Zebrafish App protein organization, gene expression, and mutant generation

(A) There are 2 App orthologs, appa and appb, in zebrafish. Appa contains the Kunitz-type protease inhibitor (KPI) domain and thus has a similar gene organization to the human APP770 isoform. Zebrafish Appb lacks the KPI domain similar to the brain-enriched human APP695 isoform. Both Appa and Appb have the functional App domains, including the heparin-binding domain (HBD), copper-binding domain (CuBD), extracellular E2 domain (E2), the conserved YENTPY motif, and amyloid-beta region (AB).

(B) Alignment of Aβ regions of zebrafish Appa and Appb to human APP695 and APP770 shows high conservation within the Aβ region and the proteolytic cleavage sites (indicated with black arrows): α-secretase cleavage site (α), β-secretase cleavage site (β), γ-secretase cleavage sites (γ), and ε-cleavage sites (ε). Black, dark gray, and light gray boxes indicate strictly, highly, and moderately conserved amino acid residues, respectively.

(C) As detected by multiplexed hybridization chain reaction (HCR), appa (green) and appb (red) are both expressed widely but in non-overlapping regions of the 5dpf larval brain, including the cerebellum and nuclei in the hindbrain (C, cerebellum; H, hindbrain). Shown is a representative image from a single brain taken at one z-plane (z140/420) (dorsal view, above) and through the midline of the same brain (lateral view, below) from an experiment with a total of n = 22 fish. A, anterior; D, dorsal; L, left.

(D) CRISPR/Cas9 targeting of zebrafish appa resulted in a 5-bp deletion. The target guide RNA (gRNA) sequence is shown in bold, and the obligatory PAM sequence (AGG) is in red. The predicted translation of appaΔ5 leads to a premature stop codon within the Aβ region (Appa amino acid 665). TALEN targeting of zebrafish appb resulted in a 14-bp deletion (dashes) and 4-bp insertion (red). The left and right sites targeted by the TALENs are highlighted, and the spacer sequence, where cleavage occurs, is bolded. The predicted translation of appbΔ14+4 leads to a frameshift and a premature stop codon.

(E) Western blot analysis of APP in brain homogenates from wild-type (WT) and appa^Δ5/Δ5; appb^−/− double mutants.

(F) Elisa detection of Aβ42 (left) and Aβ40 (right) levels in adult brain homogenates from WT controls, WT animals treated with the γ-secretase inhibitor DAPT for 24 h, appa^Δ5/Δ5 and appb^−/− mutants quantified as picograms (pg) of Aβ per μg total protein extracted. Each dot is an independent biological replicate, and error bars represent the mean ± SEM. The bottom panel plots the ratio of Aβ40 to Aβ42 in WT zebrafish adult brain homogenates, n = 5. ∗p ≤ 0.05, ∗∗p ≤ 0.01, Dunnett’s test.

appa^Δ5/Δ5 mutants have reduced day waking activity but no sleep phenotype

Exemplar 48 h traces of average waking activity taken from a single experiment of appa^Δ5/Δ5 mutants, heterozygous, and wild-type siblings (5–7dpf) on a 14 h:10 h light:dark cycle. Each line and shaded ribbon show the mean ± SEM.

(B) Day waking activity and C) night waking activity for appa^Δ5/Δ5 mutants and siblings from appa^+/Δ5 incrosses, combined across N = 5 experiments. Each dot is a single larva, normalized to the mean of their experimentally matched WTs. At the bottom are plotted the effect sizes (±95% confidence interval [CI]) relative to WT.

(D) Exemplar 48 h traces of average sleep for the same experiment shown in (A).

(E) Day sleep and (F) night sleep of WT, heterozygous, and appa^Δ5/Δ5 mutants normalized to their WT siblings, as in (B) and (C). ^nsp > 0.05, ∗p ≤ 0.05, Kruskal-Wallis, Tukey’s post hoc test. n = number of larvae.

appb^−/− mutants have altered waking activity and sleep across the day-night cycle

Exemplar 48 h traces of average waking activity taken from a single experiment of appa^Δ5 mutants, heterozygous, and wild-type siblings (5-7dpf) on a 14h:10h light:dark cycle. Each line and shaded ribbon show the mean ± SEM.

(B) Day waking activity and (C) night waking activity for appb^−/− mutants and siblings from appb^+/− incrosses, combined across N = 5 experiments.

(D) Exemplar 48 h traces of average sleep for the same experiment shown in (A).

(E) Day sleep, (F) night sleep, (G) day sleep length, (H) night sleep length, (I) day sleep bout number, and (J) night sleep bout number of WT, heterozygous, and appb^−/− mutants normalized to WT siblings as in (B) and (C). At top, each dot is an animal normalized to the mean of their experimentally matched WT. At bottom, shown are the effect size ±95%CI relative to WT. ^nsp > 0.05, ∗p ≤ 0.05, ∗∗p ≤ 0.01, Kruskal-Wallis, Tukey’s post hoc test. n = number of larvae.

The γ-secretase inhibitor DAPT shortens sleep bout lengths at night in WT but not in appb^−/− mutants

(A and B) Exemplar 48 h traces on a 14h:10h light:dark cycle of the average waking activity of WT (A) and appb^−/− mutants (B) continuously exposed to either 100 μM DAPT or DMSO vehicle control.

(C and D) Exemplar 48 h traces of the average sleep from the same experiment shown in (A) and (B).

(E) Night sleep and (F) night sleep bout length of WTs and appb^−/− mutants exposed to either 100 μM DAPT or DMSO vehicle. At top, each dot represents a single larva normalized to its experiment-matched WT (-DAPT) mean; error bars indicate ±SEM. At bottom, the within-genotype effect size and 95%CIs of DAPT treatment are plotted. n = the number of larvae. Data are pooled from N = 4 independent experiments, omitting the first day and night to account for any delay in drug action. ^nsp > 0.05, ∗p ≤ 0.05, ∗∗p ≤ 0.01, two-way ANOVA, Tukey’s post hoc test. n = number of larvae.

The β-secretase inhibitor lanabecestat increases sleep at night in WT but not in appb^−/− mutants

(A and B) Exemplar 48 h traces on a 14h:10h light:dark cycle of the average waking activity of WT (A) and appb^−/− mutants (B) continuously exposed to either 0.3 μM lanabecestat or DMSO vehicle control.

(C and D) Exemplar 48 h traces of the average sleep from the same experiment shown in A and B.

(E) Night sleep and (F) night sleep bout length of WTs and appb^−/− mutants exposed to either 0.3 μM lanabecestat or DMSO vehicle. At top, each dot represents a single larva normalized to its experiment-matched WT (-lanabecestat) mean, and error bars indicate ±SEM. At bottom, the within-genotype effect size and 95%CIs of 0.3 μM lanabecestat treatment are plotted. n = the number of larvae. Data are pooled from N = 4 independent experiments, omitting the first day and night to account for any delay in drug action. ^nsp > 0.05, ∗p ≤ 0.05, ∗∗p ≤ 0.01, two-way ANOVA, Tukey’s post hoc test. n = number of larvae.

Intraventricular injection of P3 decreases sleep and shortens sleep bout lengths

Exemplar 24 h traces of the average waking activity of WT larvae injected with either P3 or vehicle control on a 14h:10h light:dark cycle. Shown is a single exemplar experiment.

(B and C) Average waking activity during the day (B) and night (C) across three independent experiments. Each dot represents a single larva normalized to the experiment-matched WT mean. The bars represent the mean.

(D) Exemplar 24 h trace of the average sleep for the experiment shown in (A).

(E) Day sleep, (F) night sleep, (G) day sleep bout length, (H) night sleep bout length, (I) day sleep bout number, and (J) night sleep bout number. n = the number of larvae. ^nsp > 0.05, ∗p ≤ 0.05, ∗∗p ≤ 0.01, Kruskal-Wallis, Tukey’s post hoc test. n = number of larvae.