Generation and characterization of a Shank3-deficient mouse
We made use of gene targeting in Bruce4 C57BL/6 embryonic stem (ES) cells [32
] to generate a mouse line that has loxP sites inserted before exon 4 and after exon 9 (Figure ). For all studies reported here, the floxed allele was excised and a line was maintained with a deletion of exons 4 through 9. This line, which completely deletes the ankyrin repeat domains of Shank3, produced wild-type (+/+), heterozygous (+/-) and knockout (-/-) animals with Mendelian frequencies from heterozygote-heterozygote crosses.
qPCR showed 50% reduction of full-length Shank3 mRNA in the heterozygotes and complete loss in knockouts (Figure ). Moreover, there was no expression of full-length Shank3 protein in PSD fractions from Shank3-knockout mice and reduced expression in the heterozygotes, using antibodies which cross-react either with an epitope downstream of the PDZ domain (antibody N69/46; see Figure ) (Figure ) or with the COOH terminal (data not shown), consistent with haploinsufficiency.
Heterozygous and homozygous animals were viable and showed no obvious alterations in gross brain structure or hippocampal cytoarchitecture, nor were there any obvious seizures. There was evidence for subtle motor abnormalities in homozygotes, which is being further characterized in ongoing studies. Both genotypes were normal on measures of general health, developmental milestones and exploratory activity, as well as on social approach as measured in an automated three-chambered social approach task.
Because SHANK3 haploinsufficiency is responsible for the neurobehavioral phenotype in individuals with 22q13DS or SHANK3 mutations, we focused the studies reported here on Shank3 heterozygous mice generated by crossing wild-type mice with heterozygotes to be most relevant to the clinical syndromes and to be most useful in ultimately assessing potential therapeutic interventions in preclinical studies. A comprehensive investigation of the knockout mice obtained through an alternate breeding strategy (heterozygote-heterozygote matings) is now in progress in independent experiments.
Basal glutamatergic synaptic transmission is reduced in Shank3 heterozygous mice
To examine the role of Shank3 in regulating synaptic glutamate receptor function, we studied glutamatergic synaptic transmission in hippocampal slices. We examined the properties of basic transmission at Schaffer collateral-CA1 synapses in hippocampal slices from 3- to 4-week-old Shank3 heterozygous mice and their littermates using extracellular recordings. Plotting field excitatory postsynaptic potential (fEPSP) slope versus stimulus intensity demonstrated a reduction in the I/O curves in the heterozygotes (data not shown), prompting us to then further examine synaptic transmission in the presence of inhibitors of specific subtypes of glutamate receptors. In the presence of the NMDA receptor antagonist APV, a decrease in AMPA receptor-mediated field potentials in the heterozygous mice was seen, reflected as a 50% decrease in the average slope of I/O function compared with wild-type mice (n = 4 mice per genotype, two to three slices per mouse; P = 0.001; Figure ). In contrast, when the I/O relationship was analyzed in the presence of the competitive AMPA/kainate receptor antagonist CNQX to measure synaptic NMDA receptor function, there was no difference between genotypes (n = 4 mice per genotype, two to three slices per mouse; P = 0.1; Figure ). These results indicate that there is a specific reduction in AMPA receptor-mediated basal transmission in the Shank3 heterozygous mice.
Figure 2 Altered basal synaptic properties in Shank3 heterozygous mice. (A) Reduced α-amino-3-hydroxyl-5-methyl-4-isoxazole-propionic acid (AMPA) receptor responses in Shank3 heterozygous mice. Slices were incubated in the presence of 2-amino-5-phosphonopentanoic (more ...)
We asked whether the impairment of evoked synaptic transmission in heterozygous mice is caused by alterations in presynaptic and/or postsynaptic parameters. We performed whole cell patch-clamp recordings in 3-month-old littermates and monitored spontaneous miniature postsynaptic currents in the presence of tetrodotoxin (TTX). The amplitude of miniature excitatory postsynaptic currents (mEPSCs) from hippocampal CA1 pyramidal neurons of heterozygous mice were significantly smaller than those in control mice (n = 7 for wild-type and n = 8 for heterozygous mice, P < 0.01; Figures and ), which was evident by the significant shift of the cumulative probability to the left (Figure ), again indicating a reduction in basal transmission. However, in heterozygous mice, the frequency of miniature excitatory postsynaptic currents was significantly higher (n = 7 for wild-type and n = 8 for heterozygous mice, P < 0.03; Figures and ) and paired-pulse ratio was decreased (n = 6 for wild-type and n = 7 for heterozygous mice, P < 0.05; Figure ), which revealed an additional, presynaptic alteration in the heterozygotes as well.
Long-term potentiation is impaired in Shank3 heterozygous mice
We next examined long-term potentiation (LTP) with extracellular fEPSP recordings at Schaffer collateral/CA1 synapses. In the first set of experiments, LTP was induced by tetanic stimulation of the Schaffer collaterals (four trains of 100 Hz separated by 5 min). While initial expression of LTP was identical across the two genotypes, the maintenance of LTP was clearly impaired in the heterozygous mice (average percentage of baseline 120 min after tetanus: 165.1 ± 8.8% in wild-type and 117.1 ± 9.5% in heterozygous mice, P = 0.004; Figure ). In an additional set of experiments, we further tested TBS LTP (10 bursts of four pulses at 100 Hz separated by 200 ms), which also showed a significant decrease in the potentiation at 60 min after TBS in heterozygous mice (156.3 ± 9.2% of baseline in wild-type and 126.0 ± 8.9% in heterozygous mice, measured 60 min after TBS, P = 0.007; Figure ).
Figure 3 Reduced long-term potentiation in Shank3 heterozygous mice. (A) Long-term potentiation (LTP) following high-frequency stimulation. Field recordings of LTP induced with high-frequency stimulation (HFS; 4 times 100 Hz, separated by 5 min) as a function (more ...)
In contrast to the altered synaptic plasticity observed with LTP, long-term depression (LTD) induced by either low-frequency stimulation (LFS) (82.6 ± 1.35% of baseline in wild-type and 79.9 ± 2.5% in heterozygous mice, measured 60 min after LFS, P > 0.1; Figure ) or paired-pulse LFS (PP-LFS) stimulation (81.6 ± 6% of baseline in wild-type and 82.7 ± 1.9% in heterozygous mice, P > 0.1; Figure ) was not significantly changed in heterozygotes.
Previous studies have shown that LTP is accompanied by spine enlargement [35
]. Therefore, it was of interest to determine whether the deficits in LTP in the Shank3
heterozygous mice were associated with altered spine remodeling. We first established that spines from wild-type mice were capable of structural modification by simultaneously monitoring spine size and synaptic responses in CA1 neurons before and after TBP [34
]. We found that TBP produced a rapid and persistent increase in spine volume concurrent with an immediate increase in EPSP slope in whole cell recordings, which gradually reached a plateau by ~30 min (Figures ). However, in recordings from heterozygous mice, we found that the stabilization in synaptic potentiation and spine expansion were impaired in CA1 neurons (Figures ). In the heterozygous mice, both EPSP slope and spine volume increased immediately to values comparable to those of control spines, but synaptic potentiation and spine expansion failed to be sustained.
GluR1 immunoreactivity is decreased in Shank3 heterozygous mice
We then asked whether Shank3
haploinsufficiency can lead to alterations in the numbers of AMPA receptor-positive puncta, given the results from the electrophysiological experiments and the relationship between synaptic strength and AMPA receptor subunit trafficking [49
]. We carried out immunolabeling for GluR1 (an AMPA receptor subunit) and quantified GluR1-immunoreactive puncta (Figures ). Neurons from Shank3
heterozygous mice showed significantly fewer GluR1-immunoreactive puncta (P
< 0.005) consistent with the electrophysiological data.
Figure 4 Decreased density of GluR1-immunoreactive puncta in Shank3 heterozygous mice. (A and B) High-resolution confocal images of puncta from wild-type and heterozygous mice. (C and D) The same images from Figures 4A and 4B are shown after deconvolution. (E (more ...)
Behavioral analyses of Shank3 heterozygotes
To more extensively study social interactions in Shank3
heterozygous mice, we examined male-female social interactions during a 5-min session of freely moving reciprocal social interactions with an estrus B6 female (Figure ). Cumulative duration of total social sniffing by the male test subjects was lower in Shank3
heterozygotes than in wild-type littermates (P
= 0.02). In addition, fewer ultrasonic vocalizations were emitted by heterozygotes than by wild-type littermates during the male-female social interaction session (P
= 0.003). Note that while the equipment used could not distinguish between calls emitted by the male subject and female partner, the preponderance of calls during male-female interactions in mice is usually emitted by the male [50
Figure 5 Reduced social behaviors in Shank3 heterozygous and wild-type littermate mice. (A) Adult male-female social interactions. Left: Total duration of social interactions, scored as cumulative seconds spent by the male subject in sniffing the nose, anogenital (more ...)
As a control task to ensure that subject mice can detect the social pheromones that elicit approach and vocalizations, we measured olfactory abilities using the olfactory habituation/dishabituation test. Habituation and dishabituation were normal for social and nonsocial odor cues in both wild-type and Shank3 heterozygous mice (n = 8/genotype; Figure ). Both genotypes displayed the expected habituation, indicated by decreased time spent in sniffing the sequence of three same odors, and the expected dishabituation, indicated by increased time sniffing the different odor (water habituation, wild-type: P < 0.001; water habituation, heterozygotes: P < 0.001; almond habituation, wild-type: P = 0.015; almond habituation, heterozygotes: P = 0.04; banana habituation, wild-type: P < 0.001; banana habituation, heterozygotes: P = 0.033; social odor 1 habituation, wild-type: P < 0.001; social odor 1 habituation, heterozygotes: P < 0.001; social odor 2 habituation, wild-type: P < 0.001; social odor 2 habituation, heterozygotes: P < 0.001; water to almond dishabituation, wild-type: P = 0.061; water to almond dishabituation, heterozygotes: P = 0.118; almond to banana dishabituation, wild-type: P = 0.005; almond to banana dishabituation, heterozygotes: P = 0.046; banana to social odor 1 dishabituation, wild-type: P < 0.001; banana to social odor 1 dishabituation, heterozygotes: P = 0.046; social odor 1 to social odor 2 dishabituation, wild-type: P < 0.001; social odor 1 to social odor 2 dishabituation, heterozygotes: P < 0.001).