The Interactome of Palmitoyl-Protein Thioesterase 1 (PPT1) Affects Neuronal Morphology and Function

Abstract

Palmitoyl-protein thioesterase 1 (PPT1) is a depalmitoylation enzyme that is mutated in cases of neuronal ceroid lipofuscinosis (NCL). The hallmarks of the disease include progressive neurodegeneration and blindness, as well as seizures. In the current study, we identified 62 high-confident PPT1-binding proteins. These proteins included a self-interaction of PPT1, two V-type ATPases, calcium voltage-gated channels, cytoskeletal proteins and others. Pathway analysis suggested their involvement in seizures and neuronal morphology. We then proceeded to demonstrate that hippocampal neurons from Ppt1-/- mice exhibit structural deficits, and further investigated electrophysiology parameters in the hippocampi of mutant mice, both in brain slices and dissociated postnatal primary cultures. Our studies reveal new mechanistic features involved in the pathophysiology of this devastating neurodegenerative disease.

Keywords: PPT1; hippocampal neurons; mass-spectrometry; palmitoyl-protein thioesterase 1; palmitoylation.

Figure 1

(A) Schematic presentation of the proteomics work flow. From left to right: HEK 293 cells were transfected with either GFP-PPT1 or GFP expression plasmids. Cell were lysed, cell lysates were immunoprecipitated with anti-GFP antibodies coupled to protein A/G beads. The beads were loaded onto a column, washed and then protein lysates from P0 cortices derived from either Ppt1+/+ or Ppt1−/− were applied to the GFP- and to the GFP-PPT1- bound beads and then washed. The bound proteins were eluted. The samples were then subjected to MS. The list of the identified proteins was then subjected to bioinformatic analysis. (B,C) Characterization of palmitoyl-protein thioesterase 1 (PPT1)-interacting proteins. (B) Among the total of 64 PPT1 interacting proteins, all contain at least one cysteine residue, 45 (70%) are known to be palmitoylated, nine (14%) contain sites that are predicted to be palmitoylated. (C) Ingenuity pathway analysis revealed that the high confidence proteins may contribute to neuronal morphology and to seizures. Proteins colored in red were enriched in the fractions from Ppt1−/− and proteins colored in green were enriched in fractions from wild-type. Note: the origin of PPT1 is from the HEK293 cells used for the initial immunoprecipitation.

Figure 2

Reduced complexity of cultured Ppt1−/− hippocampal neurons. The number of dendrites crossing concentric rings with regular radial increments centered in the neuronal cell body (Sholl analysis) revealed a more simplified dendritic tree in cells lacking Ppt1−/− (B,C) in comparison to cells derived from control littermates (A,C). Calibration bar, 100 μM.

Figure 3

Spine density is modified in Ppt1−/− cultured hippocampal neurons. (A–C) Hippocampal neurons form control animals (A,A’) or Ppt1−/− littermates (B,B’) were transfected with GFP and grown for 14 days in culture. Dendrites (green) and dendritic spines (blue) were traced using Imaris Filament Tracer tool (A’,B’). (C) The numbers of spines branching off dendritic segments were computed from n = 6 cells in each group. Averages of the number of spines per each dendritic segment (branch points are highlighted in purple, see “Materials and Methods” section for details) and SEM are plotted, **P = 0.005. (D–L) Hippocampal neurons transfected with GFP (green, D,G) to mark dendritic spines, and immuno-stained with PSD95 (red, in E,H) to mark post-synaptic sites and SYNCRIP (SYP) to locate positive puncta along the dendrites. (J,K) SYNCRIP signal (Red) in control or Ppt1−/− hippocampal neurons, was masked outside the neurons, showing puncta accumulating in the head of control dendritic spines (J) vs. Ppt1 KO cells (K). (L) Colocalization of SYNCRIP in n = 36 WT and n = 29 Ppt1 KO dendritic spines, was measured and calculated as percentage of the general GFP signal, *P = 0.0101. The Dot plot chart is depicting all measured spines. Lines indicate average and SEM. Size markers (A’) 10 μM, (G) 1 μM, (K) 2 μM.

Figure 4

Summary of recordings from wild-type controls and Ppt1 KO cells. (A) Illustration of spontaneous miniature EPSCs recorded from a cultured neuron voltage clamped at −60 mV. Left, a continuous record at low speed, right a zoomed trace taken from the left record. Scale 20 pA for both, 10 s and 5 ms for the left and right traces, respectively. (B) Means of discharge rates of 16 controls and 18 Ppt1−/− cells. (A) Significant reduction in discharge rate is seen (t-test, *p < 0.02). (C,D) No difference in amplitudes (C) or rise time (D) between the same two groups is detected.

Figure 5

Electrophysiology of transverse hippocampal slices from control and Ppt1 KO animals. (A) Input-output curves of field excitatory postsynaptic potential (fEPSP) slopes for WT and Ppt1 KO mouse hippocampal slices showing no significant difference between groups. (B) Paired-pulse facilitation (PPF) of the fEPSP slopes expressed as response to the 2nd stimulation over the 1st at different interpulse intervals (IPIs; 10, 20, 40, 60, 80, and 100 ms) in WT and PPT1-KO slices. No significant difference in PPF was found. (C) Short tetanic stimulation, subthreshold to long-term potentiation (LTP) induction (35 pulses, 100 Hz), was delivered twice to one pathway. The arrows denote the points at which the tetanic stimulation was delivered. The stimulation produced a transient, short term potentiation (STP) which was larger in WT than Ppt1 KO slices (1.59 ± 0.05 and 1.37 ± 0.06 (n = 4 in both groups, P < 0.05), correspondingly). (D) LTP was induced by high-frequency stimulation, (HFS, 100 Hz, 1 s) at twice the test intensity. The arrows denote the points at which tetanic stimulation was delivered twice to one pathway. The second train produced a further increase in the response. Ppt1 KO slices expressed a significantly lower magnitude LTP than WT [1.38 ± 0.004 (n = 7; P < 0.001, F value 201.16) compared with 1.71 ± 0.01 (n = 7) in WT].