Rare variants in BMAL1 are associated with a neurodevelopmental syndrome

Abstract

Through international gene-matching efforts, we identified 10 individuals with ultrarare heterozygous variants, including 5 de novo variants, in BMAL1, a core component of the molecular clock. Instead of an isolated circadian phenotype seen with disease-causing variants in other molecular clock genes, all individuals carrying BMAL1 variants surprisingly share a clinical syndrome manifest as developmental delay and autism spectrum disorder, with variably penetrant sleep disturbances, seizures, and marfanoid habitus. Variants were functionally tested in cultured cells using a Per2-promoter driven luciferase reporter and revealed both loss-of-function and gain-of-function changes in circadian rhythms. The tested BMAL1 variants disrupted PER2 mRNA cycling, but did not cause significant shifts in cellular localization or binding with CLOCK. Conserved variants were further tested in Drosophila, which confirmed variant-dependent effects on behavioral rhythms. Remarkably, flies expressing variant cycle, the ortholog of BMAL1, also demonstrated deficits in short- and long-term memory, reminiscent of the highly prevalent developmental delay observed in our cohort. We suggest that ultrarare variants in the BMAL1 core clock gene contribute to a neurodevelopmental disorder.

Keywords: BMAL1; Drosophila; circadian rhythms; developmental delay; neurodevelopmental disorder.

Fig. 1.

Characteristics of BMAL1 variants presented in this study. (A) The molecular clock is composed of a transcription–translation feedback loop. BMAL1 and CLOCK promote the transcription of the PER and CRY families of genes, and then are inhibited by those protein products over a period of 24 h. (B) Schematic of human BMAL1 protein with major protein domains depicted in blue. Missense variants are listed on top and frameshift or predicted truncations are depicted on bottom. (C) Amino acids implicated in this study demonstrate marked conservation across animal phyla. (D) SpliceAI was used to predict changes in splice acceptors or donors for the three rare splice site variants presented in this study, as well as three nearby common splice site variants. Rare variants are likely to lead to a loss of splice acceptor and/or donors. 0.8 indicates a stringent cutoff, 0.5 indicates a moderate cutoff, and 0.2 indicates a low cutoff value.

Fig. 2.

U2OS cells expressing Per2-dLuc reporter and BMAL1 variants reveal altered BMAL1 function. (A) Raw luminescence signal after dexamethasone synchronization was recorded for 6 d. The same genetic control condition BMAL1^+/+ is plotted in each trace to aid in comparison to variants. Traces indicate average values and thickness of the line depicts SE of the mean. n = 5 to 9 experiments/condition. (B) Circadian parameters calculated through BioDare2. One-way ANOVA with Benjamini, Krieger, and Yekutieli’s two-stage step-up method to control the false discovery rate for multiple comparisons made to the control genotype. *P < 0.05 **P < 0.01 ***P < 0.001. Data are presented as mean values ± SEM.

Fig. 3.

Altered PER2 expression caused by BMAL1 variants. qPCR results PER2 expression at six timepoints through the day in wild-type and variant BMAL1 U2OS cells. Daily oscillation of PER2 mRNA in control cell is in gray and replotted in each trace to allow for comparison to variant lines. Circadian parameters were calculated through BioDare2 and significant results are listed in respective graphs. n = 3 samples/timepoint/condition. One-way ANOVA with Benjamini, Krieger, and Yekutieli’s two-stage step-up method to control the false discovery rate for multiple comparisons made to the control genotype. To compare mRNA levels independent of circadian metrics, a mixed-effects model with Geisser–Greenhouse correction and Dunnett’s multiple comparisons test were used. *P < 0.05 **P < 0.01 ***P < 0.001. Data are presented as mean values ± SEM.

Fig. 4.

BMAL1 expression and interaction with CLOCK. (A) Representative immunoblots depicting expression of BMAL1 and CLOCK, as well as β-actin loading control, in wild-type, and variant BMAL1 cell lines (Left). Immunoprecipitation of CLOCK and probing for BMAL1 demonstrates preserved protein–protein interactions (Right). (B) Quantification of BMAL1 band intensity in input as normalized to β-actin loading control reveals decreased BMAL1 expression caused by BMAL1 c.141-2A>G, p.(Tyr340CysfsTer35), and p.(Gly535GlufsTer) variants. n = 3 replicates/condition. One-way ANOVA with Holm–Sidak’s multiple comparisons test. ***P < 0.001. Data are presented as mean values ± SEM.

Fig. 5.

Flies harboring orthologous cyc variants exhibit circadian alterations. (A) cycle is the fly ortholog of BMAL1 with a DRSC Integrative Ortholog Prediction Tool (DIOPT) score of 13, indicating high homology. The BMAL1 p.I201T variant in the PAS1 domain corresponds with the cycle p.I161T variant in the PAS1 domain. The BMAL1 p.Y340CfsTer35 variant in the PAS2 domain corresponds with the cycle p.F311fs variant in the PAS2 domain. (B) Representative actograms demonstrating that loss-of-function cyc⁰¹/cyc⁰¹ flies lack rhythmic behavior in total darkness, but when wild-type cyc (UAS-cyc^WT) or cyc p.I161T (UAS-cyc^I161T) is expressed, behavioral rhythmicity is improved in light–dark cycles. Yellow shading indicates presence of light and resumption of ZT9-9 light:dark cycle. (C) Quantification of behavioral rhythmicity. Left: percentage of flies demonstrating rhythmic behavior as measured by an FFT value above 0.01. Middle and Right: FFT values for flies with rhythmic behavior in total darkness (DD) and light–dark cycles (LD). Note that rhythmic flies were rarely seen in the cyc⁰¹/cyc⁰¹;tim-Gal4 and cyc⁰¹/cyc⁰¹;tim-Gal4> UAS-cyc^F311fs conditions in DD, so FFT values are not provided for the Middle panel. n = 82-96 flies/condition. The Mann–Whitney test was used when comparing two groups given lack of normal distribution. The Kruskal–Wallis test was used when comparing three or more groups with Dunn’s multiple comparisons test given lack of normal distribution. *P < 0.05. Data are presented as mean values ± SEM.

Fig. 6.

Short- and long-term memory is impaired in flies expressing variant cycle. (A) Quantification of short-term memory in cyc⁰¹/cyc⁰¹ flies expressing wild-type cyc (UAS-cyc^WT), cyc p.I161T (UAS-cyc^I161T), or cyc p.F311fs (UAS-cyc^F311fs) using the tim-Gal4 driver. cyc p.I161T and cyc p.F311fs disrupts short-term memory. (B) Quantification of long-term memory in cyc⁰¹/cyc⁰¹ flies expressing wild-type cyc, cyc p.I161T, or cyc p.F311fs using the tim-Gal4 driver. cyc p.I161T and cyc p.F311fs impairs long-term memory. n = 7 to 8 replicates/condition. One-way ANOVA with Tukey’s multiple comparisons test. Data are presented as box and whiskers and showing all points.