4 April 2007. The science of frontotemporal dementias generated perhaps the loudest buzz at the 8th International Conference AD/PD 2007, held 14-18 March in Salzburg, Austria. The excitement grew out of last year’s twin discoveries that mutations in the progranulin gene account for a large chunk of heretofore unexplained cases of this diverse group of diseases (Baker et al., 2006; Cruts et al., 2006; Mukherjee et al., 2006) and that the protein TDP-43 is the main constituent of their characteristic misfolded protein inclusions (see Neumann et al., 2006; Arai et al., 2006). These were major findings that gave impetus to neurogeneticists, neuropathologists, and molecular neurobiologists around the world. Follow-on studies are gearing up, and big discoveries on how those proteins function and how they interact in the chain of pathogenic events are not in yet. But already, six papers on TDP-43 and 19 papers on progranulin were published on the heels of the initial reports, and the Salzburg conference featured nearly 30 presentations on the two proteins. Together, this work is beginning to paint a picture of these formerly obscure forms of neurodegeneration.

As an observer, this reporter was struck that the results coming in from several different groups all seem to jibe rather well, with few contradictory data sets as yet. A trend afoot in this field appears to be that the findings on progranulin and TDP-43 are prompting a realignment of the overlap generated by the different ways to characterize these diseases, i.e., the clinical description, the genetics, and the pathology. Finally, this growing field reminds us yet again that scientists and clinicians have to come to grips with the complex reality of spectrum disorders. Neurodegenerative disease occurs on continuums, not in the neatly defined boxes that an ordering mind tends to prefer. Below is a fairly comprehensive summary of new developments in Salzburg.

In her plenary lecture on TDP-43, Virginia Lee from the University of Pennsylvania School of Medicine, Philadelphia, noted that she expects this protein to define a new type of neuropathology with a similarly broad sweep as tau and α-synuclein have done before it. These better-known proteins first explained pathology, then opened a window on pathogenesis, and then became drug targets. TDP-43 may follow the same path, Lee said. Frontotemporal dementias (FTDs) were recognized in the mid-1980s. They account for close to half of all cases of dementia in people younger than 65. FTDs are distinct from Alzheimer disease, though differential diagnosis frequently is a problem. FTDs can be dramatic for families in that in addition to functional deficits, patients exhibit bizarre personality changes and psychiatric symptoms that are difficult to cope with. FTDs have a strong familial component. Their genetics first broke with the discovery in the early 1990s of mutations in the tau gene (causing FDTP-17), and since then some additional cases were pinned on mutations in the gene for valosin-containing protein, the CHMP2 gene, and an unknown gene on chromosome 9 (see AD/FTD mutation database). But those latter gene changes are rare, and part of the reason why progranulin energized scientists last summer is that it explains a large fraction of the remaining genetic burden of familial FTDs.

TDP-43 did for pathology what progranulin did for genetics. Its discovery suddenly accounted for the largest fraction of previously mysterious protein deposits in all FTDs. These are the ubiquitin-positive but tau-negative intranuclear, cytoplasmic, and neuritic inclusions that mark the brains of people with what was accordingly called FTLD-U. FTLD-U is the most common form of the FTDs. TDP-43 pathology occurs in a spectrum of people who have FTLD-U, FTLD-U with motor neuron disease, and even in sporadic cases of ALS. Sometimes, a given family will have one affected member suffering from frontotemporal dementia and another from ALS, but both people will have TDP-43 deposits in their respective sets of degenerating neurons. Taken together, the cytoplasmic and intranuclear inclusions seen across many different neurodegenerative diseases now fall into four main categories, Lee said. They are the

• tauopathies (which include AD and some FTDs);
• α-synucleinopathies (which include PD, dementia with Lewy bodies or DLB);
• TDP-43opathies (while in need of a better name, this category subsumes FTLD-U and some forms of ALS under one disease entity); and
• poly-Q disorders (which include Huntington’s and ataxias).

The finding that TDP-43 constitutes the main ingredient of ubiquitin-positive inclusions in the brains of people with progranulin mutations gratified researchers. As was the case with amyloid-β and tau before, a convergence of genetics and pathology always gives science a boost. But the connection between progranulin and TDP-43 is not exclusive. TDP-43 is broader. It also deposits in FTLD-U cases with mutations in the gene for valosin-containing protein, in FTLD-U cases without mutations in the known genes (Seelaar et al., 2007; Neumann et al., 2007), as well as in familial ALS without SOD1 mutations (Tan et al., 2007). And, dropping a small fly into the ointment for those who like to keep things nicely separated, Dennis Dickson of the Mayo Clinic in Jacksonville, Florida, reported in Salzburg that TDP-43 inclusions also occur in a subset of cases with AD. As a take-home message from the current pathology data, Lee suggested that TDP-43 defines FTLD-U and ALS as syndromic variants of the same pathology (also Davidson et al., 2007). That is, TDP-43 deposits in cortex and hippocampus underlie symptoms of FTD; in spinal cord and motoneurons they give rise to symptoms of ALS; and a mixed pathology gives rise to a mixed clinical phenotype in the middle of this spectrum. Moreover, both neurons and glia express TDP-43, and indeed the protein reveals a prominent, previously underappreciated white matter pathology in the cortex, spinal cord, and brain stem of people with FTLD-U/ALS (Neumann et al., 2007). TDP-43 constitutes a robust pathologic marker, and smart neurochemists may soon be trying to develop a brain imaging tracer for it.

TDP-43 is a conserved nuclear RNA-binding protein thought to be involved in transcriptional repression and in scaffolding structures called nuclear bodies. In deposits, TDP-43 is hyperphosphorylated, ubiquitinated, and truncated into C-terminal fragments. The disease process by which it becomes adulterated in this way remains unclear, nor is it known which genes it normally represses. TDP-43 gets drawn into a pathogenic process, partly by not being degraded properly anymore. However, it does not at present appear to be a major genetic cause of neurodegeneration. In Salzburg, neither a Belgian team led by Marc Cruts and Christine van Broeckhoven at the VIB-University of Antwerp, nor a U.S.-Canadian team led by Roos Rademakers and Mike Hutton at the Mayo Clinic in Jacksonville, Florida, reported finding disease-causing TDP-43 mutations in FTLD-U/FTD/FTD-ALS cases analyzed so far. Sequencing and association studies are ongoing, so a small genetic role for TDP-43 may still turn up. Progranulin, by contrast, is fast becoming a heavy-hitter in FTD genetics. The relationship between TDP-43 and progranulin is a mystery investigators would love to solve. For more on progranulin, see Part 2 of this meeting report.—Gabrielle Strobel.