Neurodegenerative diseases are characterized by complex cellular perturbations involving synaptic, axonal, and mitochondrial dysfunction as well as transcription changes, among others. In contrast to these disruptive events, misfolded proteins can also launch adaptive, protective responses, including inflammation, Ubiquitin–Proteasome-dependent protein degradation, autophagy, and UPR. We are particularly interested in understanding the role of the UPR in disease because several recent studies have linked ER stress to some of the most prevalent neurodegenerative diseases, such as AD, PD, and amyotrophic lateral sclerosis (ALS) [28
]. For instance, the brains of AD patients accumulate elevated levels of the ER chaperone Grp78/BiP, and phosphorylation of the UPR sensor PERK and its target eIF2α [8
]. In addition, the ER chaperone PDI and phospho-eIF2α are elevated in the brain of PD patients [10
] and in the spinal cord of ALS patients [20
has only recently been used as a UPR marker based on the diagnostic value of the small intron regulated by the IRE1 sensor. XBP1s
is elevated in the frontal cortex of AD patients, but not in mice expressing mutant APP [24
]. We also showed that transgenic flies expressing human Aß42 and rat PC12 cells treated with Aß42 oligomers induce unconventional splicing of XBP1
]. Moreover, reduction of endogenous XBP1
increased Aß42 toxicity in flies, while XBP1
misexpression ameliorated it [6
]. In a chemical model of PD, mice treated with the toxin MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) exhibited XBP1
upregulation in the brain, while adenoviral expression of XBP1s
protected dopaminergic neurons in these mice [29
]. Mice inoculated with several strains of prions showed increased levels of the XBP1s
], suggesting the involvement of the IRE1-XBP1 pathway in PrP pathogenesis. These recent results suggest that XBP1
is activated in tissues undergoing neurodegeneration and support the idea that XBP1
activation is a neuroprotective response to amyloid insults. However, patient and animal studies are not ideal models to identify the conformations and assemblies directly responsible for inducing ER stress.
The purpose of the present study was threefold: (1) To develop a sensitive assay to detect XBP1 activation at the RNA level, (2) compare the ability of several amyloidogenic proteins to induce XBP1 splicing in the same experimental conditions, an (3) determine which assemblies are responsible for this activity. We show here that amplifying both XBP1 isoforms by RT-PCR and running the PCR products on polyacrilamide gels eliminates the XBP1u/XBP1s hybrid, thus removing the main obstacle to exploiting RNA isoforms for diagnostic purposes. Our results demonstrate that changes in the relative abundance of XBP1 isoforms are highly reproducible, supporting the use of RNA to accurately determine XBP1 unconventional splicing. We are currently developing a quantitative PCR method to increase the sensitivity and precision for detecting XBP1 splicing.
To answer the next two questions, we first confirmed that oligomeric preparations from α-Syn, PrP106–126, and ABri1-34 induced similar levels of cell toxicity (around 50 % lethality). On the other hand, monomers showed no toxicity at all and fibers induced a small but significant cell loss. These results support the idea that oligomers from different protein sources share unique biological properties that make them highly toxic. In contrast to the consistent cell toxicity of oligomers, the ability to induce XBP1 activation was sequence-dependent. Of all the conditions tested, only α-Syn oligomers were potent inducers of XBP1s, resulting in a dramatic decrease in the levels of XBP1u. α-Syn fibers induced slightly higher levels of XBP1s than the untreated cells, but that effect was very modest compared to the oligomers. On the other hand, PrP106–126 and ABri1-34 assemblies were poor inducers of XBP1 splicing. However, PrP106–126 and ABri1-34 oligomers induced a mild transcriptional upregulation of XBP1u, which could be due to the activation of other ER stress sensors, like ATF6, which is a known transcriptional regulator of XBP1. α-Syn may also induce transcriptional activation of XBP1, but since most of it is spliced, we do not appreciate an increase in XBP1u. These experiments uncover unexpected differences among amyloidogenic proteins, subdividing them into those that induce potent XBP1 splicing (α-Syn, Aß42) and those that do not (PrP106–126, ABri1-34).
If the ability to induce XBP1 splicing is highly dependent on specific structures only found in some oligomers, why did α-Syn fibers induce a slight activation of XBP1? There are two possible explanations for the mild effect of α-Syn fibers. One is that the fibril preparations may contain small amounts of pre-aggregated oligomers or that the oligomers are actively released from fiber breakage. This small amount of oligomers may explain the weak activation of XBP1, suggesting that fibers have no role in the induction of ER stress. Alternatively, highly pure fibrillar preparations may be directly responsible for XBP1 activation, arguing for the preservation of oligomeric structures in the fibers that allow them to interact with the same cellular pathways.
Whereas all oligomers showed similar cell toxicity, a highly specific biological assay (XBP1
activation) uncovered the contribution of the protein sequence to the activity of oligomers from four protein sources. The different ability of Aß42 and α-Syn oligomers to induce XBP1
splicing compared to ABri1-34 and PrP106–126 oligomers support the existence of some degree of variation in the conformation of these two groups of oligomers. Unfortunately, it is unclear at this point what makes Aß42 and α-Syn capable of activating IRE1-XBP1 and why ABri1-34 and PrP106–126 do not. The available experimental evidence suggests that there may be little structural variation among the oligomeric conformations. This is supported by the ability of a few conformational antibodies to recognize multiple oligomeric species obtained from synthetic or biological sources and prepared by different methods [12
]. These results argue for the existence of few stable conformations compatible with the formation of neurotoxic oligomers. Also, most oligomers show the ability to perturb membrane integrity and disrupt ion metabolism [11
], pointing to common biological activities [13
]. Since activation of UPR requires the perturbation of an internal organelle (the ER), exogenous Aß42 and α-Syn may be more efficiently transported into the ER by endocytic mechanisms. If this were the case, this would indicate the differential recognition of some oligomeric conformations, but not all, by specific receptors or transporters. Thus, we report here that XBP1 and the ER stress play different roles in neurodegenerative diseases, although the mechanisms underlying these differences are not clear. Additional structural approaches in the future may contribute to resolve in more detail the similarities and differences among these conformers critical in many chronic disorders.
Overall, we report here a strong connection of α-Syn to induction of ER stress and the XBP1-IRE1 pathway. Importantly, α-Syn misfolding and aggregation is an salient pathological feature of other neurological disorders, including dementia with Lewy bodies and multiple systems atrophy, suggesting that ER stress may be a common component of other synucleinopathies. Thus, identification of the signals that result in UPR and amelioration of this cellular response may contribute to the treatment of several synucleinopathies. On the other hand, there seems to be less consensus on the role of ER stress in prion diseases. The inability of PrP106–126 to induce XBP1
splicing agrees with the observation that elimination of XBP1
in mice did not alter the course of prion disease [16
], suggesting that XBP1
plays no physiological role in prion diseases. Finally, FBD is a rare dementia and little is known about its specific pathobiology. Our results indicate that despite the strong similarities between Aß42 and ABri1-34 (two small, secreted, amyloidogenic peptides that cause neurodegeneration), they may cause toxicity through different cellular pathways. In conclusion, we describe here the differential activation of XBP1
by four amyloidogenic proteins, suggesting a complex involvement of UPR in disease, a pathway that in the last few years has been connected to a wide array of human maladies, including cancer, ischemia, and several chronic disorders [38