Mutation of Semaphorin-6A disrupts limbic and cortical connectivity and models neurodevelopmental psychopathology

Abstract

Psychiatric disorders such as schizophrenia and autism are characterised by cellular disorganisation and dysconnectivity across the brain and can be caused by mutations in genes that control neurodevelopmental processes. To examine how neurodevelopmental defects can affect brain function and behaviour, we have comprehensively investigated the consequences of mutation of one such gene, Semaphorin-6A, on cellular organisation, axonal projection patterns, behaviour and physiology in mice. These analyses reveal a spectrum of widespread but subtle anatomical defects in Sema6A mutants, notably in limbic and cortical cellular organisation, lamination and connectivity. These mutants display concomitant alterations in the electroencephalogram and hyper-exploratory behaviour, which are characteristic of models of psychosis and reversible by the antipsychotic clozapine. They also show altered social interaction and deficits in object recognition and working memory. Mice with mutations in Sema6A or the interacting genes may thus represent a highly informative model for how neurodevelopmental defects can lead to anatomical dysconnectivity, resulting, either directly or through reactive mechanisms, in dysfunction at the level of neuronal networks with associated behavioural phenotypes of relevance to psychiatric disorders. The biological data presented here also make these genes plausible candidates to explain human linkage findings for schizophrenia and autism.

Figure 1. Prefrontal cortex dysconnectivity in Sema6A mutants.

We visualized pAC projections using the placental alkaline phosphatase (PLAP) marker encoded in the Sema6A mutant allele . During development, PLAP is expressed in the neurons that form the pAC (Figure 1A, B), while in adults, the PLAP marker is expressed by oligodendrocytes and thus labels all myelinated axons (Figure 1C, D). In newborn Sema6A^−/− mice (B), the PLAP-stained pAC axons travel ventrally into hypothalamic areas and fail completely to cross the midline as seen in Sema6A^+/− mice (A). Such misrouted pAC axons are still present in adult Sema6A^−/− brains (C, compare with Sema6A^+/− in D) and can be found through extended levels of the caudo-rostral axis within inappropriate regions, such as the septum (E-I, arrowheads). The aAC and stAC are normal in young and adult Sema6A^−/−.a/p/stAC: anterior/posterior/stria terminalis arm of the anterior commissure; Pir: piriform cortex; Scale bar in (A) is for (A, B), in (C) is for (C, D), and in (E) is for (E–I): 500 µm.

Figure 2. Defects in piriform cortex and olfactory projections.

(A, B) NeuN-immunohistochemistry of coronal P10 brains sections demonstrates a greatly exaggerated folding of the piriform cortex in Sema6A^−/− (B), compared to Sema6A^+/− mice (A). (C–F) E16 Sema6A^+/− (C, D) and Sema6A^−/− (E, F) brain sections immunostained for L1 (green) to label all fibres and for Neuropilin-1 (NP-1) to label more specifically LOT axons. In Sema6A^−/− mutants, the L1/NP-1 double-stained LOT (yellow) appears rounded and more embedded into layer 1 of the piriform cortex (E) in contrast to a superficially located LOT in Sema6A^+/− mice (C). In addition, the LOT extends far further caudally than normal (compare F and D). (G–N) Series of PLAP-stained adult brain sections, showing persistence of a more rounded and displaced LOT (arrowheads) that extends further caudally in Sema6A^−/− mutants (K–N) when compared with Sema6A^+/− mice (G–J). Note also that adult Sema6A^−/− mutants show an exaggerated folding of the pirifom cortex, which extends far caudally. Ia: agranular insular cortex; lot: lateral olfactory tract; Pir: piriform cortex. Scale bar in (A) is for (A, B): 200 µm; in (C) is for (C–D): 100 µm; in (G) is for (G–N): 500 µm.

Figure 3. Aberrant lamination of the neocortex.

NeuN-immunostaining of coronal sections of adult Sema6A^+/− (A, A1–3) and Sema6A^−/− (B, B1–3) mouse brains in overview (A, B) and detailed view (A1–3, B1–3; position in A, B is indicated as boxes). While the border between layers 1 and 2 of the neocortex is sharply visible in Sema6A^+/− animals (A1–3), it is largely obliterated in Sema6A^−/− mutants (B1–2), and the neuropil of layer 1 appears infiltrated by neurons from deeper layers with a gradient in severity from lateral to medial (B). Laterally, many ectopic neurons form repetitive clusters (asterisks in B1) bordered by neuropil. More medially these neurons are more loosely scattered within the neuropil over a considerable depth (arrowheads in B2, B3). 1–6: Cortical layers 1 to 6. Au: auditory area; S1: primary somatosensory area; PtA: parietal association area. Scale bar in (A) is for (A, B): 500 µm; in (A1) is for (A1–3,B1–3): 200 µm.

Figure 4. Alterations in hippocampal lamination and projections.

(A–F) Immunohistochemistry for NeuN (A, D; green, neurons) and Prox1 (B, E; red, dentate granule cells; overlay in C and F) on coronal brain sections from P10 Sema6A^+/− (A–C) and Sema6A^−/− (D–F) mice. Sema6A^−/− mice show a broadening of the granule cell layer at the tip of the infrapyramidal blade of the dentate gyrus as well as more isolated clusters of ectopic granule cells at the surface of the molecular layer of the same blade (arrows in D–F). (G–H) Overlays of NeuN- (green) and Prox1- (red) immunostained adult brain sections of Sema6A^+/− (G) and Sema6A^−/− (H–J). Similarly distributed ectopic granule cells (arrows in H–J) are still present in the adult dentate gyrus. (K–L) PLAP-stained adult brain sections showing a reduced and defasciculated fornix (Fx) in Sema6A^−/− (L, left fornix) compared with Sema6A^+/− mice (K). The fornix is significantly reduced in size by 20% (M; p<0.05) and is significantly less compact (N; measured as the ratio of actual and circular perimeter; p<0.01) in Sema6A^−/− (n = 10) compared to Sema6A^+/− mice (n = 8). Error bars in (M, N) represent ± SD; *: significant. Dots plotted on the right of columns in (N) are individual data points. Scale bar in (A) is for (A–F), in (G) is for (G–J), and in (K) is for (K, L): 200 µm.

Figure 5. Altered cortical activity patterns.

(A) Representative bilateral cortical EEG traces from Sema6A^−/− and Sema6A^+/− animals in spontaneous, freely moving conditions. (B) Quantification of total spectral power in Sema6A^+/− (n = 3) and Sema6A^−/− (n = 4) animals, p<0.005. (C) Quantification of relative power of spectral bands in Sema6A^+/− (n = 7) and Sema6A^−/− (n = 8); a significant increase in the alpha band was observed p<0.001. (D) Increase in relative power in the alpha band in Sema6A^−/− is reversed by administration of clozapine, p<0.05. No significant effect of clozapine on alpha power in Sema6A^+/− animals was observed (n = 4 for all groups). Clozapine at this dose did not affect the other spectral power bands in either genotype (not shown). *p<0.05; ns: non-significant. Error bars represent ± SEM.

Figure 6. Gait abnormalities and hyperlocomotion.

(A) Gait analysis indicated significantly shorter hind-stride length in Sema6A^−/− mutants compared to controls (p<0.05). (B) There is a trend towards decreased hind-front overlap in Sema6A^−/− mutants (p = 0.07). (C) Sema6A^−/− mutants displayed increased number of ambulatory counts in the open-field over a six-hour period (n = 20 per genotype; p<0.05). (D) Sema6A^−/− mutants demonstrated increased exploratory behaviour in the open-field, as indexed by greater distance travelled over a six-hour period (p<0.05). (E) Clozapine (0.25 mg/kg) significantly reversed the hyper-exploratory phenotype in Sema6A^−/− mutants without altering exploration in controls (n = 8 per treatment with vehicle or drug, and per genotype; p<0.05). (F) MK-801-induced hyperlocomotion (0.2 mg/kg) did not differ across the genotypes (n = 8 per treatment with vehicle or drug, and genotype). *p<0.05, **p<0.01; ns: non-significant. Error bars represent ± SEM.

Figure 7. Memory defects, decreased anxiety, and increased social interaction.

(A) In the novel object recognition task, Sema6A^−/− mutants demonstrated increased absolute levels of exploration of a novel object at both retention intervals (1 hr, 4 hr; WT: n = 9, Sema6A ^−/−: n = 8; p<0.05). (B) Sema6A^−/− mutants displayed a disruption in novel object recognition memory, as indexed by decreased relative preference to explore the novel object at the 1-hr retention interval, when compared to controls (p<0.05). No interaction with sex was observed for either of these effects. NOR performance at the 4-hr retention interval did not differ from chance (50%) across the genotypes. (C) Spontaneous alternation test of working memory revealed a sex-specific deficit in working memory performance in Sema6A^−/− males (n = 8), as indicated by a decrease in percent alternation, relative to controls (n = 10; p<0.01). (D) A hyper-exploratory phenotype of male Sema6A^−/− mutants was demonstrated by increase in number of overall arm entries in the alternation task relative to controls (p<0.05). (E) In the elevated plus maze test of anxiety, male Sema6A^−/− mutants exhibit decreased anxiety as measured by an increase in the number of open arm entries (WT: n = 10 Sema6A ^−/−: n = 9; p<0.05). (F) Sema6A^−/− mutants demonstrate an increase in number of overall arm entries in the elevated plus maze relative to controls (p<0.01). (G) Sema6A^−/− mutants of both sexes (n = 19) exhibited decreased anxiety in the marble burying task, as indicated by a decreased number of marbles buried compared to controls (n = 22 WT, 19 Sema6A^−/−; p<0.05). (H) Assessment of social interaction with a novel conspecific revealed a decrease in duration of time spent engaged in walkover behaviours (p<0.05) in Sema6A^−/− mutants (n = 11) relative to controls (n = 9). (I) Sema6A^−/− mutants displayed an increase in number of anogenital sniffing (p<0.05) and decrease in walkover (p<0.01) episodes relative to controls. An increase in general exploration of the novel environment was also observed (wall and free rearing). Error bars represent ± SEM.