5 June 2009. Scientists wrestling the complexities of α-synuclein fluid biochemistry (see Part 6 of this series) might be forgiven for looking with some envy to a different protein of the neurodegenerative disease spectrum. At the 9th International Conference AD/PD, held last March in Prague, the field learned that crafting such a test for progranulin might actually be comparatively easy—incredible as that sounds in the field of neurodegeneration where generally speaking nothing is easy. Progranulin is the protein behind a sizable fraction of frontotemporal dementia and probably also a small, still-unknown fraction of cases diagnosed clinically as early-onset Alzheimer’s or related disorders. In Prague, three independent groups presented results of their fledgling ELISA tests—one in serum, one in plasma, and one in cerebrospinal fluid. Incredible as it may seem to a biomarker field plagued by inconsistent findings on blood tests for other proteins such as Aβ, all three groups reported the same overall result: it works just fine, thank you.

Progranulin surfaced independently in the laboratories of Christine van Broeckhoven at the VIB-University of Antwerp, Belgium, and of Michael Hutton, then at the Mayo Clinic Jacksonville, Florida, as the gene for the tau-negative form of frontotemporal dementia 17 (FTLD-U). This highly familial disease frequently strikes people younger than 65 (see ARF related news story). Since then, 66 pathogenic mutations in progranulin have turned up (see mutation database). Importantly, progranulin causes neurodegeneration by a different mechanism than Aβ, tau, or α-synuclein (see Part 1 of this series). Whereas these latter proteins are thought to become toxic as their concentration rises (i.e., the more protein, the earlier one gets sick), progranulin leads to neurodegeneration when there is not enough of it. With Aβ, tau, and α-synuclein, mutations that drastically increase expression cause familial early-onset disease, whereas risk alleles influence sporadic disease. In contrast, the general theme emerging from progranulin genetics is that null mutations that slash protein levels in half cause familial FTLD-U, while milder missense mutations that cause a partial loss of function have a susceptibility role in Alzheimer’s, amyotrophic lateral sclerosis, and perhaps Parkinson disease, van Broeckhoven said in her talk in Prague.

Progranulin’s different mechanism should translate into differences in diagnosis and treatment. The gene itself is plenty complicated, and the six different granulin proteins resulting from it have physiological functions throughout the body. But diagnosis might be straightforward. “We thought progranulin protein levels should be decreased in the blood of people with mutations that cause loss of function,” Kristel Sleegers in van Broeckhoven’s group said in her talk. Sleegers started with an ELISA against full-length progranulin developed originally by Philip van Damme (van Damme et al., 2008). She put it to work on blood samples from a large Belgian founder family whose 43 patients showcase the dramatic clinical heterogeneity of progranulin mutations. Their clinical diagnoses range from FTD, AD, PD, primary progressive aphasia (PPA), and progressive non-fluent aphasia (PNFA)—all from having inherited the same mutation. Pathologically, this family runs the gamut, too, with Lewy bodies, ubiquitin-positive FTLD-U inclusions, amyloid pathology, and of course TDP-43 (Brouwers et al., 2007).

From this family, Sleegers had serum of six patients, eight younger still-unaffected mutation carriers, and nine non-carriers. The ELISA distinguished carriers and non-carriers unequivocally, Sleegers showed. The groups were completely separate and apart by a large distance. Interestingly, the non-symptomatic carriers had the same progranulin levels as their affected relatives, suggesting the ELISA may be able to detect preclinical disease and presymptomatic mutation carriers. Genetic testing can do this, too, but it is more complicated to interpret, as scientists need to know whether a change in the gene sequence is pathogenic or a harmless variation, and genetic deletions require further analysis. Besides capturing all types of progranulin mutation, a blood-based ELISA could also be cheaper than genetic testing.

In her talk, Rosa Rademakers of the Mayo Clinic Jacksonville, Florida, reported the same results in a different, larger group of patients. Rademakers is a neurogeneticist also formerly of van Broeckhoven’s group; she received a Young Investigator Award at the conference (see ARF related news story). Her team optimized a commercial ELISA for human progranulin and tested plasma of 219 patients clinically diagnosed with FTD. In this study, too, all patients carrying a loss-of-function progranulin mutation had only about one-third as much progranulin in their blood as did patients without a progranulin mutation. The ELISA predicted with 100 percent certainty that everyone with less than 112 nanogram/milliliter (ng/ml) of the growth factor carried a progranulin mutation. This is nearly identical to the cutoff suggested in an earlier Italian study led by Giuliano Binetti at the Centro San Giovanni di Dio-Fatebenefratelli in Brescia, which tested plasma and CSF ELISAs in a group of FTLD patients (Ghidoni et al., 2008).

Working in parallel, both groups next studied whether their ELISAs could tease apart some of the multiple disease processes underlying a clinically defined disease, in other words, serve as a new tool to better define the spectrum of neurodegeneration. For example, Sleegers reported that progranulin was low in the serum of a person who had been diagnosed with AD but later proved to have a loss-of-function progranulin mutation. Conversely, Rademakers showed that the plasma test revealed abnormally low progranulin in one of 72 people clinically diagnosed with AD and that this man, upon sequencing, proved to have a new loss-of-function progranulin mutation. Likewise, a French man with clinical Parkinson disease and a progranulin mutation also had low plasma progranulin. “Regardless of how a person presents clinically, the ELISA detects a progranulin null mutation,” Sleegers said.

Lastly, both studies explored whether their ELISAs were able to pick up more subtle genetic flaws in progranulin. For example, missense mutations cause less than haploinsufficiency, which results from a mutation that aborts protein production entirely. Researchers are exploring different kinds of missense mutations in this gene. Some hasten the degradation of the protein, others reduce its secretion, and in-silico modeling points to misfolding at the protein’s internal disulfide bridges as a possible cause for these cellular problems with the protein (Shankaran et al, 2007; van der Zee et al., 2007). In Prague, Sleegers closed her talk with data showing that both in people with clinical FDT and AD, missense mutations that such research had predicted to be pathogenic came with reduced serum progranulin levels in their carriers, though the drop was less precipitous than with a null mutation. Missense mutations predicted to be harmless corresponded to normal levels of serum progranulin. Both Sleegers’ and Rademaker’s studies appeared recently online (Sleegers et al., 2009; Finch et al., 2009).

Last but not least, also in Prague, German researchers led by Anja Capell and Christian Haass at Ludwig-Maximilian University in Munich presented ongoing work on a third ELISA to quantify progranulin in the CSF. Compared to serum and plasma, where progranulin levels ranged in the low hundreds of ng/ml for controls and 50 to 90 ng/ml in null mutation carriers, CSF levels are much lower, around 5-7 ng/ml in controls. Previous work has reported this same range in controls and about 2 ng/ml in mutation carriers (van Damme et al., 2008). In clinical practice, it is not clear if a spinal tap will be necessary eventually, because blood-based tests appear to work well so far, Rademakers wrote by e-mail.

All told, these studies suggest that blood tests could reveal an underlying progranulin-driven disease process regardless of how it manifests clinically. It is simpler than genetics because it picks up the loss of the protein no matter which of a myriad of different genetic changes might be to blame. Such a blood test could show whose early-onset dementia is due to this particular protein, and predict future neurodegeneration in people who are still cognitively healthy but at risk because of their family history. Viewed broadly beyond FTD, progranulin tests could help explain a slice of the neurodegenerative disease spectrum.—Gabrielle Strobel.

This is Part 7 of a nine-part series. See also Part 1, Part 2, Part 3, Part 4, Part 5, Part 6, Part 8, Part 9.