Project
Drosophila models for TDP-43 and FUS proteinopathy, new therapeutic targets for neurodegeneration
TDP-43 is well known as a nuclear RNA/DNA binding protein involved in many aspects of
RNA metabolism. Accumulation of TDP-43 have been identified in the large majority of
patients within the amyotrophic lateral sclerosis-frontotemporal dementia spectrum of
disorders. TDP-43 proteinopathy is also found in more than 30% of Alzheimers disease
brains, suggesting a broad role for TDP-43 in neurodegeneration in the general population.
TDP-43 positive inclusions are the major pathological hallmark of the disease and are
typically located in the neuronal cytoplasm and accompanied by a loss of nuclear TDP-43
expression. ALS (and sometimes FTD) can be caused by heterozygous mutations in hTDP-
43
, although in the majority of patients that suffer from TDP-43 proteinopathy no hTDP-43mutations are present. Although this proteinopathy has been discovered ~8 years ago, its
pathogenetic mechanism is still not clear and a major unresolved question is whether TDP-43-
mediated neurotoxicity is caused by a gain or loss-of-function mechanism. TDP-43 is
evolutionary well-conserved and in Drosophila, there is a homolog called TBPH or dTDP-43
which, similar to its human counterpart, has two RNA recognition motifs, a glycine rich
region and displays similar nucleic acid binding and mRNA splicing properties.
In the first part of this PhD thesis, we further clarified the normal function of dTDP-43 in
Drosophila
. We therefore combined a phenotypical analysis with next generationtranscriptome analysis (RNA-seq) of dTDP-43 gain and loss of function flies. We found in
Drosophila
that dTDP-43 coordinates the switching of ecdysteroid receptor (EcR)-dependenttranscriptional programs from a pupal to an adult pattern and that a disturbance of this
function by gain or loss of dTDP-43 results in lethality and enhanced neuronal apoptosis.
dTDP-43 controls this switching by regulating the expression of a microtubule-binding
protein that is reponsible for the correct subcellular localization of EcR. Our results establish
disrupted EcR signaling as a cellular mechanism underlying dTDP-43 neurotoxicity in
Drosophila
and identify steroid hormone receptor signaling as a potentially importantpathway in ALS and related TDP-43 proteinopathies. Since increasing and decreasing
expression of dTDP-43 lead to largely overlapping transcriptomic alterations, cytoplasmic
EcR accumulations and neurotoxic developmental phenotypes, our study suggests that TDP-
43 aggregation results in its loss-of-function.
Secondly, we wanted to clarify the pathogenic character of mutations in TDP-43. Most
heterozygous hTDP-43 mutations located in the C-terminal glycine-rich region of the protein
cause ALS, but other atypical variants such as hTDP-43A90V (located in the nuclear
localization signal) and hTDP-43D169G (located in the first RNA-binding domain) are
described as well. In order to unravel the pathogenic nature of the different variants as well as
their gain- and/or loss-of-function properties, we have used site-specific transgene integration
to express hTDP-43WT, two typical ALS-causing mutations (hTDP-43G287S and hTDP-43A315T)
and two atypical variants (hTDP-43A90V and hTDP-43D169G) in dTDP-43 loss-of-function flies
and checked their ability to rescue the loss-of-function phenotypes. We discovered that
hTDP-43A90V
, hTDP-43G287S and hTDP-43A315T failed to rescue the neuronal loss, while hTDP-43WT
and hTDP-43D169G could rescue, suggesting that the hTDP-43A90V, hTDP-43G287S andhTDP-43A315T
mutations have loss-of-function properties. We also reported a shift of hTDP-43 from the nucleus to the cytoplasm in ~10% of the bursicon neurons in hTDP-43A90V,
hTDP-43G287S
and hTDP-43A315T but not hTDP-43D169G and hTDP-43WT flies. Next, weprovide additional evidence that the rare variant hTDP-43A90V is indeed pathogenic and might
increase the risk to develop Alzheimers disease (AD) in the French-Belgian population.
Together these in vivo data suggest that typical ALS-causing mutations (G287S, A315T) and
the rare variant A90V might be pathogenic through a loss-of-function mechanism and that
pathogenic mechanism in TDP-43 associated neurodegeneration is rather caused by a loss of
the normal molecular function of TDP-43 than a novel toxic gain of function.
In conclusion, we show that D. melanogaster is an excellent model to study TDP-43
proteinopathies and we provide substantial evidence that TDP-43-associated
neurodegeneration is rather caused by a loss of the normal molecular function of TDP-43 than
a novel toxic gain of function.