This is Part 2 of a two-part meeting report. See Part 1.
5 April 2007. The 8th International Conference AD/PD 2007, held 14-18 March in Salzburg, Austria, offered a good opportunity for scientists at large to get up to speed on progranulin, the growth factor whose gene on chromosome 17 is quickly proving to be a major cause of frontotemporal dementia. Since the discovery last summer that mutations in this gene account for many cases of FTLD-U marked by tau-negative but ubiquitin-positive protein inclusions, scientists all over the world have jumped at the chance to understand progranulin better. In Salzburg, the news came in three flavors: There were clinico-pathological observations, genetics, and a budding interest in how a loss of progranulin might cause neurodegeneration. First, the clinical observations.
Several different research groups described their patients with progranulin mutations. The general tenor of these talks and posters was that from a pathological perspective, the patients all had similar inclusions containing ubiquitinated TDP-43. Brain imaging always revealed an asymmetrical pattern, with atrophy and hypoperfusion in certain areas of only one side of the patient’s frontal lobes, indicating that one side of the brain degenerates selectively. And yet, despite these commonalities, these patients came to see their doctors with a surprisingly wide range of complaints that gave rise to an equally wide range of clinical diagnoses. Often there was a spectrum in a given family, and often a person’s diagnosis changed in the course of his or her illness as additional symptoms emerged.
Consider some examples. B.J. Kelley from the Mayo Clinic in Rochester, Minnesota, recounted the story of a family with eight affected members in three generations. One had a dementia diagnosis; two had an Alzheimer or first-MCI-then-AD diagnosis, one had an AD diagnosis that was later changed to FTD. In the third generation, one person is living with amnestic MCI, one received a diagnosis of primary progressive aphasia (PPA) that progressed to PPA/FTD, and one had a diagnosis of FTD. In this family, the affected members all had memory impairment in addition to their other symptoms. In a second family, the diagnoses ranged from AD to behavioral features more typical of FTD, and to behavioral features followed eventually by a Parkinson’s diagnosis. In a third family, the first case first had symptoms of idiopathic Parkinson’s and later was diagnosed as having Parkinson’s with dementia. A sibling had FTD, another had FTD with parkinsonism, and a third had parkinsonism, dementia, and disinhibition and personality changes typical of FTD. All told, Kelley noted that 31 people with a wide range of clinical problems all turned out to have a progranulin mutation, and they also all had the FTLD-U pathology.
Michael Hutton, of the Mayo Clinic in Jacksonville, Florida, said that the most frequent clinical diagnoses people with progranulin mutations receive are for frontotemporal dementia and primary progressive aphasia (PPA); indeed, language difficulties are often what prompt the patient/family to come to the clinic. Corticobasal syndrome also can be an expression of progranulin mutations, a different Canadian/U.S. team reported. Of the patients that Hutton’s group follows, the mean age at death was 63. The mean age of onset was 59, but as with familial AD, the onset range spans some 30 years. Curiously, several scientists noted that later generations in an affected family may show onset at younger ages than their parents, as if they anticipated disease in themselves. This is distinct, however, from triplet repeat diseases, where the abnormal DNA repeats are known to grow in length from one generation to the next, driving the onset age down. The penetrance, that is, a person’s chance of getting sick, also increases with age, Hutton noted. By age 60, about half of all progranulin mutation carriers are ill, and nine of 10 are by age 70.
Ian Mckenzie of the University of British Columbia in Vancouver, Canada, sounded a similar note. Mackenzie emphasized how the characteristic pathology of people with progranulin mutations unites with striking consistency a broad swath of clinical presentations. Last November, Mckenzie and colleagues published the neuropathological features of 13 of their cases (Mackenzie et al., 2006) and in Salzburg he presented an expanded version of that data. In summary, his group, too, found atrophy of the frontal lobes, as well as degeneration of the substantia nigra and medial thalamic nuclei. Mckenzie described abundant, lens-shaped neuronal inclusions in the neocortex, striatum, and hippocampus that stain with antibodies against ubiquitin and TDP-43. The hallmark lesions are inclusions in the nuclei of neurons, but the cytoplasm, some neurites, and some glia have them, as well. The inclusions do not contain progranulin protein and are granular, not filamentous, a finding Virginia Lee’s group, too, saw by electron microscopy. FTLD-U cases without a progranulin mutation had a similar but milder pathology, Mckenzie said.
Samir Kumar-Singh and colleagues from the VIB-University of Antwerp presented pathological data of the Belgian founder family carrying a progranulin-null mutation, described in Cruts et al., 2006. Describing six patients from this extended family, Kumar-Singh emphasized that he observed pathological heterogeneity in the predominant lobar atrophy (parietal vs. frontal lobe atrophy) and also in the subcellular localization of the inclusions (nuclear vs. cytoplasmic) or pathology within neurites. Again, all these ubiquitinated deposits were tied together by the presence of TDP-43 protein. While TDP-43 is always there, it is not the only deposited protein in the ubiquitinated inclusions. Cruts noted that other known proteins were also abundant within them, notably p62.
On the genetics front, the data is branching out from the initial null alleles published last summer by the international teams of Hutton, Howard Feldman at the University of British Columbia, Vancouver, Canada, and Christine Van Broeckhoven at the VIB-University of Antwerp, Belgium. Thirty-nine different mutations in more than 70 families have been found so far, Hutton said, all of which cause a loss of protein function. The scientists traced the most frequent progranulin mutation, R493X, back to a founder effect in England, and were able to follow its distribution to the countries of English emigration, such as the U.S. and Australia. To date, progranulin mutations in Hutton’s samples explain roughly 5 percent of frontotemporal dementia and 13 percent of cases with a family history; this compares with 9 percent and 22 percent, respectively, in a French-Belgian case series led by Alexis Brice at INSERM in Paris and Van Broeckhoven. As always in such studies, occasional cases thought to be sporadic also turn out upon sequencing to have previously discovered familial mutations. A study by Stuart Pickering-Brown from Manchester University estimated the genetic contribution of progranulin to FTD to equal that of tau in a British series of 270 cases, while in a Belgian series, progranulin’s contribution is larger than tau’s. Van Broeckhoven’s team added progranulin genomic deletions to the spectrum of null mutations that may explain an additional 2 percent of FTD. By contrast, researchers led by Lena Skoglund at Uppsala University in Sweden reported that initial analysis of Scandinavian FTD patients turned up little, suggesting progranulin is not a major cause of FTD in Europe’s northern realms.
Other presentations in Salzburg reflected geneticists’ ongoing search for other types of genetic flaws that would reduce a person’s progranulin protein levels beyond outright null alleles, such as missense mutations leading to dysfunctional protein or promoter mutations that depress transcription. All Belgian studies mentioned below came from Van Broeckhoven’s group. One Belgian-French collaboration reported finding three new missense and three promoter variants after direct sequencing of the progranulin gene in 332 FTD patients from these two countries (see van der Zee et al., 2007). Scientists are also exploring the genetic contribution of progranulin to diseases other than FTLD-U. A different sequencing study in a Belgian sample of 779 clinical AD cases turned up new progranulin missense mutations in addition to finding some of the original progranulin-null mutations linked to FTLD-U in a Belgian founder family. Besides further reinforcing the clinical heterogeneity of progranulin mutations, this finding raises the possibility that the gene might end up accounting for a small fraction of AD cases, as well. A separate Belgian study sequencing progranulin in 270 clinical PD cases produced fewer hits, suggesting that progranulin mutations are not a major cause of PD but should be considered in cases where PD patients show simultaneous clinical symptoms of dementia. The role of progranulin mutations in ALS is in flux. Van Broeckhoven noted that sequencing the gene in 230 Belgian ALS patients turned up 11 progranulin mutations, some of which were predicted to affect progranulin protein sequence or levels, plus a common variant that appeared to drive down age of onset and make the disease more aggressive. By contrast, last month a U.S. consortium of scientists concluded that progranulin mutations are not a common cause of ALS in 361 cases of either sporadic or familial ALS, or ALS-FTD (Schymick et al., 2007).
All these different mutations boil down to lowering the amount of progranulin protein in the brain. How does too little of this secreted growth factor cause disease? In addressing this open question, Hutton laid out opportunities for mechanistic and, eventually, therapeutic studies (see also Ahmed et al., 2007). Both neurons and microglia express progranulin, while astrocytes and oligodendrocytes typically do not. Any injurious stimulus that activates microglia will induce a dramatic increase in progranulin expression, Hutton noted. This happens in a host of diseases, and also in mouse models of neurodegeneration. For example, P301L tau transgenic mice massively overexpress progranulin compared with wild-type mice by 12 months of age, and this goes along with activated microglial staining and tau pathology. Aging itself also increases progranulin expression, but less so. In AD brain as well as in transgenic mouse models, progranulin in microglia accumulates in large amounts around amyloid plaques.
Progranulin comprises seven conserved repeats, which, when cleaved by the enzyme elastase, can release smaller peptides called granulins. Progranulin is known to play an important role in tissue repair. It appears as if the whole protein and its smaller granulin offspring have opposing roles in regulating the inflammation that can accompany wound healing. Why is progranulin present in such high levels in activated microglia? Its established role in peripheral wound healing might point to the answer of this question, Hutton noted.
In the periphery, progranulin stimulates the growth of epithelial cells close to the wound, and the site of injury becomes bathed in the growth factor. With wound repair typically comes an increase in inflammatory cytokines, and the conversion from progranulin to individual granulins serves as a molecular switch to control inflammation. If a similar regulatory system operated in the brain, one could imagine that progranulin released from neurons normally is neurotrophic, but in an inflamed or stressed situation, elastase-generated granulins become phagocytotic. This idea is currently under study.
Hutton noted that the clear-cut impact of the various known mutations had his group excited because it implied that restoring progranulin levels might be therapeutic. One way in which that appears doable from an industrial drug development perspective would be to find a small-molecule drug that boosts expression of the remaining normal progranulin gene. After all, the mutation is heterozygous. Initial studies surprised Hutton’s group when a large number of compounds from well-characterized chemical groups turned out to be able to do just that. For example, ibuprofen and other NSAIDs increase progranulin production in cells, Hutton said. In theory, if one can safely induce sustained progranulin elevation in brain, then that could potentially prevent disease in patients who have mutations. Much work remains to be done on this concept. For now scientists are enjoying the early days, when the limited knowledge on progranulin still fits into a simple picture, and the proverbial devil in the detail has not yet reared its ugly head.—Gabrielle Strobel.