Synaptic Function of Rab11Fip5: Selective Requirement for Hippocampal Long-Term Depression

Abstract

Postsynaptic AMPA-type glutamate receptors (AMPARs) are among the major determinants of synaptic strength and can be trafficked into and out of synapses. Neuronal activity regulates AMPAR trafficking during synaptic plasticity to induce long-term changes in synaptic strength, including long-term potentiation (LTP) and long-term depression (LTD). Rab family GTPases regulate most membrane trafficking in eukaryotic cells; particularly, Rab11 and its effectors are implicated in mediating postsynaptic AMPAR insertion during LTP. To explore the synaptic function of Rab11Fip5, a neuronal Rab11 effector and a candidate autism-spectrum disorder gene, we performed shRNA-mediated knock-down and genetic knock-out (KO) studies. Surprisingly, we observed robust shRNA-induced synaptic phenotypes that were rescued by a Rab11Fip5 cDNA but that were nevertheless not observed in conditional KO neurons. Both in cultured neurons and acute slices, KO of Rab11Fip5 had no significant effect on basic parameters of synaptic transmission, indicating that Rab11Fip5 is not required for fundamental synaptic operations, such as neurotransmitter release or postsynaptic AMPAR insertion. KO of Rab11Fip5 did, however, abolish hippocampal LTD as measured both in acute slices or using a chemical LTD protocol in cultured neurons but did not affect hippocampal LTP. The Rab11Fip5 KO mice performed normally in several behavioral tasks, including fear conditioning, but showed enhanced contextual fear extinction. These are the first findings to suggest a requirement for Rab11Fip5, and presumably Rab11, during LTD.

Keywords: LTD; Rab11; Rab11Fip5; autism.

Figure 1.

Characterization of Rab11Fip5 shRNA knockdown. A, Diagram of the treatment conditions (see also Fig. 2). The “KD+Rescue” condition is encoded in a single lentivirus. For the last three conditions, two virus infections were performed at days 2 and 4. B, mRNA measurements for Rab11Fip5. Of the two probes, probe 1 does not recognize the rescuing cDNA. This allows measurements of the endogenous mRNA levels as well as expression of the rescuing construct. Rab11Fip5 mRNA levels were assayed by qRT-PCR and are normalized to those of GAPDH. C, Spontaneous mIPSCs recorded in cultured hippocampal neurons. Sample traces (left), event frequency (middle), and average amplitude (right) are shown. D, Spontaneous mEPSCs after Rab11Fip5 shRNA KD. The number of cells/cultures used is indicated in each graph. Graphs represent mean ± SEM. ***p < 0.001 (one-way ANOVA).

Figure 2.

Characterization of inhibitory transmission after Rab11Fip5 shRNA knockdown. A, Evoked IPSCs recorded in cultured hippocampal neurons are severely reduced by a Rab11Fip5 shRNA. Sample traces (left) and peak current amplitude and charge transfer during one action potential (right two panels) are shown. Notch marks in sample traces indicate the time of stimulation; the stimulation artifact has been removed for clarity. B, eIPSC responses during a 10 Hz stimulus train. Sample traces (left), total charge transfer during the train (middle), and peak amplitudes for each stimulus normalized to the first peak (right) are shown. C, IPSCs triggered by a 30 s application of 0.5 M sucrose, sample traces (left) and quantification of the charge transfer during the 30 s sucrose application (right). The timing of sucrose application is indicated by the black rectangle. Recordings are performed in the presence of TTX to block action potentials in addition to AP5 and CNQX that are used to isolate inhibitory currents. The number of cells/cultures recorded is indicated in each graph. Graphs represent mean ± SEM. ***p < 0.001 (one-way ANOVA).

Figure 3.

Rab11Fip5 KO mice are viable and have normal synaptic protein levels. A, Strategy for generating Rab11Fip5 conditional KO animals. The genomic locus is shown with exon 1 containing the ATG start site. loxP sites are indicated by triangles while frt sites by ovals. B, mRNA expression levels are not altered in cultured hippocampal neurons from Rab11Fip5 w/w and cKO/cKO littermates. Rab11Fip5 mRNA levels are normalized to those of NeuN. Three independent cultures were infected with no virus, a lentivirus carrying an inactive form of Cre (ΔCre), or a Cre virus, total RNA was harvested and assayed by qRT-PCR. C, Rab11Fip5 KO alleles do not affect survival rates. Dashed line indicates expected ratio, which is not statistically different from that observed, >100 offspring were assessed. D, Representative ¹²⁵I immunoblots of synaptic proteins from whole-brain lysates of Rab11Fip5 w/w and KO/KO littermates. Loading controls are β-actin and valosin-containing protein (VCP). E, Quantification of protein levels in immunoblots from three different littermate pairs shows no significant changes. B, E, Graphs represent mean ± SEM. ***p < 0.001.

Figure 4.

Rab11Fip5 KO facilitates contextual fear extinction. A–H, Male, littermate Rab11Fip5 WT and KO animals after fear conditioning show no differences in freezing rates during exposure to context (A, E), tone (B, F), or an altered context (C, G). Reexposure to the context results in fear extinction, which is facilitated in Rab11Fip5 KO mice only for the second day trial (D). Fear extinction during multiple exposures to the conditional stimulus (tone) was unaltered in the KO mice (H). I–K, As assayed by an open-field test, wild-type and mutant mice had no statistically significant differences in baseline locomotion (I) and anxiety (J, K). L, The number of spontaneous alternations in a Y-maze test, a measure of short-term memory, was undistinguishable between WT and KO Rab11Fip5 mice. M, Rab11Fip5 KO mice showed normal responses in a social preference task, favoring the animal chamber versus the object chamber at the same rate as wild-type littermates. In all panels, graphs represent mean ± SEM. Statistical significance is assessed by two-way repeated-measures ANOVA for the extinction trials, one-way ANOVA in the other cases: *p < 0.05; **p < 0.01.

Figure 5.

Rab11Fip5 KO cultures show no major electrophysiological changes. A, Evoked IPSCs recorded from Rab11Fip5 cKO-cultured hippocampal neurons: sample traces (left) and peak current amplitude (right). The conditions in this and subsequence panels are as follows: lentivirus carrying a catalytically inactive Cre (ΔCre), Cre lentivirus (Cre), and a single lentivirus carrying Cre and a Rab11Fip5 rescue cDNA (Cre+Fip5). B, eIPSC responses during a 10 Hz stimulus are not altered by Rab11Fip5 KO. Sample traces (top), total charge transfer during the train (bottom left), and normalized peak amplitudes (bottom right) are shown. C, Sucrose-triggered IPSCs are also unaffected by the KO: sample traces (left) and quantification of the charge transfer during the 30 s sucrose application (right). D, Evoked EPSCs in Rab11Fip5 cultures: traces (left) and peak amplitude (right). Because recordings are performed in mixed cultures, network activity is observed soon after the first stimulus. E, F, Spontaneous mIPSCs (E) and mEPSCs (F) have normal amplitude and frequency in Rab11Fip5 cKO cultures. G, H, The Rab11Fip5 shRNA acts in a nonspecific manner. Evoked IPSCs (left) and EPSCs (right) recorded from cKO cultures (G) infected with two viruses (carrying Cre and the KD shRNA) are reduced compared with the Cre virus-infected cultures. The phenotype is likely not due to the Rab11Fip5 mRNA levels because the mRNA levels are similar in the two conditions (H). The number of cells/cultures recorded is indicated in each graph. Graphs represent mean ± SEM. ***p < 0.001 (one-way ANOVA).

Figure 6.

LTP is unchanged in Rab11Fip5 KO mice. A, Sample experiments (left two panels) and summary graphs (right two panels) of LTP obtained by extracellular field recordings in stratum radiatum of acute hippocampal slices from heterozygous (control) and Rab11Fip5 KO mice (control, 168 ± 7%, n = 5; KO, 175 ± 17%, n = 5). Arrow indicates the time of tetanic stimulation. Sample-averaged fEPSPs during the baseline (1) and 50 min after LTP induction (2) are shown above the sample experiments. B, Sample experiments (left two panels) and summary graphs (right two panels) of LTP obtained by whole-cell patch-clamp recordings of CA1 pyramidal neurons in acute hippocampal slices from heterozygous (control) and Rab11Fip5 KO mice (control, 197 ± 14%, n = 6, KO, 200 ± 18%, n = 7). Arrow indicates the time of tetanic stimulation. Sample-averaged EPSCs during the baseline (1) and 30 min after LTP induction (2) are shown above the sample experiments. The number of slices (A) or cells (B) recorded is indicated in each graph. Graphs represent mean ± SEM. No significant statistical differences were found using Student's t test.

Figure 7.

LTD is impaired in Rab11Fip5 KO mice. A, Sample experiments (left two panels) and summary graphs (right two panels) of LTD obtained by extracellular field recordings in stratum radiatum of acute hippocampal slices from heterozygous (control) and Rab11Fip5 KO mice. Arrow indicates the time of onset of LTD induction stimulation. Sample-averaged fEPSPs during the baseline (1) and 50 min after LTD induction (2) are shown above the sample experiments. B, LTD induction requires NMDA receptor activity. Same as A, but LTD was induced in Rab11Fip5 heterozygous animals, in the absence (Control) or presence (Control + AP5) of AP5 to block NMDA receptor. C, Sample experiments (left two panels) and summary graphs (right two panels) of LTD obtained by whole-cell patch-clamp recordings of CA1 pyramidal neurons in acute hippocampal slices from heterozygous (control) and Rab11Fip5 KO mice. Arrow indicates the time of onset of LTD induction stimulation. Sample-averaged EPSCs during the baseline (1) and 30 min after LTD induction (2) are shown above the sample experiments. The number of slices (A, B) or cells (C) recorded is indicated in each graph. Graphs represent mean ± SEM. *p < 0.05 (Student's t test).

Figure 8.

Rab11Fip5 is not required for basal transmission. A, Sample traces (left) and summary graph of the linear fit slopes (right) (control, 7.2 ± 0.3, n = 7; KO, 6.6 ± 0.3, n = 7) for input–output measurements obtained by extracellular field recordings in acute hippocampal slices from heterozygous (control) and Rab11Fip5 KO mice. B, Sample traces (left) and mean amplitude (right) (control, 11.0 ± 0.5 pA, n = 16; KO, 10.7 ± 0.5 pA, n = 16) and frequency (control, 0.23 ± 0.03 Hz; KO, 0.20 ± 0.02 Hz) of mEPSCs, which are unchanged in Rab11Fip5 KO mice. C, The ratio of NMDAR- to AMPAR-mediated EPSCs is unchanged in Rab11Fip5 KO mice. Representative EPSCs recorded at −60 mV and 40 mV are shown with scatter plots of individual recordings. The different colored point indicates mean ± SEM (Control, 0.50 ± 0.05, n = 10; KO, 0.51 ± 0.05, n = 9). D, Paired-pulse ratios of AMPAR EPSCs are unchanged in Rab11Fip5 KO mice. Representative EPSCs are shown on the left (PP20: control, 1.8 ± 0.1; KO, 1.9 ± 0.2; PP50: control, 1.6 ± 0.1; KO, 1.6 ± 0.1; PP100: control, 1.42 ± 0.06; KO, 1.4 ± 0.1; PP200: control, 1.25 ± 0.03; KO, 1.24 ± 0.08; control, n = 10; Rab11Fip5 KO n = 9). Individual points represent mean ± SEM. The number of cells recorded (B, C) is indicated in each graph. Data are mean ± SEM. No significant statistical differences were found using Student's t test.

Figure 9.

Rab11Fip5 is required for NMDA-triggered chemical LTD in neuronal cultures. A, Synapse number and size are not altered by Rab11Fip5 KO as determined by immunohistochemistry against GluA1. Representative images are shown on the left. Number of GluA1 puncta (center) and the size of these puncta (right) were not different in the two conditions. Both measurements are normalized to the control ΔCre condition. Data are from four cultures and at least 6 random fields of view for each culture. B, Schematic of the NMDA-dependent GluA1 endocytosis protocol. To prevent cell toxicity, neurons are incubated in TTX, LY341495, and DNQX. After fixing, external GluA1 primary antibodies are labeled with secondary antibodies, followed by permeabilization and labeling of the internalized pool of GluA1 antibodies with another secondary antibody. C, NMDA-induced GluA1 internalization is blocked in the absence of Rab11Fip5. Representative images (left) and quantification (right) of the internalized GluA1 receptors. The level of internalization is normalized to that observed in the absence of NMDA application. A 15 min chase was performed. D, Initial internalization of GluA1 receptors appears unaffected in absence of Rab11Fip5. Same as in C, but a 5 min chase was performed. Data are from three independent cultures, and the number of random fields of view analyzed is indicated. **p < 0.01 (Student's t test).