Synapse-associated protein 102/dlgh3 couples the NMDA receptor to specific plasticity pathways and learning strategies

Abstract

Understanding the mechanisms whereby information encoded within patterns of action potentials is deciphered by neurons is central to cognitive psychology. The multiprotein complexes formed by NMDA receptors linked to synaptic membrane-associated guanylate kinase (MAGUK) proteins including synapse-associated protein 102 (SAP102) and other associated proteins are instrumental in these processes. Although humans with mutations in SAP102 show mental retardation, the physiological and biochemical mechanisms involved are unknown. Using SAP102 knock-out mice, we found specific impairments in synaptic plasticity induced by selective frequencies of stimulation that also required extracellular signal-regulated kinase signaling. This was paralleled by inflexibility and impairment in spatial learning. Improvement in spatial learning performance occurred with extra training despite continued use of a suboptimal search strategy, and, in a separate nonspatial task, the mutants again deployed a different strategy. Double-mutant analysis of postsynaptic density-95 and SAP102 mutants indicate overlapping and specific functions of the two MAGUKs. These in vivo data support the model that specific MAGUK proteins couple the NMDA receptor to distinct downstream signaling pathways. This provides a mechanism for discriminating patterns of synaptic activity that lead to long-lasting changes in synaptic strength as well as distinct aspects of cognition in the mammalian nervous system.

Figure 1.

Generation of SAP102-targeted mice. a, SAP102 has multiple protein–protein interaction domains, including three tandem PDZ (PSD-95/Discs large/zona occludens-1), an Src homology 3 (SH3), and a guanylate kinase (GK) domain (top). We replaced SAP102 exons 2–8 with a selection cassette in targeted mice, deleting the majority of the PDZ-coding sequence and creating a frame-shift mutation between exons 1 and 9 (middle and bottom). The 5′ probe outside the targeting region and primers P1–P3, used for Southern blot and PCR genotyping respectively, are shown. D, DraI restriction sites. b, Southern blots of DraI-digested genomic DNA from wt (+/Y) and targeted (−/Y) ES cells, and wt, heterozygous (+/−), and hemizygous (−/Y) mice confirm the structure of the mutant allele. c, PCR genotyping of targeted SAP102 mice using a common forward primer (P1) and separate reverse primers, P2 and P3, to amplify the wt and mutant alleles, respectively. d, SAP102 is undetectable in Western blots of forebrain extract from hemizygous mutant mice. e, Immunohistochemical staining shows loss of SAP102 throughout the brain of mutant mice. Expression patterns of other NMDA receptor-associated MAGUKs, PSD-95 and PSD-93, are unaffected. Immunohistochemical staining is brown, and hematoxylin counterstain is blue. Scale bars, 2 mm.

Figure 2.

Impaired spatial learning in SAP102 mutant mice. a, Watermaze timeline. Dark triangles indicate probe trials conducted at the end of the training trials; vertical lines indicate training trials. b, Results of training sessions for each day of the visible platform, hidden platform, and reversal platform stages of the task. c, The mean percentage time spent in the training quadrant for the probe test blocks. The dotted line represents chance performance. * and § indicate performance significantly better than chance for wt and mutants, respectively. ** indicates both a group effect and performance significantly different from chance. Error bars are ±SEM.

Figure 3.

Strategy choices in the water maze. Shown are the percentages of wt and SAP102^−/Y mice that choose a particular search strategy. a, Sample traces of the paths for each of the seven search strategy categories. DF, Direct finding; A, approaching target; O, orienting; S, scanning; RS, random search; C, circling; T, thigmotaxic. b, Choices during the visible training trials. c, Choices for the first two probe trials (H1+H2) after the initial training. d, Choices for the early reversal probe trials (R1+R2). e, Choices for the reversal probe trials, analogous to the initial probe trials (H1+H2). Modal choices (M) are shown for wt and for SAP102^−/Y.

Figure 4.

Basal synaptic transmission and postsynaptic receptor function are normal in SAP102 mutant mice. a, The input–output curves for wt (open symbols; n = 5 mice, 10 slices) and SAP102 mutant mice (filled symbols; n = 5 mice, 8 slices) are shown. Presynaptic fiber volleys and postsynaptic fEPSP slopes were determined at four different stimulation intensities that elicited fEPSPs corresponding to 25, 50, 75, and 100% of the maximum fEPSP amplitude. b, Paired-pulse facilitation is normal in SAP102 mutant mice. c, NMDA receptor-mediated EPSCs are normal in SAP102 mutant mice. Whole-cell voltage-clamp techniques were used to record EPSPs at two different postsynaptic membrane potentials (−80 and +40 mV). NMDA receptor-mediated EPSCs (estimated from the EPSC amplitude 50 ms after EPSC onset) are expressed relative to the AMPA receptor-mediated component of the EPSC (estimated from the EPSC amplitude 5 ms after EPSC onset). There is no difference in EPSCs between wt (open bars; n = 15 cells from 3 mice) and SAP102 mutants (filled bars; n = 22 cells from 3 mice) at either membrane potential. The inset shows EPSCs recorded at −80 and +40 mV in wt and SAP102 mutant CA1 pyramidal cells. Calibration: 20 ms, 50 pA. d, Comparison of single-exponential curves fitted to the decaying phase of the synaptic currents recorded at +40 mV. No difference in the decay characteristics was observed between wt and SAP102 mutant mice.

Figure 5.

Loss of SAP102 results in specific enhancement of hippocampal synaptic plasticity induced by theta frequency and spike timing-dependent stimulation. a, High-frequency stimulation-induced LTP is normal in SAP102 mutants. Sixty minutes after 100 Hz stimulation (2 trains, each of 1 s duration, delivered at time = 0) fEPSPs were potentiated to 204 ± 13% of baseline in wt slices (n = 5 mice, 8 slices) and were potentiated to 230 ± 20% of baseline in SAP102 mutant slices (n = 4 mice, 9 slices; not significant compared with control, p = 0.28). b, Low-frequency stimulation (900 pulses at 5 Hz) induces modest LTP in wt animals (fEPSPs potentiated to 150 ± 5% of baseline; n = 8, 15 slices) but significantly enhanced LTP in SAP102^−/Y slices (201 ± 4% of baseline; n = 6 mice, 16 slices; p < 0.005). c, Single postsynaptic APs paired with single pulses of presynaptic fiber stimulation fails to induce LTP in wt pyramidal cells but induces robust LTP in SAP102^−/Y cells. Pairing (100 paired stimulations delivered at 10 Hz with pre/post interval of 10 ms) was delivered at time = 0. Inset shows superimposed EPSPs recorded during baseline and 30 min after pairing in a pyramidal cell from a wt (left) and a SAP102^−/Y (right) mouse. d, Bursts of postsynaptic APs paired with presynaptic fiber stimulation induce similar amounts of LTP in wt and SAP102^−/Y mutant cells. EPSPs were paired with bursts of three to four postsynaptic action potentials elicited by a 50 ms depolarizing current pulse delivered via the recording electrode.

Figure 6.

Altered postsynaptic signaling in SAP102 mutant mice. a, No change in total hippocampal levels of postsynaptic proteins in mutant mice. i, MAGUKs; ii, postsynaptic receptors; iii, postsynaptic signaling proteins. b, Elevated basal ERK phosphorylation in SAP102 mutant mice. Western blotting of hippocampal protein extracts with an antibody against phospho-ERK shows a consistent elevation of the phosphorylated form of the protein with no change in total ERK levels. Representative samples from 15 wt and 15 SAP102^−/Y animals are shown. c, A sandwich ELISA assay confirms the increase in ERK phosphorylation (t₍₁₇₎ = 3.38; p = 0.003) without a change in total ERK (t₍₂₃₎ < 1; NS). d, The MEK inhibitor U0126 blocks the enhancement of LTP in SAP102 mutant mice. Slices were continuously bathed in ACSF containing 20 μm U0126. fEPSPs were potentiated to 137 ± 12% of baseline in wt slices (open symbols; n = 4) and 143 ± 12% of baseline in slices from SAP102 mutant mice (filled symbols; n = 5; p = 0.76 compared with wt). The histogram on the right shows the magnitude of LTP measured 45 min after theta pulse stimulation in wt and SAP102 mutant slices under control conditions and in the presence of U0126. e, Phospho-ERK response to NMDA stimulation is attenuated in SAP102 mutant mice. NMDA was bath applied to hippocampal slices, and phospho-ERK was analyzed by Western blotting. SAP102^−/Y levels at time = 0 are normalized to those of wt slices. Phospho-ERK levels in −/Y slices are reduced compared with wt 5 min and 20 min after stimulation. Asterisks indicate statistically significant differences. f, Increase in phospho-MEK is also attenuated in response to NMDA application. g, Independent experiments without normalization of wt and SAP102^−/Y levels at time = 0 show that phospho-ERK response to NMDA stimulation is attenuated notwithstanding its basal elevation.

Figure 7.

Overlapping but distinct functional roles of SAP102 and PSD-95. a, SAP102 protein expression is elevated in the hippocampus of PSD-95 mutant mice. Shown are Western blots on hippocampal extracts from wt and PSD-95 homozygous animals. b, Increased PSD-95 associated with NMDARs in SAP102 mutant mice. The NMDA receptor complex was immunoprecipitated (IP) from whole forebrain extracts with an NR1 antibody and then blotted with PSD-95. Both a and b show representative results from 10 wt and 10 mutant animals. c, SAP102 and PSD-95 double mutation is lethal. Crosses between SAP102 and PSD-95 mutant mice produced a very low proportion of female PSD-95^−/− mice, SAP102^+/− mice, and no male double knock-outs. A χ² test for goodness of fit shows that the distribution of offspring genotypes is severely skewed (χ² = 34.4; n = 88; p = 0.0001).