INTERNET

internalization and reduced synaptic transmission

We next addressed how AMPA-receptor trafficking is regulated in neurons that lack ephrinB2. The genomic ablation of ephrinB2 results in embryonic lethality (E11) due to a severe developmental defect in the vasculature (Adams et al., 2001; Wang et al., 1998). Therefore, to analyze AMPA-receptor endocytosis, we made use of a previously generated conditional ephrinB2 knockout mouse (ephrinB2lox/lox_{)crossed to a Nestin-Cre line that} specifically deletes ephrinB2 from the nervous system (Grunwald et al., 2004).

74 5.2.2.1 EphrinB2 protein is absent in Nestin-Cre+ _ephrinB2lox/lox _mice

First, we confirmed the successful ablation of the ephrinB2 protein in the nervous system of conditional ephrinB2 knockout Nestin-Cre+ _{animals (eB2KO) by various methods} (Figure 5-12). Cortices from E17 eB2KO or control litter mates were homogenised and analyzed by Western blot as total lysates, or after a specific EphB4-Fc-pulldown that enriches the ephrinB2 protein (Figure 5-12, a+b). Neither the total lysates nor the specific pulldown showed a detectable level of ephrinB2 protein in these animals.

Figure 5-12: EphrinB2 protein is absent in eB2KO neurons. (a) Total cortex lysates (TL) of E17 embryos analyzed by Western blot using anti-ephrinB2 (R&D) antibodies. (b) EphrinB2 specific pulldown (EphB4-Fc) showed complete absence of ephrinB2 protein E17 cortex lysates of eB2KO mice in Western blot. (c) Primary hippocampal neurons 17DIV isolated from eB2KO or control litter mates stimulated with pre-clustered EphB4-Fc to induce ephrinB2 clusters. Cluster formation was visualized using anti-hFc cy2 antibodies. Scale bar, 5 µm.

Additionally, we analyzed primary hippocampal neurons of eB2KO and control littermates for ephrinB2-expression. Neurons (17DIV) were stimulated with pre- clustered EphB4-Fc or Fc (control) for 20 minutes and ephrinB2 clusters were visualized

75

by indirect detection of EphB4-Fc using anti-hFc-cy2 antibodies. While control cells showed punctuated, positive signals, typical for ephrinB2 surface clusters, the eB2KO neurons were completely devoid of any signal (Figure 5-12, c).

5.2.2.2 eB2KO mouse neurons show enhanced constitutive AMPA-receptor internalization

After verifying the absence of the ephrinB2 protein we analyzed the levels of AMPA- receptor internalization in hippocampal neurons from eB2KO and control litter mates. Hippocampal neurons isolated from E17 eB2KO mice and control littermates were cultured for 15-17 days and the levels of AMPA-receptor endocytosis were determined by the antibody feeding assay method as in Figure 5-10 , a+b.

Figure 5-13: EphrinB2 KO neurons show increased AMPA-receptor internalization. (a) Hippocampal neurons from eB2KO and control litter mates cultured for 16DIV and analyzed for AMPA- receptor endocytosis by the antibody feeding assay under the indicated conditions as in Figure 5-10, b. Scale bars, 20 µm whole neurons, 5 µm enlargements. (b) Quantification of AMPA-receptor internalization pictured as percentage of internalized GluR2 versus total GluR2 under the conditions presented in a (SEM, *** P < 0.0001). (c) Total level of GluR2 was not affected in eB2KO neurons analyzed in Western blot using anti-GluR2 and anti-tubulin (control) antibodies.

76 Under control condition (Fc), eB2KO neurons already showed an increased level of AMPA-receptor internalization (55.4 % ± 1.7) compared to control cells (30.3 % ± 1.4). As expected, simultaneous stimulation with pre-clustered EphB4-Fc and 100 µM AMPA did not inhibit AMPA-induced receptor internalization since ephrinB2 is not present in these neurons. The lack of ephrinB2 protein resulted in a destabilization of AMPA receptors at the cell surface but had no effect on the total levels of GluR2 (

Figure 5-13, c). When total lysates of cultured hippocampal neurons (17DIV) from eB2KO mice and control litter mates were analyzed by Western blot, the levels of GluR2, in relation to tubulin, were unchanged. Thus, ephrinB2 regulates AMPA-receptor trafficking by influencing the levels of receptors at the surface.

5.2.2.3 eB2KO neurons show reduced synaptic transmission

The endocytosis rate and, consequently the actual surface presence of AMPA receptors, turned out to be affected in eB2KO neurons. Since the amount of AMPA receptors at synaptic sites influences synaptic transmission, we next investigated the electrophysiological properties of these eB2KO neurons. We recorded miniature excitatory postsynaptic currents (mEPSCs) that are caused by the spontaneous release of single presynaptic vesicles. The amplitude of the mEPSCs correlates with the number of synaptic AMPA receptors and can be used as a read-out for synaptic transmission properties. The electrophysiological analysis of mEPSCs in cultured eB2KO, and control litter mate neurons, was performed by Matthias Traut (MPI of Neurobiology, AG Stein). The eB2KO cells showed smaller mEPSCs and their mean amplitude was significantly reduced from 19.3 pA ± 1.1 in control neurons to 16.6 pA ± 0.6 in eB2KO neurons (Figure 5-14, a+b). Even though the synaptic transmission was reduced in KO neurons, the number of synapses, determined by the number of PSD-95 (post-synaptic marker) and synapsin1 (pre-synaptic marker) positive puncta per 100 µm-dendrite stretch, remained equal (Figure 5-14, c). Thus the lack of ephrinB2 in these neurons leads to increased AMPA-receptor endocytosis resulting in a decreased number of AMPA receptors at the synapse at a given time point.

77

Figure 5-14: eB2KO neurons show reduced synaptic transmission. (a) Cumulative probability histogram of mEPSC amplitudes of control (blue) and eB2KO (red) hippocampal neurons (Kolmogorov- Smirnov test, P < 0.05, 80 events per cell, control n=10 cells, eB2KO n=13 cells). (b) Mean of mEPSC amplitudes in control and eB2KO hippocampal neurons (SEM, * P < 0.05). (c) Number of synapses in cultured hippocampal neurons of eB2KO and control litter mates. Quantified were numbers of PSD-95 positive or synapsin1 positive clusters per 100 µm-dendrite stretches. (d) Stimulation with 10 mM KCl induces ephrinB2-cluster formation in hippocampal neurons. Surface clusters visualized via indirect detection with EphB4-Fc and anti-hFc antibodies. Scale bars, 20 µm whole neurons, 5 µm enlargements. (e) Quantification of ephrinB2 clusters after KCl stimulation pictured as number of clusters per 100 µm- dendrite stretch (SEM, *** P < 0.00001).

5.2.2.4 Hyperpolarization leads to ephrinB2 cluster formation

The synaptic strength depends on the number of AMPA receptors at the synaptic surface. Insertion of additional AMPA receptors or endocytosis, respectively, is regulated by activity and assures the plasticity of synaptic contacts. Since ephrinB2 activation seemed to

78 stabilize AMPA receptors at the cell surface, we wanted to determine whether ephrinB2 clustering and reverse signaling could be influenced by neuronal activity. Therefore, we induced a change of the membrane potential in cultured hippocampal neurons by applying a hyperpolarizing 10 mM KCl solution for 10 minutes and examined the induction of ephrinB2 clusters. EphrinB2-cluster formation was visualized via indirect detection through EphB4-Fc and anti-hFc cy2 antibodies and was quantified as number of clusters per 100 µm-dendrite stretch (Figure 5-14, c). The hyperpolarization resulted in a significant increased number of ephrinB2 clusters from 52.8 ± 2.6 at control neurons to 77.6 ± 2.6 after KCl-stimulation.