Leucine-rich repeat transmembrane proteins are essential for maintenance of long-term potentiation

Abstract

Leucine-rich repeat transmembrane proteins (LRRTMs) are synaptic cell adhesion molecules that trigger excitatory synapse assembly in cultured neurons and influence synaptic function in vivo, but their role in synaptic plasticity is unknown. shRNA-mediated knockdown (KD) of LRRTM1 and LRRTM2 in vivo in CA1 pyramidal neurons of newborn mice blocked long-term potentiation (LTP) in acute hippocampal slices. Molecular replacement experiments revealed that the LRRTM2 extracellular domain is sufficient for LTP, probably because it mediates binding to neurexins (Nrxs). Examination of surface expression of endogenous AMPA receptors (AMPARs) in cultured neurons suggests that LRRTMs maintain newly delivered AMPARs at synapses after LTP induction. LRRTMs are also required for LTP of mature synapses on adult CA1 pyramidal neurons, indicating that the block of LTP in neonatal synapses by LRRTM1 and LRRTM2 KD is not due to impairment of synapse maturation.

Figure 1

In vivo LRRTM1 and -2 Double Knockdown (DKD) in Neonatal CA1 Pyramidal Neurons Blocks LTP. (A) Schematic of shRNA and replacement lentiviral backbone and the experimental approach for LTP experiments in young mice. (B) Low magnification images of an acute hippocampal slice in DIC (left) and epifluorescence (right) modes showing specific CA1 infection. (C) High magnification (60x) images of the boxed region in (B) showing mosaic GFP expression and a patched CA1 pyramidal neuron. (D) Time courses of representative LTP experiments for control (left) and DKD neurons (right). Averages of 30 consecutive EPSCs during the baseline (1) and 46–50 minutes after tetanic stimulation (2, delivered at time 0) are shown above each graph. (E) Summary time course (left), cumulative fraction of all experiments in the set (middle) and quantification of the LTP magnitude (right) for DKD cells and interleaved controls. In this and all subsequent figures, summary data is presented as mean ± SEM and numbers in parentheses represent number of cells. * p < 0.01. (F–G) As in (D–E), for LTD experiments.

Figure 2

The Extracellular Domain of LRRTM2 is Sufficient for its Function in LTP. (A, C, E, G) Diagrams of the lentiviral vector and recombinant LRRTM2 constructs used for molecular replacement and over expression experiments (top left). Sample, average EPSCs during baseline and after LTP expression (bottom left) and time course (right) of representative, single LTP experiments following the indicated molecular manipulations. (B, D, F, H) Summary time course (left), cumulative fraction of all experiments in the set (middle) and quantification of the LTP magnitude (right) for molecularly-manipulated and corresponding interleaved control neurons. The DKD-LRR2 (A) and DKD-LRR2Ex (E) sets were performed in parallel and therefore share the same control group. For clarity, and to facilitate visual comparison, this control group was plotted in both panels (B and F). *p < 0.01 (See also Figure S1–S3.)

Figure 3

LRRTM1 and -2 DKD Prevents cLTP and Alters Surface GluA1 Expression in Cultured Neurons.(A) Representative images from hippocampal neuronal cultures infected with lentiviruses expressing the indicated constructs and immunostained for the AMPAR subunit GluA1 20 minutes after treatment with control (−cLTP) or glycine-containing solution (+cLTP). (B) Summary graph showing surface GluA1 levels in the three sets of cultures in basal conditions (−cLTP) and following cLTP (+cLTP). Bars represent mean ± SEM. p < 0.0001. (C) Representative images of dendrites from cultured neurons infected with lentiviruses expressing the indicated constructs and immunostained for GFP, GluA1 and vGluT1 in basal conditions and following cLTP. (D–F) Summary quantification of the percentage of GluA1 puncta that are synaptic (D) and the intensity of synaptic (E) and total (F) GluA1 puncta in basal conditions and following cLTP. (*p < 0.001). (G) Schematic of the outside-out voltage clamp configuration and fast glutamate perfusion set up. P = perfusion pipette, R = recording pipette, Con = control solution (ACSF), Glu = ACSF + glutamate, AP-5 and cyclothiazide. (H) Representative glutamate-evoked currents obtained from control and DKD patches (left) and summary quantification (right) of currents. (I) Quantification of surface GluA1 levels in neurons expressing GFP, DKD and DKD-LRR2 in basal conditions and following cLTP induction. An increase in surface GluA1 levels can be detected in DKD neurons 10 but not 20 minutes after cLTP induction. (J) Quantification of the change in relative GluA1 surface levels following cLTP at both time points. (*p < 0.05) (See also Figures S4–S7.)

Figure 4. In Vivo LRRTM DKD Impairs LTP in Young Adult Hippocampus

(A) Diagram showing a mouse on stereotaxic apparatus for injection of lentiviruses at P21. (B) High magnification (60x) of CA1 pyramidal neurons in acute hippocampal slice in DIC (left) and epifluorescence (right) modes showing a patch pipette on an infected neuron from which a whole recording was obtained. (C, D, F, H) Representative EPSCs (left) and time courses (right) of LTP experiments obtained from a control neuron (C) and neurons infected with DKD (D), DKD-LRR2 (F) and DKD-LRR2Ex (H) lentiviruses, respectively. (E, G, I) Summary time course (left), cumulative fraction of all experiments in the set (middle) and quantification of the LTP magnitude (right) for neurons expressing the indicated constructs and the corresponding controls. The DKD and DKD-LRR2 manipulations (D and F) were performed in parallel and share the same group of control neurons. For clarity and to facilitate visual comparison, these control data are plotted in both panels E and G. *p < 0.001