Dendritic spine head diameter is reduced in the prefrontal cortex of progranulin haploinsufficient mice

Abstract

Loss-of-function mutations in the progranulin (GRN) gene are an autosomal dominant cause of Frontotemporal Dementia (FTD). These mutations typically result in haploinsufficiency of the progranulin protein. Grn^+/- mice provide a model for progranulin haploinsufficiency and develop FTD-like behavioral abnormalities by 9-10 months of age. In previous work, we demonstrated that Grn^+/- mice develop a low dominance phenotype in the tube test that is associated with reduced dendritic arborization of layer II/III pyramidal neurons in the prelimbic region of the medial prefrontal cortex (mPFC), a region key for social dominance behavior in the tube test assay. In this study, we investigated whether progranulin haploinsufficiency induced changes in dendritic spine density and morphology. Individual layer II/III pyramidal neurons in the prelimbic mPFC of 9-10 month old wild-type or Grn^+/- mice were targeted for iontophoretic microinjection of fluorescent dye, followed by high-resolution confocal microscopy and 3D reconstruction for morphometry analysis. Dendritic spine density in Grn^+/- mice was comparable to wild-type littermates, but the apical dendrites in Grn^+/- mice had a shift in the proportion of spine types, with fewer stubby spines and more thin spines. Additionally, apical dendrites of Grn^+/- mice had longer spines and smaller thin spine head diameter in comparison to wild-type littermates. These changes in spine morphology may contribute to altered circuit-level activity and social dominance deficits in Grn^+/- mice.

Keywords: Dendritic spines; Frontotemporal Dementia; Prefrontal cortex; Progranulin; Social dominance.

Fig. 1

Dendritic spine morphology changes in Grn^+/– mice. A) Grn^+/– mice showed a low-dominance phenotype in the tube test at 10 months of age (n = 5–6 mice, 2–3 mice of each sex per group, two-tailed Mann-Whitney U test, p = 0.0368). B) Representative 20X image of a Lucifer yellow injected neuron (scale bar represents 100 μm) and imaging parameters. Representative 60X dendritic segment with its corresponding reconstruction (scale bar represents 5 μm). Thin spines are shown in blue, mushroom spines in green, stubby spines in orange, filopodia in yellow. C–F) The distribution of spine types on apical dendrites of Grn^+/– mice was shifted relative to Grn^+/+ mice (C, D, chi-square, p = 0.0186) but there was no difference in the distribution of spine types on basal dendrites between genotypes (E, F, chi-square, p = 0.3208). G–J) Grn^+/– mice had decreased dendritic spine head diameter on apical dendrites (G, K-S test, p = 0.0001, H, chi-square, p = 0.0003), but not on basal dendrites (I, K-S test, p = 0.1608, J, chi-square, p = 0.2611). K–N) Grn^+/– mice showed increased spine length on apical dendrites (K, K-S test, p = 0.0356, L, chi-square, p = 0.014). On basal dendrites, analysis of cumulative frequency distribution (M, K-S test, p = 0.034), though not relative frequency distribution (N, chi-square, p = 0.3849) indicated an increase in spine length. O-R) Grn^+/– mice had reduced thin spine head diameter on apical dendrites (O, K-S test, p = 0.0003, P, chi-square, p = 0.0029), but not basal dendrites (Q, K-S test, p = 0.1868, R, chi-square, p = 0.1692). S-V) Grn^+/– mice did not exhibit a significant increase in thin spine length on the apical dendrites (S, K-S test, p = 0.1048, T, chi-square, p = 0.0669). On basal dendrites, analysis of cumulative frequency distribution (U, K-S test, p = 0.0197), but not relative frequency distribution (V, chi-square, p = 0.3040) revealed an increase in thin spine length. n = 2175–3325 thin spines, 336–370 mushroom spines, 280–541 stubby spines, 40–68 filopodia per apical or basal dendrite for each genotype, from 5–6 mice per genotype. Relative frequency distributions are shown as Gaussian curve fits. * = p < 0.05, ** = p < 0.01, *** = p < 0.0001