|
FIGURE 1. Functional characterization of M4-deleted and M4-co-expressed GluN1/GluN2A NMDA receptors. (A) Structure of the TMD region of a GluN1 (yellow)/GluN2A (red) NMDAR showing the peripheral localization of the M4 compared to M1, M2 and M3 of its own subunit. Structure in the middle and right show the M4-truncated receptor (GluN1ΔM4/GluN2A) and exemplary M4-truncated receptor with the separate M4 segment (GluN1ΔM4+M4N1/GluN2A). Structural analysis was performed from PDB 5UP2 using UCSF Chimera. (B) Agonist response properties of wt GluN1/GluN2A receptor (left), nonfunctional truncated GluN1ΔM4/GluN2A (middle), and GluN1ΔM4+M4N1/GluN2A receptor (right) rescued by M4 segment co-expression recorded by superfusion of increasing concentrations of glutamate in the presence of glycine 2–3 days after injection. (C) Glutamate or glycine dose response curves for wt GluN1/GluN2A and GluN1ΔM4+M4N1/GluN2A NMDARs in the presence of glycine (left) or in the presence of glutamate (right). Data represent mean ± SEM values. (D) Representative TEVC recordings of GluN1/GluN2AΔM4+M4N2A (left) and GluN1/GluN2AΔM4+M4N1 (middle) show rescued functionality of GluN2A-M4 truncated receptors with M4N1 or M4N2A. No measurable currents could be obtained for GluN1/GluN2AΔM4+M4N3A (right), highlighting the differences between the M4 segments of GluN1, GluN2A, and GluN3A. Right panel shows Imax currents of co-expression of M4-deleted GluN1/GluN2A NMDARs in the presence of M4 fragments (n = 5–12). Data represent mean values ± SEM.
|
|
FIGURE 2. Influence of M4 on the surface expression of NMDARs. (A) Representative images of GluN1/GluN3A, nonfunctional GluN1ΔM4/GluN3A, and GluN1ΔM4+M4N1/GluN3A receptor combinations. M4N1 failed to rescue functionality in M4-truncated GluN1/GluN3A receptors. (B) SDS-PAGE of metabolic (35S) methionine-tagged GluN1/GluN3A and GluN1ΔM4/GluN3A receptors with simultaneous purification of C-terminal His-tagged receptor subunit constructs by metal affinity chromatography (His-tag purification applies to B–E.). Both the GluN1/GluN3A and GluN1ΔM4/GluN3A receptors were correctly expressed. (C, D) SDS-PAGE of metabolic (35S) methionine-labeled GluN1/GluN3A receptor combinations. Glycosylation status was determined by PNGaseF or EndoH treatment and showed that GluN1ΔM4/GluN3A and GluN1/GluN3AΔM4 were N-glycosylated, indicating cell surface localization. Co-expression of the respective M4 segment resulted in impaired receptor expression or biogenesis in the ER. (E) SDS-PAGE of Cy5 surface-stained GluN1/GluN3A and GluN1ΔM4/GluN3A and GluN1ΔM4+M4N1/GluN3A receptors (left) and GluN1/GluN3A and GluN1ΔM4/GluN3A and GluN1ΔM4+M4N3A/GluN3A receptors (right) purified via His-tag with/without PNGase F or EndoH treatment. Only coexpression of the M4N1 segment resulted in loss of surface expression. (F–H) SDS-PAGE was performed with the surface proteins isolated by biotin affinity purification. Western blot performed with a primary antibody against the GluN1-NTD shows surface-expressed GluN1/GluN3A and GluN1ΔM4/GluN3A receptors and a complete loss of surface expression for the GluN1ΔM4+M4N1/GluN3A co-expressed with the M4 segment. (G) Western blot performed with a primary antibody against the GluN1-NTD shows correct surface expression for GluN1/GluN2A, GluN1ΔM4/GluN2A, and GluN1ΔM4+M4N1/GluN2A without affecting M4 segment co-expression (left) and for GluN1ΔM4/GluN2AΔM4 (right). (H) Western blot analysis using a primary antibody against the GluN1 CTD showing that the M4 segment was well expressed at the cell surface when GluN1ΔM4+M4N1/GluN2A and M4N1 were expressed alone. The GluN1/GluN2ΔM4 receptor was also well expressed at the surface.
|
|
FIGURE 3. Examination of the possible interaction residues of GluN1-M4 and GluN2A-M1/M3. (A) Sequence alignment of the human GluN1, GluN2 and GluN3 M4 segments. Sequences were taken from Uniprot and sequence alignment was performed using the Multiple Sequence Alignment Tool from EMBL-EBI. The residues identified as possible attachment sites, M813 and F817, are highlighted in yellow. (B) Representative TEVC images of GluN1ΔM4+M4N1/GluN2A, GluN1ΔM4+M4N1-M813A/GluN2A, and GluN1ΔM4+M4N1-F817A/GluN2A show the impaired rescue effect of the two point mutations. M4N1-M813A resulted in a complete loss of rescue effect, while M4N1-F817A resulted in a ∼90% decrease in rescued current. (C) Western blot of surface proteins isolated by biotinylation, with a primary antibody against the GluN1 CTD, showing that M4N1, M4N1-M813A, and M4N1-F817A are all well expressed at the cell surface, ruling out that impaired expression is responsible for the loss of the rescue effect. (D) Representation of GluN1-M4 (yellow) and GluN2A-M1/M3 (gray). Structural analysis was performed using UCSF Chimera with the GluN1/GluN2A structure 5UP2. Amino acids M813 and F817 are colored green, and potential interaction partners M1-M560 (partner of F817) and M3-F637 (partner of M813) are highlighted in red. (E) Quantification of the rescue effect of alanine substitutions normalized to the unmutated M4 segment in co-expressed M4N1 showed that for M4 mutations M818A and L819A, in contrast to M4N1-M813A and M4N1-F817A, there was no effect on the rescued currents. The GluN2AM560A mutation also showed no effect on the rescue effect, suggesting a stronger influence of the F817-M560 interaction. GluN2AF637A showed a strong decrease in rescued Imax currents by ∼50%, highlighting the importance of the M813-F637 interaction for M4 binding to the truncated core receptor. Data were analyzed by one-way ANOVA followed by Dunnett´s multiple comparisons test. Number of experiments is given in brackets. (F) Western blot of isolated surface proteins, with a primary antibody against the GluN1-CTD. Samples were equally divided and loaded untreated and DTT-treated. The potential interaction partners M813-F637 and F817-M560 were mutated to cysteines to confirm close proximity. GluN1ΔM4+M4N1-M813C/GluN2AF637C and GluN1ΔM4+M4N1-F817C/GluN2AM560C both showed both a ∼180 kDa and a ∼15 kDa band under denaturing conditions without DTT. In both cases, the 180 kDa band disappeared after DTT treatment with an increase in the 15 kDa band, indicating a disulfide bond between the M4M813C (or F817C) with the GluN2AF637C (or M560C). The GluN1ΔM4+M4N1-F817C/GluN2AF637C sample loaded as a control did not result in a band around the 180 kDa mark, indicating that no disulfide bond was formed here. (G) Quantification of Imax currents of wt- and alanine-substituted full-length GluN1/GluN2A receptors showing no significant decrease, indicating that the M813A, F817A, and M818A mutations do not affect receptor activity in the full-length receptor. Data were analyzed by one-way ANOVA followed by Dunnett´s multiple comparisons test. Number of experiments is given in brackets. Data represent mean ± SEM values.
|
|
FIGURE 4. Identification of a NAM recognition site of pregnenolone sulfate at GluN1-M4. (A) Representative TEVC images of Imax potentiation of PS at GluN1/GluN2A and GluN1ΔM4+M4N1/GluN2A receptors showing a strong increase in potentiating effect for M4 segment co-expression. (B) Quantification of Imax potentiation of PS at GluN1/GluN2A, GluN1ΔM4+M4N1/GluN2A, M4 segment mutants F817A, M818A, and L819A, and GluN1/GluN2AΔM4+M4N2A. Coexpression of M4N1-F817A and M4N1-L819A showed no effect on the potentiating effect compared with unmutated M4N1. Co-expression of M4N1-M818A showed a strong increase in Imax potentiation compared to GluN1ΔM4+M4N1/GluN2A [t (12) = 3.08; p = 0.0095], indicating an impaired NAM recognition site. GluN1/GluN2AΔM4+M4N2A was not different from the wt full-length receptor. Statistical analysis was performed using paired t-tests between Imax [-PS] and Imax [+PS] (*). Potentiation of the different mutants was analyzed by one-way ANOVA followed by Dunnett´s multiple comparisons test (#). (C) Dose-response analysis of PS for wt GluN1/GluN2A, GluN1ΔM4+M4N1/GluN2A, and GluN1ΔM4+M4N1-M818A/GluN2A showing nearly equal EC50 values of GluN1/GluN2A and GluN1ΔM4+M4N1/GluN2A (12.4 ± 1.3 µM, n = 5 and 15.0 ± 1.1 µM, n = 3; p < 0.05). The M4N1-M818A mutation resulted in a significant decrease in PS affinity (EC50: 23.2 ± 1.4 µM; n = 4; p < 0.001) compared to the non-mutated M4N1. (D) Quantification of Imax potentiation of PS for GluN1/GluN2A, GluN1M813A/GluN2A and GluN1M818A/GluN2A showed significantly increased potentiation for both mutants compared to wt (M813A 1.40 ± 0.09-fold; n = 9; p < 0.05 and for M818A 1.72 ± 0.19-fold; n = 12; p < 0.01 compared to wt 1.10 ± 0.09-fold; n = 8). Comparison of GluN1ΔM4+M4N1/GluN2A and GluN1ΔM4+M4N1/GluN2D showed a significant difference in PS potentiation (2.89 ± 0.2-fold, n = 8 compared to 1.45 ± 0.19-fold, n = 11; p < 0.0001). The normally inhibitory overall effect of PS on GluN1/GluN2D was converted to a potentiating one for GluN1ΔM4+M4N1/GluN2D. Statistical analysis was performed using paired t-tests between Imax [-PS] and Imax [+PS] (*). Potentiation of the different mutants was analyzed by one-way ANOVA followed by Dunnett´s multiple comparisons test (#). (E) Representative TEVC uptake of the inhibition of GluN1/GluN3A by PS. Dose-response curves of PS at GluN1/GluN3A, GluN1M813A/GluN3A, GluN1F817A/GluN3A, GluN1M818A/GluN3A showing no differences between wt and the F817A mutant, but a strong decrease in PS affinity at the M813A and M818A mutants (111 ± 11 μM, n = 4; p < 0.0001 and 19.5 ± 1.7 µM, n = 4; p < 0.0001 compared to 8.5 ± 0.9 µM, n = 7 for wt). Data represent mean ± SEM.
|
|
Fig. 1 Impact of M4-truncation and M4-segement coexpression on GluN1/GluN2A receptor expression.
SDS-PAGE of metabolic [35S]methionine-tagged GluN1/GluN2A, GluN1ΔM4/GluN2A and
GluN1ΔM4+M4N1/GluN2A receptors with purification of C-terminal His-tagged receptor subunit
constructs by metal affinity chromatography. All constructs were properly expressed, M4-Segement
coexpression did not alter GluN1ΔM4+M4N1/GluN2A expression level.
|
|
Fig.2 Impact of PS on the Glutamate Affinity.
Dose-response analysis showed that agonist affinity of GluN1/GluN2A and GluN1ΔM4+M4N1/GluN2A
were similar without PS modulation (see results Fig. 1C). PS modulation of both GluN1/GluN2A and
GluN1ΔM4+M4N1/GluN2A resulted in an increase of agonist affinity (GluN1/GluN2A [-PS] EC50:
4.2±0.47 µM to [+PS] EC50: 1.87±0.29 µM (t(5) = 7.731; p= 0.0006; and GluN1ΔM4+M4N1/GluN2A
[+PS] EC50: 1.31±0.2 µM; t(6) =11.64; p <0.0001). The results show a similar shift of the agonist affinity
for both wt and M4-segment coexpression. Statistics done by unpaired t-test. Data represent mean
±SEM.
|