LARP1 haploinsufficiency is associated with an autosomal dominant neurodevelopmental disorder

Abstract

Autism spectrum disorder (ASD) is a neurodevelopmental disorder (NDD) that affects approximately 4% of males and 1% of females in the United States. While causes of ASD are multi-factorial, single rare genetic variants contribute to around 20% of cases. Here, we report a case series of seven unrelated probands (6 males, 1 female) with ASD or another variable NDD phenotype attributed to de novo heterozygous loss of function or missense variants in the gene LARP1 (La ribonucleoprotein 1). LARP1 encodes an RNA-binding protein that post-transcriptionally regulates the stability and translation of thousands of mRNAs, including those regulating cellular metabolism and metabolic plasticity. Using lymphocytes collected and immortalized from an index proband who carries a truncating variant in one allele of LARP1, we demonstrated that lower cellular levels of LARP1 protein cause reduced rates of aerobic respiration and glycolysis. As expression of LARP1 increases during neurodevelopment, with higher levels in neurons and astrocytes, we propose that LARP1 haploinsufficiency contributes to ASD or related NDDs through attenuated metabolic activity in the developing fetal brain.

Keywords: ASD; LARP1; NDD; RBP; RNA binding protein; autism; metabolism; neurodevelopmental; plasticity; proband.

Graphical abstract

Figure 1

LARP1 isoforms and pathogenic variants in probands (A) LARP1 is a positive strand gene on chromosome 5 with two major isoforms: long/full length (blue), in which the first three exons encode amino acids, and short (red), in which the first three exons are non-coding. Mean expression across 176 postmortem human prefrontal cortex samples is shown at the top (BrainVar²²), and 100-way conservation across species (PhyloP) is shown at the bottom. (B) Zoomed-in view of the differing transcription start sites from (A). (C) Expression of the first exon from the long and short isoforms of LARP1 are shown for each of the 176 samples split into three developmental stages (PCW, post-conceptual weeks). (D) Schematic representation of LARP1 protein and the variation by proband (numbered above) described in this study and de novo genetic variants in developmental delay and ASD cohorts.^, LaM, La motif; RRM, RNA recognition motif; PAM, PABP-interacting motif; DM15, LARP1-specific HEAT-like tandem repeat region.

Figure 2

Immortalized lymphoblasts derived from proband 1 express reduced levels of LARP1 (A) Immortalized lymphoblasts derived from proband 1 express ∼50% lower levels of full-length LARP1 protein compared with his parents. Immunoblotting was conducted using an antibody raised against human LARP1 amino acids 250–350 (top) and a C-terminal antibody against amino acids 848–1096 (middle). β-Actin was used as a loading control (bottom). A variant protein corresponding to the predicted molecular weight (80.9 kDa) of the C-terminal truncated LARP1 is detected only in proband 1. (B) The western blot in (A) was quantified by densitometry relative to expression levels in proband 1’s mother and confirmed a ∼50% decrease in expression levels of LARP1 protein. (C) LARP1 mRNA levels were quantified by RT-qPCR using primer pairs corresponding to the N and C termini of LARP1. All data were normalized to expression levels in immortalized lymphoblasts from proband 1’s mother (LARP1 wild type). ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001 (n = 3).

Figure 3

Full-length, but not truncated, LARP1 promotes cellular metabolism (A and B) Oxygen consumption rate (A) and extracellular acidification rate (B) of proband 1’s lymphoblasts compared to wild-type lymphocytes from his parents. ∗p < 0.05 and ∗∗p < 0.01 (n = 3). (C) Wild-type or LARP1 knockout (LARP1KO) HEK293T cells were transfected with expression constructs expressing either wild-type LARP1 or the truncated variant (Var) of LARP1 found in proband 1. Successful expression of both constructs was confirmed by western blotting. (D) Oxygen consumption rate measured by Seahorse assay in wild-type HEK293T and LARP1KO cells transfected with wild-type and truncated LARP1 expression constructs. The low oxygen consumption rate of HEK293T LARP1KO cells was partially rescued by the wild type LARP1 expression construct but not the truncated LARP1 expression construct.

Figure 4

LARP1 expression across development and by cell type in the human cortex (A) LARP1 expression (log₂ counts per million +1) from bulk tissue RNA-seq of the postmortem human prefrontal cortex is shown for 176 samples across brain development (red points) from BrainVar. The line represents the LOESS (locally estimated scatterplot smoothing), with the 95% confidence interval shown by the shaded area. (B) LARP1 expression is shown from single-cell RNA-seq data of 8 sample of the adult human middle temporal gyrus. CT, corticothalamic; ET, extratelencephalic projecting; IT, intratelencephalic projecting; L4, L5, and L6, layers 4, 5, and 6 of the cortex; OPC, oligodendrocyte precursor cells; VLMC, vascular and leptomeningeal cells.