The impact of phosphorylated PTEN at threonine 366 on cortical connectivity and behaviour

Abstract

The lipid phosphatase PTEN (phosphatase and tensin homologue on chromosome 10) is a key tumour suppressor gene and an important regulator of neuronal signalling. PTEN mutations have been identified in patients with autism spectrum disorders, characterized by macrocephaly, impaired social interactions and communication, repetitive behaviour, intellectual disability, and epilepsy. PTEN enzymatic activity is regulated by a cluster of phosphorylation sites at the C-terminus of the protein. Here, we focused on the role of PTEN T366 phosphorylation and generated a knock-in mouse line in which Pten T366 was substituted with alanine (PtenT366A/T366A). We identify that phosphorylation of PTEN at T366 controls neuron size and connectivity of brain circuits involved in sensory processing. We show in behavioural tests that PtenT366/T366A mice exhibit cognitive deficits and selective sensory impairments, with significant differences in male individuals. We identify restricted cellular overgrowth of cortical neurons in PtenT366A/T366A brains, linked to increases in both dendritic arborization and soma size. In a combinatorial approach of anterograde and retrograde monosynaptic tracing using rabies virus, we characterize differences in connectivity to the primary somatosensory cortex of PtenT366A/T366A brains, with imbalances in long-range cortico-cortical input to neurons. We conclude that phosphorylation of PTEN at T366 controls neuron size and connectivity of brain circuits involved in sensory processing and propose that PTEN T366 signalling may account for a subset of autism-related functions of PTEN.

Keywords: PTEN; cognitive behaviour; cortical connectivity; developmental disorders; neuronal morphology.

Figure 1

Behaviour in Pten^T366A/T366A mice. (A) Pten^T366A/T366A and wild-type (wt) siblings tested for general exploratory behaviour in the open field. (B and C) Motor skills as determined in the hanging wire test (B) and in the rotarod (C). (D) Percentage of total responses to mechanical stimulation in the von Frey test in Pten^T366A/T366A and wild-type siblings. (E) Percentage of spontaneous alterations in the Y-maze in female and in male Pten^T366A/T366A and wild-type siblings. (F) Percentage of time that Pten^T366A/T366A and wild-type mice spent in the target area and other quadrants of the Barnes maze. (G) Percentage of time that Pten^T366A/T366A and wild-type mice spent in the target area and other quadrants of the Morris water maze. (H) Percentage freezing in the conditioned fear-induced freezing in female and in male Pten^T366A/T366A and wild-type siblings. (I) The average number of responses in the vibrissae-stimulated reflex test of Pten^T366A/T366A and wild-type siblings. (K) Grooming time in Pten^T366A/T366A and wild-type mice. Statistical analysis with one-way ANOVA (Bonferroni post hoc test), unpaired t-test, Wilcoxon test (grooming), *P < 0.05. For analysis details see Supplementary Table 3.

Figure 2

Cortical layers and proliferation in Pten^T366A/T366A brains. (A–D) Immunolabelling for cortical layer markers and the percentages of neurons expressing in Pten^T366A/T366A and wild-type (wt) mice. (A–D) The percentage of (A) NeuN+ neurons in cortical layers 1 to 6. (B–E) The percentage of (B) Cux1+ neurons in cortical layers 2–4, (C) CTIP2+ neurons in L5, (D) FOXP2+ neurons in L6 in relation to all neurons counterstained with HOECHST. (E) Proliferation of cortical progenitor cells in Pten^T366A/T366A and wild-type siblings. Schematic for experimental set-up, BrdU injections were performed in time-pregnant heterozygous females at E11.5, E13.5 and E15.5. Analysis was undertaken at P1 and P8. (F) Example images of BrdU labelling and HOECHST counterstain at P1 and P8. Graphs showing distribution of BrdU+ neurons in Pten^T366A/T366A and wild-type mice at P1 and P8, respectively. Each dot in graphs accounts for one brain section. Statistical analysis with unpaired t test and two-way ANOVA, Bonferroni post hoc test. For analysis details see Supplementary Table 4. Scale bars = 100 µm.

Figure 3

Soma size in Pten^T366A/T366A cortical and dentate gyrus neurons. (A) Schematic of AAV-GFP injection at P0 and expression of GFP in adult S1 cortex and dentate gyrus. Soma size was analysed at P14 and P42. Staining with HOECHST and the Allen brain mouse reference atlas were used to define cortical layers (also see the ‘Materials and methods' section). (B and C) GFP-injected neurons at P14 (B) and P42 (C) for Pten^T366A/T366A and wild-type (wt) mice. Enlarged images show somata from pyramidal neurons in L2/3 and L5 and neurons in L4. Graphs for soma size in L2/3 to L5 Pten^T366A/T366A and wild-type neurons at P14 and P42, respectively. (D and E) Analysis of soma size in dentate gyrus neurons at P14 (D) and at P42 (E) in Pten^T366A/T366A and wild-type mice. Each dot in graphs accounts for one cell. Data shown as average ± SEM. Data from three brains each genotype, six sections per brain. Statistical analysis with one-way ANOVA, unpaired t-test for dentate gyrus, ***P < 0.001. Analysis details are provided in Supplementary Table 5. Scale bars = 100 µm, enlarged images in B 20 µm, in C–E, 50 µm.

Figure 4

Dendritic arborization in Pten^T366A/T366A cortical and dentate gyrus neurons. (A–D) Dendritic arborization for L2/3 pyramidal neurons and L4 neurons at P14 (A and B) and P42 (C and D). Enlarged images showing L2/3 and L4 neurons, graphs showing quantification with Sholl analysis. (E and F) Dendritic arborization in hippocampal dentate gyrus neurons at P14 (E) and at P42 (F) with enlarged images, and quantification with Sholl analysis in graphs. Data shown as average ± SEM. Statistical analysis with two-way ANOVA, Bonferroni post hoc test, ****P < 0.0001. Analysis details are provided in Supplementary Table 7. Scale bars = 100 µm, 50 µm in enlarged images.

Figure 5

Presynaptic input to S1 cortex in Pten^T366A/T366A mice. (A) Schematic showing injection procedure of three viruses (also see the ‘Materials and methods' section). AAV1-Cre was injected (injection 1) into POm and VPm thalamus. Two weeks later, the trans-synaptically labelled S1 cortical neurons expressing Cre, i.e. receiving thalamic input, were targeted with a flex AAV to express the TVA receptor (injection 2). Two weeks after injection 2, the TVA-expressing neurons (GFP+ S1 neurons) were targeted with rabies virus (injection 3). The Cre+, TVA-expressing (green) and mCherry+ expressing (red) neurons were the starter neurons (yellow). One week after injection of rabies, brains were harvested, and local and long-range retrograde-labelled neurons could be observed (red presynaptic neurons). (B) Left graph showing number of (Cre+)/GFP+/mCherrry+ (starter) neurons, right graph showing all presynaptic mCherry+ labelled neurons. (C) Examples with enlarged images showing expression of GPF+ neurons, GFP+/mCherry+ (starter) neurons and presynaptic mCherry+ neurons in wild-type and Pten^T366A/T366A S1 cortices. (D) Graph showing the percentage of presynaptic neurons local in S1 cortex and long-range neurons from other cortical and thalamic areas in Pten^T366A/T366A and wild-type brains. (E) Pie charts showing percentages of dissected presynaptic inputs from long-range areas to S1 in Pten^T366A/T366A and wild-type brains. Graph showing input to motor cortices and thalamus in Pten^T366A/T366A and wild-type brains. (F) Example images of presynaptic neurons in motor cortices, thalamus, S1 cortex contralateral and visual cortices in Pten^T366A/T366A and wild-type brains. Higher magnification of single neurons. Statistical analysis with unpaired t-test, *P < 0.05. Analysis details are provided in Supplementary Table 8. Scale bars = 100 µm, 50 µm in enlarged images.