6 July 2009. Since it was linked to amyotrophic lateral sclerosis in 2006, TAR DNA Binding Protein-43 has become a top player in research on ALS and frontotemporal dementia. People with both conditions often accumulate TDP-43 aggregates in neurons. These two TDP-43 proteinopathies, which some have suggested are simply the two ends of a continuous spectrum, were discussed in the Third International Research Workshop on Frontotemporal Dementia in ALS, held 21-25 June in London, Ontario. In order to figure out how TDP-43 causes disease, scientists need animal models, and there was no shortage of these in the presentations. “There’s probably never been a more exciting time to work on ALS,” said Don Cleveland of the University of California, San Diego.
Cleveland, as well as many other scientists, is working to develop a mouse model for TDP-43 disease. TDP-43 is an essential protein, so simple deletions are out. Cleveland presented his lab’s progress, in collaboration with Christopher Shaw of King’s College London, on a set of transgenic mice that carry human TDP-43 along with the endogenous mouse gene. The animals express the transgene, in varying amounts, but there have been a few snags in establishing lines. Several founders failed to produce progeny. In animals expressing TDP-43 containing mutations found in human disease, the researchers have been able to make fertile animals with symptoms of neurodegenerative disease, but the phenotype is weaker in the F1 and F2 generations.
While the field anxiously awaits a useful mouse model, plenty of other animals are ready for experimentation now. Ronald Klein of the Louisiana State University Health Sciences Center in Shreveport presented his rat model, in which he uses a adeno-associated virus to transfer the human TDP-43 gene into the substantia nigra (see Tatom et al., 2009 and ARF related news story). Klein hopes to improve his model by adding a “trigger” factor that will cause the human TDP-43 to more readily leave the nucleus for the cytoplasm, as it does in disease.
Cleveland has begun using a somewhat similar approach in rodents, except his goal is to silence TDP-43 expression, to see what consequences this has for RNA regulation and physiology. By pumping antisense oligonucleotides into the cerebral spinal fluid, scientists in the Cleveland lab were able to decrease TDP-43 expression. With TDP-43 RNA interference, the initial experiments proved the protein was essential: as the levels dropped, the animals eventually died. “This was actually a disaster,” Cleveland said. “We didn’t want them to be dead.”
Another option used by Shaw is chick embryo (see Sreedharan et al., 2008 and ARF related news story). The model relies on electroporation to transfect tagged transgenes into the spinal cord. Conveniently, the genes can be targeted to only one side of the spinal cord, allowing the other side to serve as an internal control. Shaw reported that his group has overexpressed mutant TDP-43 in more than 300 embryos, with a consistent result—cell death. Using GFP-tagged versions of TDP-43, Shaw’s lab found that the wild-type protein remained in the nucleus where it belonged, while mutants were found mainly in the cytoplasm. Some mutant TDP-43 remained in the nucleus, where it formed inclusions.
Zebrafish can also provide useful information, said Philip Van Damme of VIB in Leuven, Belgium. He uses zebrafish embryos “mainly because it’s so easy,” he said—the animals are small and cheap, and convenient tools are available. Using oligonucleotides called morpholinos, scientists can block translation of mRNA. Alternatively, they can inject RNA to ramp up gene expression. There are two homologs of TDP-43 in zebrafish: TARDBP shares 74 percent homology, and TARDBPL 57 percent, with the human gene. Overexpression of wild-type human TDP-43 resulted in shorter motor neuron axons with aberrant branching, and mutant versions of TDP-43 exacerbated the phenotype. These problems matched those of animals overexpressing SOD1, another gene linked to ALS (see Lemmens et al., 2007). “We were really pleased to see we could obtain a phenotype similar to the one we saw before with SOD1,” Van Damme said. He suggested the zebrafish would make a useful “pre-model”—allowing scientists to test their hypotheses quickly before moving into time-consuming mouse studies.
There is plenty to be learned from invertebrates as well. David Morton of the Oregon Health & Science University in Portland and Emanuele Buratti of the International Centre for Genetic Engineering and Biotechnology (ICGEB) in Trieste, Italy, both presented results from studies in Drosophila. “What we’re trying to do is basically use Drosophila to understand the normal functions of TDP-43,” Morton said. “Ultimately, we would really like to understand what is going on at a cellular level.” Fruit flies also boast a host of useful genetic and genomic tools. By mixing and matching ready-made promoter-gene fusions, insect biologists can express their target gene across a variety of tissues, or knock it out via RNA interference if they prefer.
The Drosophila version of TDP-43 is TBPH. Among the 46 percent of residues conserved between human and fly are 11 of the 28 residues shown to be mutated in some forms of human disease. Silencing TBPH across all tissue types with RNA interference killed flies before adulthood, but knocking down TBPH expression selectively in neurons, by a factor of two or three, resulted in viable animals, Morton found. The flies showed neurodegeneration in the retina and lesions in the neuropil that increased with age. Although Morton’s TBPH deletions, like the RNAi knockdown models, died before reaching adulthood, the scientists were able to study them as larvae. “We saw a dramatic reduction of the total distance crawled by these larvae,” Morton reported, although just by watching the animals their process of movement appears to be normal.
In contrast, Buratti had data on TBPH knockouts that did survive to adulthood, from the laboratory of Fabian Feiguin at ICGEB. “It’s something that [we] have to sort out,” Morton said. He suspects that there may only be minor differences between the lines, though, because his knockouts developed into fully formed adults that apparently lacked the strength to climb out of the pupal cuticle after metamorphosis. Feiguin’s flies were also weak, often needing assistance in exiting the pupal cuticle, and had reduced lifespans. In Feiguin’s experiments, adults that had reduced expression of TBPH in neurons due to RNA interference also had motility defects. Co-expression of TBPH, or human TDP-43, somewhat rescued these defects.
Worms, too, have something to offer in the TDP-43 arena. Human TDP-43 causes movement defects when expressed in the nematode C. elegans, reported Brian Kraemer of the University of Washington in Seattle. They wriggled only half as far across a plate as non-transgenic animals. When the TDP-43 transgene contained disease-associated mutations, the phenotype intensified, with some animals barely traversing a millimeter. “They don’t get around very well at all,” Kraemer said. Worms also have their own TDP-43, but it lacks the carboxyl-terminal region where the disease-causing mutations cluster. The transgenic worms show several features of human disease, Kraemer said, including progressive motor neuron dysfunction and degeneration, decreased lifespan, and accumulation of aggregated, insoluble TDP-43. Worm models can be used to understand the basic biology of TDP-43 pathology in a simple system with only 302 neurons, he said, as well as to search for genes that interact with the TDP-43 pathway.
Each of these systems offers a platform for asking the most basic questions about TDP-43: What happens when its function is lost in the nucleus? How does it cause pathology? How do mutations contribute to these events? And ultimately, how can medicine halt these molecular disasters? Having fashioned the tools, scientists can now apply them to screen therapies as well as investigate basic biology.—Amber Dance.