Pin1 binds and isomerizes tau that is phosphorylated at threonine 231 (pT231), one of the earliest phospho-epitopes detected before neurofibrillary tangles form in AD (Luna-Muñoz et al., 2007). Pin1 levels drop in the brains of people with AD (Lu et al., 1999) or frontotemporal dementia (Thorpe et al., 2004). And in mice, lack of Pin1 induces tau and Aβ pathology in otherwise normal animals (ARF related news story on AbstractLiou et al., 2003; ARF related news story on Pastorino et al., 2006), whereas Pin1 overexpression prevents tauopathy in tau transgenic mice (Lim et al., 2008). In human genetic studies, Pin1 polymorphisms that reduce the protein’s expression correlate with higher AD risk (Segat et al., 2007), while gene variants that drive up Pin1 levels delay disease onset (Ma et al., 2010; see Pin1 on AlzGene). “All this suggests that Pin1 protects against AD,” said Lu in an interview with Alzforum. However, without a means to distinguish cis- or trans-specific forms of phosphorylated tau, researchers could not directly show that isomerization influences what tau does, or even what occurs in vivo.

Generating isomer-specific antibodies to cis and trans pT231-tau was the first order of business in the current study. The task was tricky; with only 9 percent of the synthetic peptide in cis form, pT231-tau would work poorly as an immunogen for making cis-specific antibodies. First author Kazuhiro Nakamura and colleagues coaxed the peptide into cis configuration by adding a methylene group that converted proline’s pentagonal structure into a six-member ring. The scientists immunized rabbits with the modified peptide, then collected isomer-specific antibodies by purifying the resulting sera with either cis- or trans-locked phospho-tau peptides. The antibodies showed little cross-reactivity with the opposing isomer in protein binding assays and Western blots. They also immunostained neurons from mouse and Alzheimer’s brains.

While neither conformation of pT231-tau was detected in the brains of healthy elderly, the cis form cropped up in degenerating neurons of people with mild cognitive impairment, and accumulated in dystrophic neurites as disease progresses. In contrast, trans phospho-tau levels were lower, increased only slightly from MCI to AD, and hardly appeared in dystrophic neurites. “This suggested to us that the cis form is pathological and trans is not,” Lu said.

The isomers functioned differently as well. Trans pT231-tau bound tubulin and promoted microtubule assembly in vitro, as normal tau does; however, cis pT231-tau did not. In neuronal cells as well as brain tissue from transgenic mice and MCI patients, cis pT231-tau resisted dephosphorylation and degradation, and aggregated readily, unlike the trans isomer.

Given that Pin1 levels go down in AD, if the cis form of pT231-tau is indeed pathological, then it stands to reason that Pin1 protects against tau pathology by converting phospho-tau from cis to trans, Lu said. This scenario seemed to play out in vivo. When tauopathy mice were crossed to Pin1 transgenic mice (Lim et al., 2008), the double transgenic progeny had more trans pT231-tau and less of the cis form in their brains, compared to tau transgenic littermates with normal Pin1 levels. Conversely, crossing to Pin1 knockouts made tauopathy mice rack up more cis pT231-tau and less of the trans isomer.

“The finding that cis pT231-tau is toxic and that Pin1 can convert it to the non-toxic trans pT231-tau is exciting and has potential to be developed for early diagnosis and perhaps treatment of AD,” suggested Rakez Kayed of the University of Texas Medical Branch, Galveston, in an e-mail to the Alzforum. Julian Thorpe and Stuart Rulten of the University of Sussex, U.K., agreed, saying the data “make a good case for using targeted antibodies against the cis form of phospho-tau in immunotherapy to prevent AD progression.” (See full comment below.)

These scientists and others praised the generation of isomer-specific antibodies as novel tools with broad application. The study “is quite elegant,” commented Charles Glabe of the University of California, Irvine. “It is another good example of how conformation-dependent antibodies give important insights into diseases of protein conformation.” Jürgen Götz of the University of Sydney, Australia, noted, “It will be interesting to see whether the findings for a role of the cis phospho-Thr231 epitope in pathogenesis can be extended to tauopathies other than AD (see full comment below).” Lu said his team has not yet looked for cis and trans phospho-tau in other tauopathies.

However, they have unpublished data showing that the antibodies pick up cis phospho-tau in cerebrospinal fluid of AD patients—and not in normal elderly controls. “We may be able to use this pathologic conformation in CSF as a marker to select high-risk patients for cis-phospho-tau immunotherapy,” Lu said.

John Trojanowski of the University of Pennsylvania, Philadelphia, warned against putting too much stock in a single modification, noting that “tau is subject to lots of molecular mischief in AD and related tauopathies.” (See, e.g., ARF related news story on Cohen et al., 2011; ARF related conference story.)

Other work by Lu and colleagues suggests Pin1 may also guard against amyloid-β accumulation. Earlier, they found that isomerization of a proline in amyloid precursor protein (APP) curbs Aβ production (Pastorino et al., 2006). In research published this month, the scientists link the Aβ changes with Pin1. One study, which appeared in the March 2 Journal of Biological Chemistry, suggests that the isomerase enhances APP turnover by inhibiting glycogen synthase kinase-3β activity, thereby reducing Aβ production (Ma et al., 2012). Their second paper—posted online March 19 in the Journal of Alzheimer’s Disease—shows that loss of Pin1 function causes APP to move from the cell surface into intracellular compartments to undergo amyloidogenic processing (Pastorino et al., 2012). And in another study—also published this month in the Journal of Alzheimer’s Disease (Giustiniani et al., 2012)—Etienne-Emile Baulieu of INSERM, Paris, France, and colleagues find reduced levels of a different tau-binding prolyl isomerase, FKBP52, in brains of people with AD or frontotemporal dementia with parkinsonism-17 (FTDP-17).—Esther Landhuis.

Reference:
Nakamura K, Greenwood A, Binder L, Bigio EH, Denial S, Nicholson L, Zhou XZ, Lu KP. Proline Isomer-Specific Antibodies Reveal the Early Pathogenic Tau Conformation in Alzheimer’s Disease. Cell 30 March 2012;149:1-13. Abstract