hippocampal neurons were transfected with a control HA vector, or REST shRNA (REST knockdown) or REST-HA vector (REST overexpression). Cellular proteins were immunoblotted for REST, GluN2B, GluN2A and NMDAR1. b and d quantitative analyses of a and c, respectively. Immunoblot density in b and d were normalized to GAPDH and expressed as the relative density to the HA vector (dashed lines) in each group. *P < 0.05; **P < 0.01; ***P < 0.001 in (b) vs the HA vector group. e Confocal immunofluorescence labeling for total REST (green), including exogenous (overexpression of REST) and endogenous REST, and mitochondria (MitoTracker Red) in Prnp+/+ and Prnp0/0 hippocampal neurons. Nuclei (blue) were stained with DAPI. Scale bars = 10 μm. f Quantification of mitochondrial number (e) in a segment of neuronal process 200 μm length beginning from the cell body of neurons (n = 30). Experiments were performed in triplicate. **P < 0.01; ***P < 0.001 vs Prnp+/+

REST effects on mitochondria partially depend on PrPC

It has been shown that Aβ induces an extrasynaptic NMDAR-dependent increase in nitric oxide, leading to mitochondrial dysfunction and synapse deficiency27,28. Overexpression of REST in primary cortical neurons alleviates neurotoxicity peptide (PrP106-126)-induced neuronal oxidative stress and mitochondrial damage24. Importantly, loss of PrPC results in decreased mitochondrial numbers and abnormal mitochondrial morphology29. In light of these previous studies, we further explored the potential relationship of PrPC with REST-regulated mitochondrial numbers and distribution.

The NMDAR antagonist dizocilpine maleate (MK-801) improves mitochondrial function and energy status30, linking NMDAR activation and mitochondrial function31. WT and Prnp0/0 hippocampal neurons were treated with either NMDA (10 μM, 24 h) or MK-801 (Dizocilpine) (10 μM, 24 h) to observe the density and distribution of mitochondria (Fig. 5a–c). In both the WT and Prnp0/0 groups, NMDA significantly decreased mitochondrial numbers by 60.49% and 63.01%, respectively, compared with controls. MK-801 markedly increased mitochondrial numbers in the WT and Prnp0/0 groups by 1.61-fold and 1.53-fold, respectively, relative to corresponding control.

Fig. 5: PrPC effects on the density and distribution of mitochondria.

a Following treatment with NMDA or MK-801 (or no treatment), Prnp+/+ and Prnp0/0 mice primary hippocampal neurons were stained with MitoTracker Red to visualize mitochondria and analyzed by fluorescence microscopy. Nuclei (blue) are stained with DAPI. Scale bars = 10 μm. b Quantification of mitochondrial number (a) in a segment of neuronal process 200 μm in length beginning from the cell body of neurons (n = 30). #P < 0.05; ##P < 0.01; ***P < 0.001 vs corresponding control. c Quantification analysis of (a) to confirm the number of cells harboring perinuclearly clustered mitochondria. For this quantification, mitochondria of at least 500 cells per experiment were determined in a blinded manner. Quantifications were based on triplicates of at least three independent experiments. **P < 0.01 vs corresponding control. d Co-transfection of PrPC and REST rescued the density and distribution of mitochondria. Scale bars = 10 μm. e Quantification of mitochondrial number (d) in a segment of neuronal process 200 μm in length beginning from the cell body of neurons (n = 30). **P < 0.01 vs Prnp+/+. #P < 0.05 is Prnp0/0+ HA-PrP vs Prnp0/0. f Quantification analysis of (d) to confirm the number of cells harboring perinuclearly clustered mitochondria. For this quantification, mitochondria of at least 500 cells per experiment were determined in a blinded manner. Quantifications were based on triplicates of at least three independent experiments. *P < 0.05 vs Prnp+/+. #P < 0.05 is Prnp0/0+ HA-PrP vs Prnp0/0

Exposure of hippocampal neurons to NMDA significantly increased the number of cells harboring clustered perinuclear mitochondria in both WT and Prnp0/0 groups, whereas the number in the latter group increased slightly. As overexpression of REST had a diminished effect on mitochondrial numbers in the Prnp0/0 group compared with the WT group, (mitochondrial number decreased by 7.92%), through a PrPC rescue experiment we directly explored the role of PrPC in REST-regulated mitochondria. A REST-HA vector was co-transfected with a HA vector into WT or Prnp0/0 hippocampal neurons, or co-transfected with a PrP-HA vector into Prnp0/0 hippocampal neurons. As expected, co-overexpression of PrPC and REST in Prnp0/0 cultures partially restored mitochondrial numbers compared with the Prnp0/0 group transfected with REST alone (Fig. 5d, e). Moreover, the number of cells harboring clustered perinuclear mitochondria also significantly decreased when PrPC and REST were co-transfected into Prnp0/0 mice (Fig. 5d, f).

PrPC plays a critical role in REST-regulated GluN2A and GluN2B expression

Previous reports have demonstrated that native PrPC mediates an important neuroprotective role through its ability to inhibit NR2D subunits16. Although NR2B subunits did not immunoprecipitate with PrPC in that study, PrPC may interact with NMADRs subunits through other and/or indirect ways, an interaction that needs further exploration. So, to examine whether there is an association between NR2 subunits and PrPC, we analyzed the surface expression of Glu2A, Glu2B, and in WT hippocampal neurons using immunolabel reactivity (Fig. 6a). Although Glu2A expression was slight, Glu2B and PrPC could well localize in the place where Glu2A existed (Fig. 6b). We then stained total Glu2A and total Glu2B along dendritic processes in WT and Prnp0/0 neurons. The specific dendritic marker, microtubule-associated protein (MAP2) was used to reveal dendrites by IF. Glu2A-positive and Glu2B-positive puncta along processes were quantitatively analyzed. PrPC rescue experiments were also used to explore the function of PrPC by IF and WB. First, WT cultures were transfected with three different short hairpin RNA (shRNA)-mediated PrP knockdown vectors (Prnp+/+ + Sh-PrP) and compared with the PrP-null cultures (Prnp0/0) to exclude the interfering factor of genetic background in the knockout mice. In a second experiment, PrP-null cultures were transfected with HA-PrP vector (Prnp0/0+ HA-PrP) to observe the function of full-length PrPC. Third, PrP-null cultures were transfected with a WT PrPC vector32(Prnp0/0+ WT) and compared with the HA-PrP vector group to remove the potential influence of the HA tag to the location of PrPC and to further confirm the function of full-length PrPC. In a final experiment, PrP-null cultures were transfected with a PrPC mutant (D177N point mutation)32 [Prnp0/0+  PrP(D177N)] that is less efficiently trafficked to the surface than the WT PrP and accumulates in the cytoplasm even without proteasome inhibition. These cultures were compared with the WT PrPC to further confirm the role of full-length PrPC. Remarkably, both in the Prnp+/+ + Sh-PrP group and in the Prnp0/0 group, Glu2A puncta significantly decreased by 37.70% and 52.17% relative to the WT group (23.07 ± 0.05 per 100 μm of dendrites), respectively. Conversely, Glu2B was present in more continuously abundant puncta, increased 1.25-fold (Prnp+/+ + Sh-PrP) and 1.30-fold (Prnp0/0) over the WT group (33.67 ± 0.06 per 100 μm of dendrites) (Fig. 6c, d). On the other hand, overexpression of HA-PrP vector or WT PrP vector had similar effects. They both partly restored the densities of GluN2A puncta (Prnp0/0 versus Prnp0/0 + HA-PrP: 11.01 ± 0.07 versus 20.33 ± 0.09 per 100 μm of dendrites; Prnp0/0 versus Prnp0/0+  WT PrP: 11.01 ± 0.07 versus 23.03 ± 0.05 per 100 μm of dendrites) and reduced the densities of GluN2B puncta (Prnp0/0 versus Prnp0/0 + HA-PrP: 43.67 ± 0.04 versus 35.07 ± 0.07 per 100 μm of dendrites; Prnp0/0 versus Prnp0/0+ WT PrP: 43.67 ± 0.04 versus 34.33 ± 0.05 per 100 μm of dendrites). However, the D177N mutant had no significant effect on the expression of GluN2A (12.33 ± 0.07 per 100 μm of dendrites) or GluN2B (42.83 ± 0.03 per 100 μm of dendrites) (Fig. 6f, g). Total neurite length (the cumulative length of all neurites of a single neuron) was not significantly different in the groups tested (n=30 randomly selected neurons per coverslip within each experiment. Each experiment was repeated with, at least, three independent cultures). The average length of neurites in each group is shown in Figure. 6e, h.

Fig. 6: PrPC plays a critical role in REST-regulated expression levels of GluN2A and GluN2B.

a Triple-labeled WT mouse hippocampal neurons for GluN2A (green), GluN2B (red) and PrPC (blue) along dendritic processes. The arrowheads highlight examples of clear colocalization of GluN2A (green), GluN2B (red), and PrPC (blue). Scale bars = 10 μm. The yellow line in the top panel of (b) indicates the position in (a) of the line scan shown in the bottom panel of (b). c-e Knockdown of PrPC disturbed the density of GluN2A and GluN2B. fh Overexpression of full-length PrPC restores Prnp0/0 induced disorder of GluN2A and GluN2B. c, f Microtubule-associated protein (MAP2) is the specific dendritic maker. GluN2A (green) protein puncta along processes (MAP2, magenta) and GluN2B (red) protein puncta along processes (MAP2, green) in neurons of Prnp+/+, Prnp+/+ + Sh-PrP and Prnp0/0, or Prnp0/0 culture transfected with HA-PrP vector, wild-type PrP vector or PrP(D177N) vector. Scale bars = 10 μm. d, g Quantifications of GluN2A and GluN2B puncta density are normalized to 100 μm neurite length, as assessed by MAP2 staining. In (d), *P < 0.05; **P < 0.01 is Prnp+/+ + Sh-PrP vs Prnp+/+, #P < 0.05; ##P < 0.01 is Prnp0/0 vs Prnp+/+. In (g), *P < 0.05 is Prnp0/0+PrP(D117N) vs Prnp+/+, #P < 0.05; ##P < 0.01 is Prnp0/0+PrP(D117N) vs Prnp0/0+wild-type PrP. e, h Average neurite length for each randomly tested neuron in each group (n = 30). i Immunoblot of REST, GluN2A, and GluN2B in Prnp+/+ neurons pre-transfected with negative control vector (NC) or three separate PrPC shRNAs (Sh-PrP; each lane represents a different shRNA). l Quantitative analyses of (i). j Immunoblot of REST, GluN2A, and GluN2B in Prnp0/0 neurons pre-transfected with HA vector or HA-PrP vector. m Quantitative analyses of (j). k Immunoblot of REST, GluN2A and GluN2B in Prnp0/0 neurons pre-transfected with HA-PrP vector, or wild-type PrP vector or PrP(D117N) mutant vector. n Quantitative analyses of (k). Immunoblot density normalized to GAPDH and expressed as a relative density to the control group in (i), or to the HA vector-transfected group in (j), or to the HA-PrP vector-transfected group in (k). Data are presented as means ± SD in triplicate experiments. In (l), **P < 0.01; ***P < 0.001 vs control. In (m), **P < 0.01; ***P < 0.001 vs HA vector-transfected group. In (n), *P < 0.01; **P < 0.01 vs HA-PrP vector-transfected group

Consistently, quantitative WB analysis revealed the following. (1) Knockdown of PrPC in WT culture remarkably induced the expression of REST (increased to 17.14-fold of the Prnp0/0 group), stimulated the level of GluN2B (increased to 20.68-fold of the Prnp0/0 group) and suppressed the level of GluN2A (decreased to 63.54% of the Prnp0/0 group). However, NMDAR1 expression was not significantly different, and negative control vector (NC) showed similar results to control group (Fig. 6i, l). (2) PrPC overexpression significantly inhibited the expression of REST (decreased to 21.94% of the Prnp0/0 group), contributing to the suppression of GluN2B (decreased to 39.50% of the Prnp0/0 group) and the promotion of GluN2A (increased to 2.33-fold of the PrP-null group). NMDAR1 expression was not significantly different in the two groups (Fig. 6j, m). (3) Both HA-PrP vector and WT PrP vector had similar degrees of functional action to REST, GluN2B and GluN2A. However, PrP (D177N) mutant failed to recover the expression of these proteins in Prnp0/0 culture (Fig. 6k, n). Overall, our data indicate a critical and novel relationship between full-length PrPC, GluN2A and GluN2B. This is mediated by REST to maintain the correct proportions of NMDAR subunits.


The importance of a developmental switch maintaining the correct ratio of GluN2B-to-GluN2A

NMDARs are necessary regulators of brain plasticity1. During the development of synapse structure and function, they transform precise patterns of neuronal activity into long-term changes that are thought to underlie higher cognitive functions33. Moreover, NMDARs switch their subunits from predominantly GluN2B to primarily GluN2A during early postnatal development34. This subunit switch is evolutionarily conserved from amphibians to mammals and occurs all over the CNS during a time window coinciding with synapse growth and neuronal circuitry refinement1,33. However, the mechanisms responsible for the GluN2B-to-GluN2A subunit exchange have yet to be fully defined12. In neurodegenerative diseases, recent studies also indicate that GluN2B-containing NMDARs have an essential role in mediating the adverse effects of Aβ27. GluN2B antagonists can rescue Aβ-induced damage of long-term potentiation and synaptic impairment35. However, all of the clinical trials of first-generation NMDAR antagonists were disappointing as a result of unendurable side-effects and short therapeutic windows36. Another potential limitation is the lack of subunit selectivity of the drugs2,35. GluN2B-specific antagonists offer significant neuroprotection with a better side-effect profile. Conversely, activation of GluN2A may exhibit pro-survival effects37 via CREB signaling, although the neuroprotective role of GluN2A is still controversial38,39. Additionally, in patients with Parkinson’s disease, the degeneration of nigral dopaminergic neurons gives rise to overactivation of glutamatergic projections40,41. In the striatal membrane of l-3,4-dihydroxyphenylalanine (l-DOPA)-treated dyskinesia animals, the increased synaptic abundance of GluN2A and redistribution of GluN2B from synaptic to extrasynaptic regions demonstrate that selectively targeting specific NMDARs might be more hopeful42. Understanding the precise molecular mechanisms responsible for the GluN2B-to-GluN2A exchange could provide new perspectives for the development of therapeutic strategies.

PrPC plays a novel and critical role in the REST-dependent developmental switch in synaptic NMDA receptors

During embryogenesis in pluripotent stem cells and neural progenitors, REST is a widely expressed gene-silencing factor. REST plays an important role for synaptic function via epigenetic remodeling by silencing coding and noncoding neuronal genes10. At the gene level, REST takes part in the postnatal switch in synaptic NMDARs by reducing GluN2B expression through epigenetic remodeling of Grin2b43. Some questions remain unanswered, such as what mechanism turns on REST expression in differentiated neurons, and which factor regulates the long-term increase in GluN2A expression during postnatal development.

Mature PrPC is a glycoprotein attached by a carboxyl(C)-terminal glycosylphosphatidylinositol (GPI) anchor to the extracellular leaflet of the plasma membrane44. PrPC can undergo different types of physiological cleavage, producing N2 and C2 fragments13. PrPC contacts with, and signals through, multiple cell surface proteins and signaling pathways. This highlights the need for a better understanding of the mechanisms of PrPC interaction with its binding factors, both physiologically and pathologically44. Comparing WT and Prnp0/0 revealed that PrPC expression at synapses contributes to hippocampal synaptic function and exhibits neuroprotection by regulating neuronal excitability. In particular, PrPC requires copper to facilitate S-nitrosylation-mediated NMDAR suppression17. However, GluN2B subunits did not immunoprecipitate with PrPC in a previous study16 and no studies have yet explored the relationship between PrPC and GluN2A. PrPC might interact with GluN2B through indirect ways, an aspect needing further research. Nucleocytoplasmic transport is thought to be important for REST as a transcriptional repressor regulating neuronal gene expression22,25. In this study, nuclear REST was markedly lost in the absence of PrPC, demonstrating that the functional translocation of REST in response to NMDA and LiCl treatment depends on the presence of PrPC. Overexpression of exogenous PrPC rescued excessive REST induced GluN2A and GluN2B imbalance by increasing the expression of GluN2A and inhibiting the level of GluN2B in PrP-null neurons. To our knowledge, our study is the first to demonstrate that PrPC-mediated REST-dependent expression of GluN2A and GluN2B contributes to maintain the correct subunit proportions of NMDARs and cellular homeostasis, deserving to be further explored as a novel and viable therapeutic target against dysfunctions of glutamatergic transmission.

Materials and methods


PrnP0/0 mice on a C57BL/6J × 129Sv genetic background were kindly supplied by Dr. Charles Weissmann45,46. Control C57BL/6J mice with no genomic modifications (WT), 6–8 weeks of age and weighing 18–25 g, were obtained from Beijing Experimental Animal Center. Mice were housed 3–4 per cage and under controlled environmental conditions (21–23 °C; 40–60 % humidity; 12-h light/dark cycle) with free access to water and standard pelleted food. All animal experiments were conducted in accordance with the guidelines of Beijing Municipality on the Review of Welfare and Ethics of Laboratory Animals and approved by the Beijing Municipality Administration Office of Laboratory Animals (BAOLA).

Plasmids and transfection

The pCMV-HA-Rest vector of full-length REST (cat no. PPL50007-2a)24,25 was obtained from Public Protein/Plasmid Library (Nanjing, China, GeneShare Technology, Co, Ltd). The pGPH1/GFP/Neo-REST-Mus shRNA vectors24,25, the pGPH1/GFP/Neo-PRNP-Mus shRNA vectors (Table 1) and the negative control vector (NC)—which has the same number of corresponding bases as shRNA-REST or shRNA-PRNP but does not target any known gene—were all obtained from GenePharma (Suzhou, China). The full-length mouse prion protein (PRNP) complementary DNA (cDNA) was originally cloned from total brain of C57BL/6J mice by PCR. The primers were designed according to the gene sequence of Mus musculus prion protein (PRNP) mRNA in GenBank (Gene ID: 19122). Oligonucleotides 5′-CGGAATTCATGGCGAACCTTGGCTACTG-3′ (forward) and 5’-GCCTCGAGTCACATGTGCTTCATGTTGGTTTTTCCCACGATCAGGAAGATGA-3′ (reverse) were used to introduce EcoRI and XhoI sites flanking the cDNA (The underlined alphabets denote the "Restriction enzyme cutting sites" of EcoRI and XhoI, respectively.). The PCR product was cloned into vector pCMV-HA (Clontech, Kyoto, Japan) to generate plasmid pCMV-HA-PRNP (HA-PrP) by standard molecular biology techniques and confirmed by sequencing. All the primers were synthesized by Sangon Company (Shanghai, China). Full-length mouse PrP with 3F4 epitope (WT PrPC) (Addgene plasmid # 13917) and moPrP(3F4) D177N (Addgene plasmid # 1319) were gifts from Susan Lindquist32. For transfection23,24,47,48, cultured primary hippocampal neurons were washed with Opti-MEM (Gibco) and then transfected with the appropriate plasmids using the LipofectamineTM Ltx & PlusTM Reagent (Thermo Fisher, Waltham, MA, USA) in Opti-MEM according to the manufacturer’s instructions. The amounts of plasmids and reagents were 2 μg and 3 μl per 12-well, respectively. Forty-eight hours after transfection, cells were observed using an upright fluorescence confocal microscopy (Olympus, Tokyo, Japan) or subjected to immunoblot analyses.

Table 1 Sequences of REST-targeting shRNA(Sh-REST), PRNP-targeting shRNA(Sh-PrP)

Statistical analysis

All assays were performed at least three times. Data are expressed as means ± S.D. For in vivo experiments, six hippocampal homogenates were separately collected and tested in each group. One-way ANOVA, followed by Dunnett’s test was performed for whole hippocampal lysate experiments in Figure 111. Two-way ANOVA was performed for Figure 2. Other comparisons for parametric data were made using Student’s test or one-way ANOVA followed by post hoc Turkey’s test or two-way ANOVA test using the SPSS software (version 13.0: SPSS Inc., Chicago, IL, USA), GraphPad Prism 5 software (La Jolla, CA, USA) and Image J (National Institutes of Health, USA). P<0.05 was considered statistically significant24,25.


  1. 1.

    Paoletti, P., Bellone, C. & Zhou, Q. NMDA receptor subunit diversity: impact on receptor properties, synaptic plasticity and disease. Nat. Rev. Neurosci. 14, 383–400 (2013).

  2. 2.

    Traynelis, S. F. et al. Glutamate receptor ion channels: structure, regulation, and function. Pharmacol. Rev. 62, 405–496 (2010).

  3. 3.

    Paoletti, P. Molecular basis of NMDA receptor functional diversity. Eur. J. Neurosci. 33, 1351–1365 (2011).

  4. 4.

    Monyer, H., Burnashev, N., Laurie, D. J., Sakmann, B. & Seeburg, P. H. Developmental and regional expression in the rat brain and functional properties of four NMDA receptors. Neuron 12, 529–540 (1994).

  5. 5.

    Dumas, T. C. Developmental regulation of cognitive abilities: modified composition of a molecular switch turns on associative learning. Prog. Neurobiol. 76, 189–211 (2005).

  6. 6.

    Espinosa, J. S., Wheeler, D. G., Tsien, R. W. & Luo, L. Uncoupling dendrite growth and patterning: single-cell knockout analysis of NMDA receptor 2B. Neuron 62, 205–217 (2009).

  7. 7.

    Von Engelhardt, J. et al. Contribution of hippocampal and extra-hippocampal NR2B-containing NMDA receptors to performance on spatial learning tasks. Neuron 60, 846–860 (2008).

  8. 8.

    Gray, J. A. et al. Distinct modes of AMPA receptor suppression at developing synapses by GluN2A and GluN2B: single-cell NMDA receptor subunit deletion in vivo. Neuron 71, 1085–1101 (2011).

  9. 9.

    Lee, M., Yasuda, R. & Ehlers, M. D. Metaplasticity at single glutamatergic synapses. Neuron 66, 859–870 (2010).

  10. 10.

    Song, Z., Zhao, D., Zhao, H. & Yang, L. NRSF: an angel or a devil in neurogenesis and neurological diseases. J. Mol. Neurosci. 56, 131–144 (2015).

  11. 11.

    Rodenas-Ruano, A., Chávez, A. E., Cossio, M. J., Castillo, P. E. & Zukin, R. S. REST-dependent epigenetic remodeling promotes the developmental switch in synaptic NMDA receptors. Nat. Neurosci. 15, 1382–1390 (2012).

  12. 12.

    Parsons, M. P. & Raymond, L. A. Extrasynaptic NMDA receptor involvement in central nervous system disorders. Neuron 82, 279–293 (2014).

  13. 13.

    Black, Stefanie A. G. et al. Cellular Prion Protein and NMDA Receptor Modulation: Protecting against Excitotoxicity. Frontiers in Cell Dev. Biol. 2, article 45 (2014); e-pub ahead of print 27 April 2018.

  14. 14.

    Song, Z., Zhao, D. & Yang, L. Molecular mechanisms of neurodegeneration mediated by dysfunctional subcellular organelles in transmissible spongiform encephalopathies. Acta Biochem. Biophy. Sin. 45, 452–464 (2013).

  15. 15.

    Soto, C. & Satani, N. The intricate mechanisms of neurodegeneration in prion diseases. Trends Mol. Med. 17, 14–24 (2011).

  16. 16.

    Khosravani, H. et al. Prion protein attenuates excitotoxicity by inhibiting NMDA receptors. J. Cell Biol. 181, 551–565 (2008).

  17. 17.

    Gasperini, L., Meneghetti, E., Pastore, B., Benetti, F. & Legname, G. Prion protein and copper cooperatively protect neurons by modulating NMDA receptor through S-nitrosylation. Antioxid. Redox Sign. 22, 772–784 (2015).

  18. 18.

    Borrelli, E., Nestler, E. J., Allis, C. D. & Sassone-Corsi, P. Decoding the epigenetic language of neuronal plasticity. Neuron 60, 961–974 (2008).

  19. 19.

    Levenson, J. M. & Sweatt, J. D. Epigenetic mechanisms in memory formation. Nat. Rev. Neurosci. 6, 108–118 (2005).

  20. 20.

    Tremblay, P., Bouzamondo-Bernstein, E., Heinrich, C., Prusiner, S. B. & DeArmond, S. J. Developmental expression of PrP in the post-implantation embryo. Brain Res. 1139, 60–67 (2007).

  21. 21.

    Salès, N. et al. Developmental expression of the cellular prion protein in elongating axons. Eur. J. Neurosci. 15, 1163–1177 (2002).

  22. 22.

    Lu, T. et al. REST and stress resistance in ageing and Alzheimer’s disease. Nature 507, 448–454 (2014).

  23. 23.

    Song, Z. et al. REST alleviates neurotoxic prion peptide-induced synaptic abnormalities, neurofibrillary degeneration and neuronal death partially via LRP6-mediated Wnt-beta-catenin signaling. Oncotarget 7, 12035–12052 (2016).

  24. 24.

    Song, Z. et al. Downregulation of the repressor element 1-silencing transcription factor (REST) is associated with Akt-mTOR and Wnt-β-catenin signaling in prion diseases models. Front. Mol. Neurosci. 10, 128 (2017).

  25. 25.

    Song Z., Yang W., Zhou X., Yang L. & Zhao D. Lithium alleviates neurotoxic prion peptide-induced synaptic damage and neuronal death partially by the upregulation of nuclear target REST and the restoration of Wnt signaling. Neuropharmacology 123, 332–348 (2017).

  26. 26.

    Kaneko, N., Hwang, J., Gertner, M., Pontarelli, F. & Zukin, R. S. Casein kinase 1 suppresses activation of REST in insulted hippocampal neurons and Halts ischemia-induced neuronal death. J. Neurosci. 34, 6030 (2014).

  27. 27.

    Li, S. et al. Soluble Aβ oligomers inhibit long-term potentiation through a mechanism involving excessive activation of extrasynaptic NR2B-containing NMDA receptors. J. Neurosci. 31, 6627–6638 (2011).

  28. 28.

    Snyder, E. M. et al. Regulation of NMDA receptor trafficking by amyloid-β. Nat. Neurosci. 8, 1051–1058 (2005).

  29. 29.

    Kim, B. et al. The cellular prion protein (PrPC) prevents apoptotic neuronal cell death and mitochondrial dysfunction induced by serum deprivation. Mol. Brain Res. 124, 40–50 (2004).

  30. 30.

    Gilland, E., Puka-Sundvall, M., Hillered, L. & Hagberg, H. Mitochondrial function and energy metabolism after hypoxia—ischemia in the immature rat brain: involvement of NMDA-receptors. J. Cereb. Blood F. Met. 18, 297–304 (1998).

  31. 31.

    McKay, S., Bengtson, C. P., Bading, H., Wyllie, D. J. & Hardingham, G. E. Recovery of NMDA receptor currents from MK-801 blockade is accelerated by Mg2+ and memantine under conditions of agonist exposure. Neuropharmacology 74, 119–125 (2013).

  32. 32.

    Ma, J. & Lindquist, S. Wild-type PrP and a mutant associated with prion disease are subject to retrograde transport and proteasome degradation. Proc. Natl. Acad. Sci. USA 98, 14955–14960 (2001).

  33. 33.

    Yashiro, K. & Philpot, B. D. Regulation of NMDA receptor subunit expression and its implications for LTD, LTP, and metaplasticity. Neuropharmacology 55, 1081–1094 (2008).

  34. 34.

    Xu, Z. et al. Metaplastic regulation of long-term potentiation/long-term depression threshold by activity-dependent changes of NR2A/NR2B ratio. J. Neurosci. 29, 8764–8773 (2009).

  35. 35.

    Talantova, M. et al. Aβ induces astrocytic glutamate release, extrasynaptic NMDA receptor activation, and synaptic loss. Proc. Natl. Acad. Sci. USA 110, E2518–E2527 (2013).

  36. 36.

    Mony, L., Kew, J. N., Gunthorpe, M. J. & Paoletti, P. Allosteric modulators of NR2B-containing NMDA receptors: molecular mechanisms and therapeutic potential. Br. J. Pharmacol. 157, 1301–1317 (2009).

  37. 37.

    Von Engelhardt, J. et al. Excitotoxicity in vitro by NR2A-and NR2B-containing NMDA receptors. Neuropharmacology 53, 10–17 (2007).

  38. 38.

    Liu, Y. et al. NMDA receptor subunits have differential roles in mediating excitotoxic neuronal death both in vitro and in vivo. J. Neurosci. 27, 2846–2857 (2007).

  39. 39.

    Terasaki, Y. et al. Activation of NR2A receptors induces ischemic tolerance through CREB signaling. J. Cereb. Blood F. Met. 30, 1441–1449 (2010).

  40. 40.

    Sgambato-Faure, V. & Cenci, M. A. Glutamatergic mechanisms in the dyskinesias induced by pharmacological dopamine replacement and deep brain stimulation for the treatment of Parkinson’s disease. Prog. Neurobiol. 96, 69–86 (2012).

  41. 41.

    Hallett, P. J. et al. Alterations of striatal NMDA receptor subunits associated with the development of dyskinesia in the MPTP-lesioned primate model of Parkinson’s disease. Neuropharmacology 48, 503–516 (2005).

  42. 42.

    Gardoni, F. et al. Targeting NR2A-containing NMDA receptors reduces L-DOPA-induced dyskinesias. Neurobiol. Aging 33, 2138–2144 (2012).

  43. 43.

    Rodenas-Ruano, A., Chavez, A. E., Cossio, M. J., Castillo, P. E. & Zukin, R. S. REST-dependent epigenetic remodeling promotes the developmental switch in synaptic NMDA receptors. Nat. Neurosci. 15, 1382–1390 (2012).

  44. 44.

    Song, Z., Zhao, D. & Yang, L. Metabolism of minor isoforms of prion proteins: cytosolic prion protein and transmembrane prion protein. Neural. Regen. Res. 8, 2868 (2013).

  45. 45.

    Weissmann, C. & Flechsig, E. PrP knock-out and PrP transgenic mice in prion research. Br. Med. Bull. 66, 43–60 (2003).

  46. 46.

    Liu, J. et al. Prion protein participates in the protection of mice from lipopolysaccharide infection by regulating the inflammatory process. J. Mol. Neurosci. 55, 279–287 (2015).

  47. 47.

    Zhu T. et al. HDAC6 alleviates prion peptide-mediated neuronal death via modulating PI3K-Akt-mTOR pathway. Neurobiol. Aging 37, 91–102 (2015).

  48. 48.

    Song, Z. et al. Overexpression of BAT3 alleviates prion protein fragment PrP106-126-induced neuronal apoptosis. CNS. Neurosci. Ther. 20, 737–747 (2014).

Download references


We thank Dr. Yongyong Cui, Dr. Ali, Dr. Chaosi Li, Dr. Ruichao Yue and Dr. Chunfa Liu for their valuable suggestions and critical reading of this manuscript. This work was supported by the Natural Science Foundation of China (Project no. 31472166) and the Foundation of Chinese Ministry of Science and Technology (Project no. 2015BAI07B02).

Author information

Conceived and designed the experiments: Z.Q.S. Performed the experiments: Z.Q.S. and W.Y. Analyzed the data: Z.Q.S. Wrote the paper: Z.Q.S. Edited the paper: Z.Q.S., W.Y., G.Y.C., X.M.Z., L.F.Y and D.M.Z. Contributed reagents/materials/analysis tools: D.M.Z. All authors read and approved the final manuscript.

Correspondence to Deming Zhao.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Further reading