Expression patterns of lgi1b in zebrafish. Semi-qRT-PCR analysis of lgi1 expression levels at different developmental ages (A) or in different zebrafish tissues (B). (C) Significant identity between the zebrafish lgi1a and lgi1b and the human and mouse genes is apparent based on the amino acid sequences within the leucine-rich repeat motif (LRR). The asterisk indicates amino acids conserved in all species. (D) The percentage comparison of identical amino acids.

Splice-site-targeted morpholino oligonucleotides alter normal lgi1b splicing. (A) A schematic diagram of the partial pre-mRNA map showing the predicted outcomes in lgi1b morphants (left panel). The exons are shown in boxes, labeled with the corresponding exon number E1–E3, and the introns are represented by solid lines. The location of the exon 2 splice site, targeted by the MO-E2 morpholino, and the primer binding sites (p1, p2, p3, arrows) for PCR validation are indicated. Semi-qRT-PCR analysis (right panel) of lgi1b transcript levels in wild-type fish, MO-E2mis and MO-E2 morphants at different time points (48 and 72 hpf) show aberrant splicing in the MO-E2 (2 ng) morphants which generates an abnormal (251 bp) mRNA (lacking exon 2) using primers p1/p3. (B) qRT-PCR analysis of lgi1b transcript levels in wild-type, MO-E2mis and MO-E2 morphants following injection of three different concentrations of MO-E2 (2, 3 and 4 ng). (C) Semi-qRT-PCR analysis of transcript levels of the zebrafish lgi1 family members after knockdown of lgi1a (left) and lgi1b (right) respectively.

Morphological changes in 2 ng lgi1b morphants after 48 and 72 hpf.

Compared with wild-type (WT) and MO-E2mis-injected embryos (A, C: lateral view and B: dorsal view), lgi1b morphants showed abnormal head development involving smaller eyes and overt hydrocephalus (arrows) in midbrain and hindbrain ventricles compared with wild-type and control morphants. These phenotypes are more obvious at 72 hpf. (D) Embryos were examined at 48 hpf after injection of two doses of MO-E2 (2 ng and 4 ng) or 4 ng of MO-E2mis. More severe hydrocephalus (demarcated by the dotted frames) and smaller eyes were seen in the high dose (4 ng) lgi1b morphants. (E) The consistent gross head and eye defects and hydrocephalus (left; arrows) in 72 hpf lgi1b morphants were rescued by co-injection of the full length lgi1b mRNA (lateral view), shown in more detail on the right. (F) Comparison of relative ventricle sizes in midbrain and hindbrain based on measurements from 10 embryos per group. ** = p<0.01. Scale bars: A, C and E = 500 μm; B = 50 μm; D = 25 μm.

Confocal imaging analysis of hydrocephalus and apoptosis in lgi1b morphants.

Representative single z-plane images (A) from (left) forebrain and midbrain, (center) hindbrain and (right) Z-stack images show MO-E2 morphants have enlarged ventricles (indicated by *) in both forebrain/midbrain and midbrain/hindbrain compared with mismatch controls (MO-E2mis). Scale bar = 50 μm. (B) Uninjected (WT), MO-E2mis and MO-E2 embryos stained with AO after 48 hpf showed increased cell death (green cells). (C) AO staining analysis showed increased apoptotic cells (White dots) could be detected in the forebrain of the lgi1b morphants after 72 hpf. (D) Whole mount TUNEL analysis further supports increased apoptosis in the forebrain at 72 hpf in lgi1b morphants. Black dots (white arrows) indicate TUNEL-positive cells. (E) AO staining shows apoptosis (arrows) was significantly reduced in lgi1b morphants co-injected with lgi1b mRNA at 36 hpf. (B–D) dorsal view, Z-stack confocal images. (E) a single lateral view image. Scale bars: B, C and D, 100 μm; E, 500 μm.

Synergistic PTZ induction of seizure-like behavior in lgi1b morphants. (A) Longitudinal plot of average activity of 72 hpf embryos over a 2 hour period using Viewpoint modeling software time series plots and a custom smoothing algorithm described by Teng et al (2010). Dotted lines depict the four classes of fish without PTZ treatment (e, f, g, h). Solid lines depict morphants and controls according to key (above). In the absence of PTZ (dotted lines) fish activity (n = 96 for each group) is relatively low in the 2 ng morphants and wild type fish. Upon treatment with PTZ (solid lines) average activity increases significantly in MO-E2 morphants (a) but not in the mismatch morphants (c) or wild type fish (d). Morphants co-injected with rescue lgi1b mRNA show a significant reduction in activity (b) compared with the MO-E3 morphants. (B) Box plots generated in R statistical software showing the distribution of average fish activity over time for different experimental conditions as shown. (C) Comparison of the longitudinal differences in mean values between experimental groups. The red trace follows the moving averages longitudinally (left). Box plots of the same data are shown (right).

Analysis of c-fos expression as a marker for seizure-like behavior. Increased expression of c-fos in lgi1a (MO-E3), but not lgi1b (MO-E2), morphants compared with corresponding mismatch (mis) MO treatment. RNA and protein extracts from 3 dpf embryos injected with 3 ng of the indicated morpholinos or wildtype were used for semi-qRT-PCR (A) and western blot assays (B), respectively. When lgi1 morphants (MO-E2) were treated with PTZ, an increase in c-fos expression was observed. No increase in c-fos was seen when mismatch morphants (MO-E3mis) were treated with PTZ.

Analysis of lgi1a and lgi1b double morphants.

Coincident knockdown of lgi1a and lgi1b results in enhanced phenotype and high mortality rates compared with the single morphants. In (A) 2 ng morphants for lgi1a shows a macroscopically normal phenotype, which is similar to lgi1b morphants, other than the presence of hydrocephalus (arrow). The double mismatch morphant (3 ng of each MO) also appears normal. When the double morphant was analyzed using low dose MO (2 ng of each) small eyes and head together with hydrocephalus (arrow) are observed as well as a curved tail phenotype not seen in single morphants. At high MO doses (3 ng) the curved tail phenotype is enhanced. (B) Quantitative PCR analysis shows that combined treatment with mismatch MO has no effect on either lgi1a or lgi1b mRNA whereas combined low dose (2 ng) treatment results in a 35–50% reduction in mRNAs from the two genes which is even further reduced after high dose (3 ng) treatment. In (C) the mortality in both low dose and high dose combined treatment is significantly different than in the single morphants or combined mismatch morphants. After 48–72 hpf the mortality is even further increased.

48 hpf lgi1b morphants (lateral view) injected with high dose (4 ng) MO-E2 show abnormal developmental phenotypes with more severe hydrocephalus (arrow), small eyes and curved tails (arrow head). Embryos injected with the MO-E2mis did not show these phenotypes. Scale bar: 500 μm.

(A) RT-PCR analysis shows that mRNA levels in lgi1b morphants (2 ng) recover after 4 dpf. (B) Severe hydrocephalus (arrow), heart edema (arrow head) and smaller eyes were still observed at 5 dpf in lgi1b morphants (lateral view). Scale bar: 500 μm.

Immunofluorescence analysis of BrdU incorporation in MO-E2 vs. MO-E2mis injected embryos at 48 hpf. Comparison of proliferating cells in forebrain regions of lgi1b knockdown morphants shows no significant difference from that in control morphants. Scale bar: 50 μm.