Tdp-43 cryptic exons are highly variable between cell types

Abstract

Background: TDP-43 proteinopathy is a prominent pathological feature that occurs in a number of human diseases including amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), and inclusion body myositis (IBM). Our recent finding that TDP-43 represses nonconserved cryptic exons led us to ask whether cell type-specific cryptic exons could exist to impact unique molecular pathways in brain or muscle.

Methods: In the present work, we investigated TDP-43's function in various mouse tissues to model disease pathogenesis. We generated mice to conditionally delete TDP-43 in excitatory neurons or skeletal myocytes and identified the cell type-specific cryptic exons associated with TDP-43 loss of function.

Results: Comparative analysis of nonconserved cryptic exons in various mouse cell types revealed that only some cryptic exons were common amongst stem cells, neurons, and myocytes; the majority of these nonconserved cryptic exons were cell type-specific.

Conclusions: Our results suggest that in human disease, TDP-43 loss of function may impair cell type-specific pathways.

Keywords: Amyotrophic lateral sclerosis; Bioinformatics; Frontotemporal dementia; Inclusion body myositis; TDP-43 –Nonconserved cryptic exons.

Fig. 1

Generation of CaMKIIα-Cre;Tardbp ^F/F and MLC-Cre;Tardbp ^F/F knockout mice. (a) Breeding strategy to cross floxed Tardbp knockout mice with CaMKIIα-Cre or MLC-Cre mouse lines to conditionally delete Tdp-43 in excitatory neuron or skeletal muscle, respectively. Hippocampal protein extracts from CaMKIIα-Cre;Tardbp ^F/F knockout mice were taken from p25 and 3-month old mice, as indicated. Protein extracts from various muscle groups, as indicated, were taken from 2-month old MLC-Cre;Tardbp ^F/F mice. Immunoblotting confirms deletion of Tdp-43 in the hippocampi of CaMKIIα-Cre;Tardbp ^F/F knockout mice (b) and the quadriceps of MLC-Cre;Tardbp ^F/F knockout mice (c); biological replicates of immunoblotting were performed in excess of n = 3 to validate knockdown. (d) Immunofluorescence staining of hippocampal sections from 3 month old CaMKIIα-Cre;Tardbp ^F/F knockout mice demonstrate specific deletion of Tdp-43 from neurons (CA region, scale bar = 50 μm). (e) Immunohistochemical staining of Tdp-43 in quadriceps from 3 month old MLC-Cre;Tardbp ^F/F knockout mice also reveals loss of Tdp-43, as indicated by asterisks (scale bar = 50 μm)

Fig. 2

Neuron-specific cryptic exons (CaMKIIα-Cre;Tardbp ^F/F knockout mice). (a) Visual examples of neuron-specific cryptic exons (Ap3b2, Camk1g). (b) Neuron-specific cryptic exons are flanked by UG repeats that are present upstream, downstream or within the cryptic exon sequence itself. (c to e) RT-PCR validation of cryptic exons (red arrows) in RNA extracted from hippocampi of 3 month old CaMKIIα-Cre;Tardbp ^F/F mice. Refer to Additional file 2 for a complete list of cryptic exons

Fig. 3

Muscle-specific cryptic exons (MLC-Cre;Tardbp ^F/F knockout mice). (a) Visual examples of muscle-specific cryptic exons (Sh3bgr, Tns1). (b) Muscle-specific cryptic exons are flanked by UG repeats that are present upstream, downstream or within the cryptic exon sequence itself. (c and d) RT-PCR validation of cryptic exons (red arrows) in RNA extracted from quadriceps of 2 month old MLC-Cre;Tardbp ^F/F mice. Refer to Additional file 2 for a complete list of cryptic exons

Fig. 4

Tdp-43 cryptic exons are highly variable between cell types. (a) While some cryptic exons are common between cell types, many cryptic exons are unique to neurons (58), muscle (79) and stem cell [22]. Of the common cryptic exons, several are highly incorporated in mRNA regardless of splicing environment (b), while other cryptic exons are incorporated at varying levels depending on the cell type (c to g)