Regulation of protein conformation by Pin1 offers novel disease mechanisms and therapeutic approaches in Alzheimer's disease

Abstract

Pin1 is a unique enzyme that changes the shape of target proteins by acting on specific amino acids that have been phosphorylated: serine or threonine residues that precede proline. Pin1 catalyzes the flip between two distinct orientations, called cis and trans, around the proline bond. This change in shape has profound effects on protein function and is a major signaling mechanism in the cell. Abnormal regulation of Pin1 has been associated with premature aging and multiple pathological processes, notably cancer and Alzheimerʼs disease (AD), two major age-related diseases. In AD, Pin1 affects two proteins thought to be key to disease pathology: the amyloid precursor protein (APP) and the microtubule-binding protein tau, by switching them from a dysfunctional shape (cis) back to a functional one (trans), which can be distinguished by cis and trans-specific antibodies. In the brains of people with AD, Pin1 is absent or inactivated and cis tau is accumulated at early stages of AD. In the absence of Pin1, APP is processed into toxic beta-amyloid and tau becomes misshapen to form tangles. As a result, Pin1-deficient mice develop age-dependent tau and Aβ pathologies and neuronal degeneration and loss. Thus, regulation of protein conformation by Pin1 has a critical neuroprotective role and offers a novel diagnostic and therapeutic target for AD. Notably, antibodies or vaccines specifically against the dysfunctional misshapen tau (while leaving the functional one untouched) may offer early diagnosis and treatment of AD and related disorders.

Over 100 years ago, Alois Alzheimer first described the hallmark lesions of the disease that bears his name: senile plaques and neurofibrillary tangles. Plaques are caused by the self-aggregation of beta-amyloid (Aβ) peptides - a product of the amyloid precursor protein (APP) - and tangles are formed of the microtubule-binding protein tau (Glenner et al., 1984; Goedert et al., 1992). The definition of AD requires the presence of both pathologies (Braak & Braak, 1990), although their relationship with each other and their precise role in the development of the disease remain obscure.

Familial AD, which accounts for less than 5% of cases, is uniformly associated with mutations in APP or genes involved in APP processing. It fits the profile of a genetic disease, with early onset and a Mendelian inheritance pattern. For many years, the field of AD research has been dominated by the hypothesis that late-onset AD is also primarily caused by an imbalance in the production and/or clearance of Aβ (Hardy & Selkoe, 2002). However, the vast majority of AD cases occur after age 65 and are not explained by mutations in genes associated with either APP or tau. The strongest risk factor for this “sporadic” type of AD is age itself. The development of plaques and tangles can be seen in “normal” aging (Davies et al., 1988), and very substantial amyloid plaque deposits can be found in older adults with normal cognition (Klunk et al., 2004; Schmitt et al., 2000).

Many molecules key to preserving neuronal health are impaired with increasing age through the process of thermodynamic entropy. Entropy leads to a loss of molecular fidelity of enzymes and other proteins, and in turn, to deficits in cellular metabolism, DNA-repair, and many other vital cellular functions (Demetrius & Driver, 2013). Pin1 is a novel enzyme that seems to play an important role in the maintenance of healthy aging and neuroprotection by correcting age-related changes in the shape of key proteins (Liou et al., 2011). Pin1-deficient mice are normal to mid-life and then begin to age prematurely. Along with osteoporosis, accelerated telomere shortening, and a wide variety of other age-related conditions, they develop age-dependent tau and Aβ pathologies and neuronal degeneration and loss (Lee et al., 2009; Liou et al., 2002; Liou et al., 2003; Nakamura et al., 2012; Pastorino et al., 2006). There is growing evidence that Pin1 plays a critical role in maintaining tau and APP in a healthy, functional form in the face of age and disease-related neuronal insults. In this article, we will review the nature and function of Pin1, explain its abnormality in AD, and discuss its possibilities as a target for novel diagnostic and therapeutic strategies.

Proline-directed phosphorylation

The addition of phosphate groups to proteins (phosphorylation) is a major method of cell signaling – the process of initiating, coordinating and controlling complex functions such as energy production and cell growth and survival. Phosphorylation can have profound effects on protein activity, function, and localization in the cell, both in health and disease presumably through changes in 3-dimensional protein structure (Pawson & Scott, 2005). However, the significance and regulation of such changes in protein conformation (shape) remained elusive until the recent discovery of Pin1. Phosphorylation of the amino acids serine or threonine that precede proline (pSer/Thr-Pro) is a central signaling mechanism that regulates many cellular processes (Lu & Zhou, 2007). This proline-directed phosphorylation is brought about by a large family of enzymes called protein kinases that includes the cyclin-dependent protein kinases (CDKs) necessary for mitosis, glycogen synthase kinase 3 (GSK3), an enzyme involved in glucose metabolism, and a large family of stress-activated protein kinases. The structure of proline includes a 5-membered ring in its chain of peptides which gives it the unique property of being able to flip its shape between two isomers (conformations), called cis and trans. Little was known about the importance of these conformational changes until the discovery of the enzyme Pin1 (Lu & Zhou, 2007).

Pin1 structure and function

Pin1 belongs to a family of enzymes called prolyl isomerases that catalyze the flip in orientation around the proline bond (Figure 1) and is unique because it exclusively binds to and isomerizes the specific pSer/Thr-Pro motifs regulated by proline-directed kinases (Lu et al., 1996; Lu & Zhou, 2007). After certain types of proline-directed phosphorylation, the rate of conversion between cis and trans is extremely slow due to the substantial energy required for conversion. Pin1 accelerates this rate by more than 1000-fold (Lu & Zhou, 2007; Pastorino et al., 2006). Thus, Pin1-catalyzed post-phosphorylation conformational change, rather than proline-directed phosphorylation per se, is the point of regulation for many phosphorylation signaling pathways. Pin1 can have various effects on target proteins as displayed in Table 1. It uses these multiple mechanisms to regulate a broad range of cellular processes including the cell cycle, cell signaling, transcription and splicing, DNA damage response pathways, germ cell development and neuronal survival (Lu & Zhou, 2007). It is clear that Pin1 activity and function are tightly regulated by multiple mechanisms, including gene transcription, protein phosphorylation, protein sumoylation and protein oxidization, and Pin1 can drive complex cellular processes by modulating multiple targets in different ways towards one direction. Pin1 can thus serve as a molecular “timer” or “switch,” turning proteins or entire pathways “on” or “off” at critical times (Lu et al., 2007).

Figure 1. Conformational change catalyzed by Pin1.

Pin1 has a two-domain structure that gives its unique substrate specificity. One domain specifically binds to the phosphorylated serine/threonine residue followed by a proline, and the other catalyzes the flip from a cis to a trans orientation around the proline bond. Pin1 is the only known enzyme thus far that can efficiently catalyze conformational change of the targets of proline-directed phosphorylation. H-bond = hydrogen bond. PT688 is the target of proline-directed phosphorylation, and other landmark amino acids are listed in black.

Table 1.

Pin1 regulates diverse cellular processes through multiple mechanisms including effects on protein processing, cellular localization, phosphorylation, stability, activity and interactions. Pin1 can catalyze changes from cis to trans or vice versa depending on the structure of the target protein. Pin 1 is strongly expressed in neurons where it plays a critical role in renewing APP and tau protein, the two proteins thought to be key to Alzheimer’s pathology. AD is characterized by the absence or dysfunction of Pin1. The proteins such as oncogenes, tumor suppressors, PKM2, β-catenin and NF-κβ are all involved in cell survival and/or proliferation. Pin1-mediated dysregulation of these proteins can lead to cancer through multiple pathways. Pin1 is therefore very tightly regulated in normal cycling cells. NF-κβ and IRAK1 are key regulators of inflammation and immunity.

Pin 1 Effect	Examples
Decrease protein stability	Tau, APP, multiple tumor suppressors
Increase protein stability	Multiple oncogenes
Increase protein dephosphorylation	Tau, Raf
Increase protein phosphorylation	RNA polymerase II
Decrease protein interaction	p-53, β-catenin
Increase protein interaction	Tau
Decrease enzymatic activity	GSK3beta
Increase enzymatic activity	CDC25C, IRAK1
Inhibit protein aggregation	Tau
Regulate protein processing	APP
Change protein localization	APP, PKM2, NF-κβ
Decrease transcription activity	Multiple tumor suppressors
Increase transcription activity	Multiple oncogenes

In the case of cell proliferation and transformation, Pin1 activates a number of growth enhancers and oncogenes while inactivating growth inhibitors and tumor suppressor genes to promote cancer development. Pin1 downstream targets include transcription activators and regulators, protein kinases, protein phosphatases, glycolytic enzymes, methyl transferases, lipid kinases, ubiquitin E3 ligases, DNA endonucleases and RNA polymerases(Lu & Hunter, 2014) . Pin1 is thus very tightly regulated in cycling cells as its overexpression and activation can cause carcinogenesis through a number of simultaneous mechanisms. Pin1 overexpression is frequently found in human cancers and is correlated with poor prognosis (Ayala et al., 2003; Bao et al., 2004). Thus, Pin1 might offer a promising unique anti-cancer target to stop numerous cancer-driving pathways at the same time.

Pin1 and neuronal health

In contrast to cycling cells, neuronal expression of Pin 1 increases during differentiation and remains unusually high throughout the lifespan (Hamdane et al., 2006; Liou et al., 2003). Although there is much to be discovered about the role of Pin1 in neurons, it is known to regulate a number of important proteins, including tau and APP. Pin1 knockout causes mice to develop a premature-age-dependent neurodegeneration similar to AD in humans that is characterized by tau hyperphosphorylation and an increase in the pathogenic processing of APP (Liou et al., 2003; Pastorino et al., 2006) On the other hand, overexpression of Pin1 in mature neurons protects against neurodegeneration induced by tau overexpression (Lim et al., 2008). Pin1 is inhibited in neurons affected by AD through stress-induced oxidation or phosphorylation by DAPK1 kinase, which is genetically linked to AD, and its expression is known to correlate with vulnerability to neurodegeneration (Liou et al., 2003). Age-related dysregulation of Pin1 thus provides a link between Aβ and tau abnormalities as well as a pathway to neuronal loss. There is also some genetic evidence that Pin1 is linked to AD risk. We found that a variant in the promoter of the Pin1 gene that increases its expression is associated with a 3 year delay in the age of onset of sporadic AD (Ma et al., 2012b). Together, these findings suggest that down-regulation or inactivation of Pin1 plays a critical role in the pathogenesis of AD.

Pin1 and tau protein

Tau protein helps with the assembly and stability of the microtubular structures that form the neuron’s “scaffolding.” In addition to providing key support for the unique neuronal structure, microtubules also provide transport channels for nutrients and other necessities down the long axon that connects the cell body to the terminals where synapses occur. When tau gets hyperphosphorylated, it becomes dysfunctional and can no longer interact with microtubules (Buee et al., 2000). The absence of functional tau leads to neurodegeneration and eventual cell death (Ballatore et al., 2007; Iqbal et al., 2009). In addition, hyperphosphorylated tau tends to self-aggregate into insoluble filaments that form neurofibrillary tangles. Once it has aggregated, tau is unable to return to its functional form. While tau hyperphosphorylation can occur as the result of normal aging, it is accelerated in the setting of neurodegeneration due to the increased activity of kinases that phosphorylate tau, such as CDK5 and GSK3β (Hernandez et al., 2010; Su & Tsai, 2011), or reduced activity of phosphatase that dephosphorylated tau such as PP2A (Liu & Gotz, 2013).

Phosphorylation of tau at the threonine (Thr) 231-proline motif (pT231-Tau) is an early step in pre-tangle pathophysiology that triggers misfolding of the protein and inhibits its ability to bind microtubules (Lu et al., 1999a; Luna-Munoz et al., 2007). It is therefore critical that the cell have mechanisms to help prevent or reverse tau hyperphosphorylation. We have shown that Pin1 binds mainly to the pT231-Pro motif in tau and greatly accelerates its cis to trans isomerization to restore the ability of pT231-Tau to promote microtubule assembly (Lu et al., 1999b), to facilitate pT231-Tau dephosphorylation (Liou et al., 2003; Zhou et al., 2000), to promote pT231-Tau degradation (Lim et al., 2008) and to prevent pT231-Tau aggregation (Liou et al., 2003) in vitro and in vivo. Importantly, Pin1 knockout and overexpression have the opposite effects on T231 phosphorylation before any other sites, which can also explain their opposite effects on tau-related pathologies both in neuron cultures and in mouse models (Lim et al., 2008; Liou et al., 2003; Pastorino et al., 2006). In fact, Pin1 neither binds tau nor affects tau function when T231 is mutated to a non-phosphorylatable amino acid (Lim et al., 2008; Lu et al., 1999a; Zhou et al., 2000) These results indicate that the pTh231-Pro motif is the prime action site for Pin1 in tau. In addition, Pin1 can prevent tau hyperphosphorylation by inhibiting tau kinases such as GSK3-beta (Ma et al., 2012a). Thus, it seems that Pin1 uses multiple mechanisms to prevent tau from being hyperphosphorylated, restore the biological function of hyperphosphorylated tau and prevent hyperphosphorylated tau from gaining toxic function to protect against neurodegeneration in AD.

Pin1 and APP processing

While APP and Aβ have been heavily studied from the perspective of AD pathogenesis, investigation of their normal function in the neuron is sorely needed. APP seems to have a role in cellular signaling and Aβ is released in response to various types of neuronal stress including brain aging, oxidative damage and ischemic and traumatic injury (Bero et al., 2011; Castellani et al., 2009; Magnoni et al., 2012; Pluta et al., 2013). In familial AD, mutations in APP or presenilin genes cause alterations in APP processing and lead to the formation of a longer form of the Aβ peptide that is more prone to aggregation. An increasing body of evidence, including the recent failure of clinical trials of therapies that target Aβ, raise questions about the amyloid hypothesis in the pathogenesis of sporadic AD (Aisen et al., 2013; Benilova et al., 2012).

Interestingly, the processing of APP is can also be regulated by phosphorylation, and proline-directed phosphorylation by CDK5 (Cruz et al., 2006; Lee et al., 2003; Lee & Tsai, 2003) and GSK3β (Phiel et al., 2003) increases amyloid pathology. Phosphorylation at this site changes the conformation of APP from trans to cis, yielding a pathogenic form of Aβ when it is processed. We have shown that Pin1 binds to Threonine 668 (pThr668-Pro) on APP and regulates its conformation in a way that favors healthy processing through cis to trans conversion, thus favoring healthy Aβ processing (Pastorino et al., 2006). Studies of APP processing in Pin 1 knockout mice showed an age-dependent shift in APP processing that led to a massive buildup of insoluble Aβ(Pastorino et al., 2006). Our subsequent experiments have shown that Pin1 helps APP stay anchored at the plasma membrane, favoring the non-amyloidogenic processing, and reducing the internalization and thus the amyloidogenic processing of APP (Pastorino et al., 2012). Moreover, Pin1 can also increase protein turnover of APP by inhibiting GSK3β activity to phosphorylate APP on the pThr668-Pro motif (Ma et al., 2012a). This finding is highly significant because increasing APP gene dosage or expression has been shown to cause familial early-onset AD (Ma et al., 2012a; Theuns et al., 2006). Thus, Pin1 again uses multiple mechanisms to protect against age-dependent Aβ toxicity by accelerating the cis and trans conformational changes of APP. The role of Pin 1 in maintaining neuronal health is summarized in Figure 2.

Figure 2. Pin1 in neuronal health and disease.

Pin1 helps to protect tau and APP from the phosphorylation which can cause protein dysfunction by suppressing glycogen synthetase kinase 3 (GSK3). Once tau or APP is phosphorylated, they can exist in two shapes – cis or trans. The trans isomer is still functional and can be dephosphorylated. However the cisconformation is resistant to dephosphorylation. Cis-ptau begins to self-aggregate and cannot be degraded, while cis-pAPP is processed into the pathogenic form of Aβ. Pin1 facilitates conversion from unhealthy cis back to functional trans and also facilitates the dephosphorylation of ptau and pAPP. If Pin1 is absent or deficient, GSK3 hyperphosphorylates tau and the cis conformation accumulates, eventually leading to plaque and tangle formation, neurodegeneration and cell death.

Conformation-specific antibodies

Protein phosphorylation and dephosphorylation are readily detectable due to a change in the phosphate group, and their value in the pathogenesis, diagnosis and therapy of diseases have been well accepted. However, there was no tool available to detect Pin1-catalyzed post-phosphorylation conformational changes and their conformation-specific function in cells until our recent discovery of cis and trans conformation-specific antibodies (Nakamura et al., 2012).

We have developed innovative technology that allows the generation of cis- and trans-specific antibodies and used them to raise antibodies specific for cis and trans isomers of the pThr231-Pro motif in tau (Nakamura et al., 2012). Using these antibodies, we have provided the first direct evidence that 1) trans tau is functional and able to interact with microtubules, 2) the cis isomer loses normal biological function and also gains toxic functions, and 3) Pin1 prevents AD tau pathology by flipping the dysfunctional cis to physiological trans in vitro and in vivo. More importantly, we have discovered that cis, but not trans, tau appears extremely early in the brain of humans with mild cognitive impairment, an early form of AD. As the disease progresses, cis, but not trans, tau accumulates exclusively in diseased neurons and localizes to dystrophic neurites, which are early hallmarks of AD and are highly correlated with cognitive loss in AD patients. Thus, cis tau is the previously unrecognized early pathogenic tau that leads to the development of tau pathology and memory loss in AD, offering a novel diagnostic and therapeutic target for the disease (Nakamura et al., 2012).

Diagnostic and therapeutic implications

Our exciting new insight into the role and regulation of p-tau conformations by Pin1 in AD might have important and novel therapeutic implications. Detecting these abnormal conformational imbalances might offer a new opportunity to identify individuals with early and even presymptomatic disease, predict AD progression and monitor treatment response. Notably, Thr231 phosphorylation is the earliest detectable tau phosphorylation event in human AD (Luna-Munoz et al., 2007). Its levels in CSF are among the most promising early biomarkers for AD, and correlate with memory loss, tangle accumulation and hippocampal atrophy, and predict conversion of MCI to AD. However, there is large variability in pT231 levels between individuals, making it difficult to standardize the test (Ewers et al., 2007; Hampel et al., 2010). Our findings that cis, but not trans, pT231-tau appears early in MCI and is pathogenic suggest that cis pT231-tau and its ratio with trans might improve the sensitivity and specificity of tau as a biomarker in spinal fluid or blood (Nakamura et al., 2012). Furthermore, strategies to increase Pin1 expression or decrease its inhibition in neurons might decrease the production of pathogenic conformations of both of the hallmark proteins of AD. More importantly, antibodies that specifically target the pathogenic conformation of tau but spare the healthy one could be developed, and might be used to prevent or treat AD in its pre-symptomatic stages (Nakamura et al., 2012). In this regard, active and passive immunization against p-tau tangle-containing epitopes show promising efficacy in mouse models (Rosenmann, 2013). However, since neuronal dysfunction occurs long before tangle formation, neutralizing antibodies and vaccines specifically against the extremely early and pathogenic cis p-tau, while leaving the healthy trans untouched, might be highly efficacious and specific for halting or even preventing tauopathy and memory loss relevant to MCI and AD at the beginning (Nakamura et al., 2012).

In summary, our work illustrates that the ability of Pin1 to prevent the accumulation of certain proteins such as tau and APP in their pathogenic conformation is vital to protecting against age-dependent neurodegeneration in AD. More work is needed to determine the upstream factors that regulate the activity of Pin1 during aging and Pin1 downstream targets and their function. The availability of conformation-specific antibodies will help to further define the role of Pin1 in AD pathogenesis and opens new horizons for diagnosis and therapy.