Functionally distinct groups of inherited PTEN mutations in autism and tumour syndromes

Abstract

Background: Germline mutations in the phosphatase PTEN are associated with diverse human pathologies, including tumour susceptibility, developmental abnormalities and autism, but any genotype-phenotype relationships are poorly understood.

Methods: We have studied the functional consequences of seven PTEN mutations identified in patients diagnosed with autism and macrocephaly and five mutations from severe tumour bearing sufferers of PTEN hamartoma tumour syndrome (PHTS).

Results: All seven autism-associated PTEN mutants investigated retained the ability to suppress cellular AKT signalling, although five were highly unstable. Observed effects on AKT also correlated with the ability to suppress soma size and the length and density of dendritic spines in primary neurons. Conversely, all five PTEN mutations from severe cases of PHTS appeared to directly and strongly disrupt the ability to inhibit AKT signalling.

Conclusions: Our work implies that alleles causing incomplete loss of PTEN function are more commonly linked to autism than to severe PHTS cases.

Keywords: Cancer: breast; Cell biology; Neurosciences.

Figure 1

PTEN lipid phosphatase activity. (A) A diagram representing PTEN protein structure and the localisation of the autism-related mutations analysed. (B) PTEN crystal structure indicating mutations. The catalytic Cys-124 is yellow. (C) U87MG cells were transduced with different units (low to high) of lentivirus particles encoding for PTEN WT or PTEN mutants. Control cells were transduced with viruses encoding for GFP and treated with the phosphoinositide 3-kinase (PI3K) inhibitor PI103 (1 µM for 30 min). PTEN expression and AKT phosphorylation were investigated by western blotting of cell lysates. The presented blotting panels for each protein are all from one membrane and are representative of three independent experiments. (D and E) The level of PTEN gene expression was measured by qPCR in U87MG cells transduced with an equal amount of lentivirus particles (two units) encoding for PTEN WT or mutants. (D) Relative PTEN expression was detected using real-time qPCR. Data are shown as mean PTEN ΔCt values relative to GAPDH±SEM. from three experiments each performed in duplicate. (E) Parallel PTEN protein immunoblot. GFP, green fluorescent protein; GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

Figure 2

Autism-associated PTEN mutations reduce protein stability. The stability of PTEN mutation proteins was determined using the cycloheximide inhibitor. (A) U87MG cells expressing PTEN WT or PTEN point mutants were treated with 200 µg/mL of the inhibitor cycloheximide for the indicated times, followed by immunoblotting analysis with anti-PTEN and anti-GAPDH antibodies. The blots shown are representative of three experiments. (B) Plot is densitometric quantitation of the cycloheximide chase assay. Data are shown as PTEN level normalised to GAPDH at each time point and compared with that at the 0 time point. The quantitation of the cycloheximide inhibitor assay is derived from three independent experiments (mean±SEM). Six independent experiments for PTEN WT, C124S, H118P and G129R (including the data shown in figure 4D, E). ****p<0.0001, ***p<0.001, **p<0.01, *p<0.05 compared with PTEN WT (Student's t test using GraphPad Prism software).

Figure 3

Characterisation of autism-related PTEN mutants in primary neurons. Re-expression at physiological level of PTEN WT and mutants in PTEN knockout neurons is able to support wild type neuronal morphology. (A) Workflow of the experiment. Neurons were transduced after 7 days with viruses encoding Cre recombinase (RFP-Cre) and PTEN WT or mutants as indicated. Cells were lipofected with yellow fluorescent protein (YFP) and then fixed and immunostained 6 days after transduction for GFP (to detect YFP), PTEN and for RFP (RFP-Cre) as showed in the representative images in (B) (scale bars 10 µm). (C) The soma size was measured using ImageJ software and the area measured (green lines) is represented in µm 2±SEM (n=9 cells for each genotype). (D) Parallel neuronal cell lysates were used for western blotting to verify the efficiency of the knockout and the re-expression of PTEN. (E–G) Effects of PTEN expression on spine density and length. (E and F) Quantification of the effects of PTEN genotype on spine density (n=8 dendrites for each genotype) and spines length (n=40 spines measured for each mutant) as mean±SEM. Spines were counted on a dendrite length of 70 µm. ****p<0.0001, ***p<0.001, **p<0.01, *p<0.05 compared with PTEN WT. (Dunnett's post hoc test following a one-way ANOVA using GraphPad Prism software). (G) Representative images of dendritic spines (scale bars 10 µm).

Figure 4

Missense PTEN mutations associated with severe PTEN hamartoma tumour syndrome (PHTS) phenotypes cause complete loss of function. (A) Studied mutations in PTEN associated with PHTS. (B) PTEN crystal structure with catalytic Cys124 in yellow and indicated mutations. (C) PTEN null U87MG cells were transduced with lentiviruses encoding PTEN WT or mutants. Specific lentiviral doses were used for each mutant to standardise PTEN expression level (5 units are equal to 50 µL of viral supernatant). Cell expressing GFP and cells treated with the phosphoinositide 3-kinase (PI3K) inhibitor PI103 (1 µM, 30 min) were controls. PTEN expression and AKT and PRAS40 phosphorylation were investigated by immunoblotting of total cell lysates. (D and E) U87MG cells were transduced with PTEN WT or mutants, treated with 200 µg/mL cycloheximide, lysed at the indicated times and then immunoblotted for PTEN and GAPDH. (D) PTEN protein levels normalised to GAPDH from three independent experiments (mean±SEM). Six independent experiments for PTEN WT, C124S, H118P and G129R (including the data shown in figure 2). ****p<0.0001, ***p<0.001, **p<0.01, *p<0.05 compared with PTEN WT (Student's t test using GraphPad Prism software). (E) The blots shown are representative of three independent experiments.