Figure 1.
Nfix+/− mice have reduced hippocampal expression of NFIX.
Anti-NFIX staining of wild-type and Nfix+/− coronal brain sections at embryonic, postnatal and adult ages (A–H’). In E14 wild-type mice (A) NFIX was detected within two regions of the hippocampal ventricular zone, the ammonic neuroepithelium (arrow in A) and the dentate neuroepithelium (arrowhead in A), as well as by post-mitotic neurons within the developing hippocampal anlage (double arrowhead in A). At P2, NFIX expression was detected in the wild-type dentate gyrus (arrow in C), CA3 (double arrowhead in C) and CA1 pyramidal cell layers (arrowhead in C) but was detected only in the dentate gyrus at P21 (arrow in E). In adult wild-type mice (G), NFIX was most strongly expressed by cells in the SGZ of the dentate gyrus (arrowheads in G’). The same expression pattern was observed in Nfix+/− mice (B, D, F, H, H’), although the level of NFIX expression was much lower than in wild-type mice at all ages. The specificity of the NFIX antibody is demonstrated by the absence of immunoreactivity in Nfix−/− knockout tissue at E17 (J, compare to panel I). G’ and H’ are magnified views of the boxed regions in G and H respectively. Scale bar (in J): A, B, I, J 150 µm; C, D 250 µm; E, F, G, H 500 µm; G’, H’ 50 µm.
Figure 2.
The dentate gyrus is reduced in E18 Nfix+/− mice.
Haematoxylin (A, B) and anti-PROX1 staining (C, D) of E18 wild-type and Nfix+/− coronal brain sections. The overall morphology of the hippocampus in Nfix+/− mice at E18 (B) was normal compared with that of wild-type mice (A), except that the dentate gyrus appeared smaller (dentate gyrus delineated by dashed lines in A, B). PROX1 immunohistochemistry (C, D) confirmed that the overall dentate gyrus area was smaller in Nfix+/− mice at this age (***p<0.001, E). No difference was detected for the number of PROX1-positive cells per unit area of the dentate gyrus (p = 0.19, F). Scale bar (in D): A, B 200 µm; C, D 100 µm.
Figure 3.
GFAP and GLAST immunoreactivity is reduced in E18 Nfix+/− mice.
Anti-GFAP (A–F) and anti-GLAST staining (G, H) of E18 wild-type and Nfix+/− coronal brain sections. There was reduced GFAP immunoreactivity in the hippocampus of Nfix+/− mice (B, D, F) compared to that of their wild-type littermates at E18 (A, C, E). This reduction in GFAP staining was evident along the rostrocaudal axis of the hippocampus in the supragranular glial bundle (arrowhead), fimbrial glial bundle (double arrowhead) and CA regions (arrow). There was also reduced staining for GLAST in the CA regions and supragranular glial bundle of Nfix+/− mice at E18 (compare panels G and H), although immunoreactivity appeared darker in the fimbrial glial bundle. Scale bar (in H): A–H 250 µm.
Figure 4.
Hippocampal development is delayed in P2 Nfix+/− mice.
Haematoxylin (A–B’) and anti-GFAP staining (C and D) of P2 wild-type and Nfix+/− coronal brain sections. In wild-type mice (A and A’) the suprapyramidal blade (arrow in A’) and infrapyramidal blade (arrowhead in A’) of the dentate gyrus had begun to resolve, whereas in Nfix+/− mice (B and B’) the dentate gyrus remained amorphous. GFAP staining was also reduced in Nfix+/− mice (D) relative to wild-type controls at P2 (C). A’ and B’ are magnified views of the boxed regions in A and B respectively. Scale bar (in D): A, B, C, D 250 µm; A’, B’ 50 µm.
Figure 5.
Hippocampal morphology is aberrant in P21 Nfix+/− mice.
Haematoxylin (A and B), anti-PROX1 (C and D) and anti-GFAP staining (E–F’) of coronal sections of P21 wild-type and Nfix+/− brains. In Nfix+/− mice the CA1 pyramidal cell layer had a more pronounced dorsal curvature (arrowhead in B) than in wild-type mice (A). The dentate gyrus was also misshapen in that the lateral region of the suprapyramidal blade met the medial region at a more acute angle (arrow in B and D) than in wild-type mice (A and C). Although no gross difference in GFAP staining was apparent at P21 (E and F), there appeared fewer GFAP-positive radial fibres in the dentate gyrus of Nfix+/− mice (arrowheads in F’) than wild-type mice (arrowheads in E’). E’ and F’ are magnified views of the boxed regions in E and F respectively. Scale bar (in F’): A, B 275 µm; C, D 180 µm; E, F 250 µm; E’, F’ 50 µm.
Figure 6.
The hippocampus of adult Nfix+/− mice remains misshapen.
No gross differences in the volume of hippocampal structures were detected between adult wild-type and Nfix+/− mice (A) using magnetic resonance imaging at a resolution of 30 µm (p>0.5). Haematoxylin staining of coronal brain sections (B and C) revealed that the CA1 pyramidal cell layer (arrowhead in C) and dentate gyrus (arrow in C) remained misshapen in adult Nfix+/− mice. Moreover the suprapyramidal blade of the dentate gyrus appeared thinner in adult Nfix+/− mice (C’) compared with wild-type mice (B’). B’ and C’ are magnified views of the boxed regions in B and C respectively. Scale bar (in C’): B, C 500 µm; B’, C’ 70 µm. Abbreviations: CA = Cornu ammonis; Py = pyramidal layer; Or = oriens layer; Rad = radiatum layer; MoDG = molecular layer of the dentate gyrus; GrDG = granule layer of the dentate gyrus; PoDG = polymorph layer of the dentate gyrus.
Figure 7.
Hippocampal neurogenesis is impaired in adult Nfix+/− mice.
Anti-PROX1 (A–B’), anti-GFAP (C–D’) and anti-DCX (E–F’) staining of coronal sections of adult wild-type and Nfix+/− brains. Immunostaining for PROX1 and subsequent quantification (A–B’) revealed that the thickness of the suprapyramidal blade of the dentate gyrus was reduced in adult Nfix+/− mice compared with wild-type mice (**p<0.01; G). There were also fewer GFAP-positive radial fibres (**p<0.01; H) and DCX-positive neurons (***p<0.001; I) per unit length of the dentate gyrus in adult Nfix+/− mice. A’, B’, C’, D’, E’ and F’ are magnified views of the boxed regions in A, B, C, D, E and F respectively. Scale bar (in F’) A, B, C, D, E, F 300 µm; A’, B’, C’, D’, E’, F’ 75 µm.
Figure 8.
Nfix+/− mice exhibit normal motor function and anxiety-related behaviour.
Performance of adult wild-type and Nfix+/− mice during the open-field test (A, B) and the elevated plus maze (C, D). In the open-field test there was no effect of genotype on the distance travelled over the 10 minute test period (p>0.5; A), or on time spent in the centre of the test apparatus (p = 0.43; B). In the elevated plus maze no effect of genotype was observed on the percentage of total trial time spent on the open arms, closed arms or the centre position of the maze (p>0.5; C), or on number of open arm entries (p = 0.29; D). (Nfix+/+: n = 20, Nfix+/−: n = 26).
Table 1.
Behaviour and appearance in viewing jar
Table 2.
Behaviour above viewing jar.
Table 3.
Body weight.
Figure 9.
Nfix+/− mice have impaired spatial learning and memory.
Training latencies (A) and probe trial performances (B–E) of adult wild-type and Nfix+/− mice during the Morris water maze task. Over the initial 5 days of hidden-platform training (days 1–5) Nfix+/− mice took significantly longer, on average, than wild-type mice to find the platform (**p<0.01), taking almost twice as long on training day 5 (**p<0.01; A). During the reversal learning procedure (days 7–8) Nfix+/− mice took longer to find the platform on training day 8 (*p<0.05; A). The average search path of Nfix+/− mice was further away from the trained position of platform during the probe trial on day 6 and day 9 (*p<0.05; C). A similar trend was observed for time spent in target quadrant (B). A representative performance of a wild-type mouse and a Nfix+/− mouse during the probe trial on day 6 is pictured in (D) and (E) respectively. (Nfix+/+: n = 20, Nfix+/−: n = 26).