Fighting Protein misfolding diseases

Cells function as highly accurate quality control (QC) machines to ensure that only correctly folded proteins are released into the physiological milieu to perform their designated functions. The efficient removal of damaged or incorrectly folded/misfolded proteins at the correct time keeps a cell viable and functioning. In the natural environment of the cell, proteins are degraded under two conditions:

• Mutation in the primary sequence.

• Postsynthetic events leading to damage to the protein architecture.

¹

Only 65–70% of the newly synthesized proteins survive the safeguards imposed by the cellular machinery. A majority of the remaining proteins are degraded within minutes of their synthesis because they are incorrectly folded. The purpose of this review is to present an overview of the diseases that arise from the incorrect folding of proteins.

Protein misfolding diseases

Cells have adequate safeguards in place to refold unfolded molecules, to prevent aggregation of unfolded molecules, or to hydrolyze the recalcitrant species into their constituent monomers. The presence of these elaborate systems indicates that a failure would result in a life-threatening situation for the cell. Incorrect folding of particular proteins appears to result in disease conditions that are grouped under the all-encompassing title of protein misfolding diseases (Table 1). Some of these diseases are caused by the absence of a protein function, as the misfolded protein has been removed by the cell via proteasome degradation. The most famous example in this class is the misfolded cystic fibrosis transmembrane regulator (CFTR) protein whose loss causes cystic fibrosis (CF).¹

Table 1 A list of protein misfolding diseases.

Other examples include Fabry's disease (FD) where the misfolded α-galactosidase is eliminated by the cellular machinery. The occurrence of some cancer types is also attributed to this route.¹ In other cases, diseases are characterized by the extracellular deposition of misfolded proteins. The major feature of this latter class of diseases, termed 'amyloidoses', is the amyloid fibril. The assembly fibrils (from protofibrils) into plaques (Figure 1) and the pattern of fibrillation and plaque formation is eerily similar in all the cases, even though a different protein is implicated in each of them.

Figure 1

In this class of diseases, the pathogenic element has not been clearly identified — it is not readily apparent if the deposition of fibres is a symptom of the disease or if it is the underlying cause of these symptoms. Thus, the pathogenicity could be a result of the 'loss of function' as the native functional protein is not formed, or it could be attributed to the toxicity brought about by amyloid plaque formation. The misfolded proteins induce pathogenesis when the cell's ability to process them is overwhelmed. Even though the cellular machinery is not able to get rid of these species, there is evidence that it is still involved in the formation of these dense insoluble particles. The chaperones and components of the ubiquitin-proteasome pathway,² including the ubiquitin–conjugated rogue protein, are all components of these macrostructures termed aggresomes.³ Many misfolded proteins, belonging to both the classes (loss-of-function or aggregates with cytotoxic effects), have been detected in these structures.

The structure of the aggresome is stabilized by formation of extensive β-sheets held together by hydrogen bonds, and is a characteristic feature of all amyloid structures. At present, no common denominator has been found for the primary, secondary or tertiary structures of the proteins that give rise to these structures. These inclusion particles are dynamic systems where one unfolded/misfolded protein is continually being replaced by the other. It has been shown that if the production of the unfolded protein is terminated, for example, by the removal of proteasome inhibitors in in vitro systems, the cells are able to remove these misfolded proteins. It appears that aggresomes function as segregation units that keep the cellular components separated from potentially toxic elements.

Accumulation of misfolded proteins in selective regions of the brain is related to the development of a number of neurodegenerative diseases. Alzheimer's disease (AD), Parkinson's disease (PD), amyotrophic lateral sclerosis (ALS), prion diseases and the polyglutamine diseases all represent proteinopathies — diseases in which a protein or set of proteins misfolds and aggregates.

Polyglutamine diseases. One surprising feature of these diseases is the appearance of symptoms in a limited number of cell types. For example, Huntington's disease (HD), with its hallmark choreiform movements and selective loss of neurons in the striatum, is caused by a dominantly inherited expanded coding for glutamine (CAG) repeat that extends an N-terminal polyglutamine segment in the protein huntingtin.⁴

The gene for the mutant protein is expressed in neurons and some other cells. However, the mutant protein appears as aggresomes only in the striatum and cerebral cortex of the brain, which implies that the mutant protein is successfully eliminated in the rest of the cells. Elucidation of the reason behind the selective removal of misfolded aggregates will go a long way in designing the therapeutic course of action. Another feature of HD is the appearance of smaller fragments containing polyglutamine stretches that result from proteolytic cleavage of the mutant protein. Thus, it appears that even though the proteolytic machinery has been able to act on the misfolded protein, the cell is unable to provide conditions that prevent oligomerization of the cleaved fragments, either among themselves or with other proteins such as huntingtin-interacting proteins (HIP 1 and 2).⁵ This is also the case for the mutant ataxin-1, which causes spinocerebellar ataxia type I. This protein is also characterized by a long polyglutamine stretch. In this case, proteasome degradation of the mutant protein does occur, but is inefficient, although the rate of ubiquitin conjugation is normal.

Alzheimer's disease. With an increase in the average age of the population, one disease that has assumed prominence is AD, characterized by extracellular senile plaques. The main component of the plaque is the β-amyloid (Aβ) peptide, which is generated following the cleavage of an amyloid precursor protein (APP) by the β-site APP cleavage enzyme (BACE) and presinilins (γ-secretase). The most acceptable hypothesis talks of the deposition of Aβ peptide in the parenchyma of the brain, leading to neural loss and dementia, the classical symptoms of AD.⁶

The mutations in APP are centred on the APP cleavage sites and generally induce a change in the ratio of Aβ₄₂ to Aβ₄₀ generated, in favour of the former. The occurrence of the more hydrophobic Aβ₄₂ (containing two extra amino acid residues) leads to faster aggregation and is the major component of senile plaques. The second major component is the intracellular neurofibrillary tangles (NFTs), which comprise of paired helical filaments of the microtubule-associated protein tau and are ubiquitinated. There is also evidence that Aβ accumulates intraneurally.⁶ Both these factors indicate that the cellular machinery does try, but is unable to cope, with the increased production of Aβ, leading to its deposition.

Parkinson's disease. PD is characterized by Lewy bodies composed of filamentous aggregates of the presynaptic protein, α-synuclein. As the autosomal-dominantly inherited PD is caused by mutations in the gene for α-synuclein, increased expression, aggregation and accumulation of α-synuclein plays a key role in the neurodegeneration associated with the disease. Oxidative damage to α-synuclein gives rise to species with nitrated tyrosine residues, and these form a major chunk of the Lewy bodies. Similar to the cases described earlier, the proteinaceous inclusions are also found to contain large amounts of proteasomes and ubiquitin, pointing again to the failure of the cellular machinery to tackle with the covalently modified protein. Another interesting point is raised by the protein parkin, which is an enzyme that performs ubiquitin ligation. This enzyme is associated with the hereditary form of PD and it would be worth discovering if α-synuclein acts as a substrate for parkin.

Therapeutic intervention

As protein misfolding diseases share certain common structural features, so therapeutic intervention aimed at their treatment may also follow some common strategies (Table 2). Many of these diseases are spread over a large part of the body, thus localized intercession may not work, and an approach based on small molecules may have to be explored. Some success at the preclinical stage using this approach has been reported. For example, trehalose has been found to stabilize partially unfolded extended polyglutamine-containing huntingtin in a transgenic mouse model of HD.⁷ Oral administration of trehalose showed a reduction in huntingtin aggregation in the cerebrum, with improvement in motor dysfunction.⁷ It is probable that trehalose will act as a prophylactic to alleviate the disease symptoms before they occur, or at least delay the age of onset.

Table 2 Common strategies for therapeutic intervention in protein misfolding diseases.

Another small molecule, Congo Red (C₃₂H₂₂N₆Na₂O₆S₂), which binds to amyloid fibrils, has been known to reduce the concentration of the pathogenic prion protein PrP^Sc and to ameliorate symptoms of HD by decreasing oligomerization of mutant huntingtin as well.^8,9 Efforts have focused on compounds that inhibit conversion to the disease-causing conformation and on compounds that help the cell to get rid of the aggregated protein, reducing toxicity. Pharmacokinetic approaches have identified quinacrine, an acridine compound that is able to cross the blood–brain barrier and to decrease the concentration of PrP^Sc in infected neuroblastoma cells.¹⁰ Trials will now be performed on patients with Creutzfeldt-Jakob disease (CJD), a prion disease, to test the efficacy of this compound.

At present, only two accepted treatment therapies for AD are available: acetylcholine esterase inhibitors and memantine, the noncompetitive glutamate receptor antagonist (Ebixa, Lundbeck; Namenda, Forest Lab) — neither of which is particularly effective. Cleavage of APP leads to the release of the pathogenic moiety, and therefore inhibition of BACE and γ-secretase constitutes one of the most tried and tested routes to therapeutic intervention. Nonsteroidal drugs such as R-flurbiprofen (Flurizan) are purported to lower the level of Aβ₄₂ by manipulating γ-secretase activity.¹¹

Phase III clinical trials are underway with this drug for AD treatment. One potential problem with the inhibition of γ-secretase is that it is involved in Notch signalling. Cleavage of the transmembrane Notch protein by γ-secretase generates a peptide that translocates to the nucleus to regulate gene expression. Inhibition of γ-secretase may affect Notch-mediated gene regulation. γ-Secretase is a complex comprising presinilin-I, which cleaves APP to generate Aβ, nicastrin and APH-2, which inhibit production of Aβ, and PEN-2, promoting production of Aβ. Therefore, PEN-2 may be a more meaningful target for inhibition.

Key points

Data of Phase I clinical trials from Eli Lily have shown that LY450139, a γ–secretase inhibitor, is successful in decreasing the concentration of Aβ in the blood by 38%, but has no effect on its level in the cerebrospinal fluid (CSF).¹² The exact point of action of the drug is still being studied and the drug is currently in Phase II clinical trials. Conversly, the aspartate protease BACE appears to be a more validated target for inhibition of Aβ production because:

• proteases such as rennin have already been used successfully in clinical trials.

• mouse knockout models for BACE show reduced production of Aβ with no concomitant side-effects.

A recent study has focused on another enzyme for which Aβ is the substrate. PKC, a serine/threonine kinase, which acts as a receptor for phorbol esters, was able to reduce plaque density and astrogliosis in transgenic mice.¹³ The activities of other enzymes, which use Aβ as a substrate, remained unchanged. It was hypothesized that amyloid deposition was decreased because of increased endothelin-converting enzyme (ECE)-mediated breakdown of Aβ. Thus, both PKC and ECE are putative targets for treatment of AD. Aggregation of Aβ is a kinetic process, and therefore it is not necessary to eliminate all the monomers, but to keep the concentration of monomeric Aβ below a threshold value.

The deposition of Aβ is promoted by its interaction with proteoglycans. Thus, glycosaminoglycan mimetics, which bind to Aβ and prevent its interaction with proteolycans, would constitute good drug molecules. A sulfated glycosaminoglycan mimetic, NC-531 (Alzhemed, Neurochem), inhibits formation of senile plaques and results in a reduced level of Aβ₄₂ in a dose-dependent manner.¹⁴ This drug is under development for treatment of Alzheimer's disease and has now reached Phase III clinical trials. Preliminary data of treatment with statins have shown minor improvement in cognitive functions with simultaneous decrease in the concentration of Aβ in CSF. Unfortunately, this successful approach could not be replicated when the sample size was increased. Another attempted approach is the use of metal chelators such as clioquinol, originally prescribed for parasitic gastrointestinal ailments, which disentangle β-amyloid fibrils by sequestering Zn²⁺ and Cu²⁺ .¹⁵

Phase II trials with this putative drug have led to lower level of the hydrophobic Aβ₄₂ in plasma coupled with improved cognitive ability. Treatment of Chinese hamster ovary cells overexpressing APP with clioquinol and Cu²⁺ or Zn²⁺ resulted in ~85–90% reduction of secreted Aβ.¹⁶ Analogous effects were seen in neuroblastoma cells overexpressing APP. The secreted Aβ was rapidly degraded through upregulation of matrix metalloproteinase-2 (MMP-2) and MMP-3 after addition of the drug and metal ions. Immunization with Aβ peptides as immunogens (active and passive) is a desirable option.

Results with transgenic mice have shown dramatic clearance rates for senile plaques.¹⁷ Improvement in cognitive functions has also been observed. Regrettably, active immunization of AD patients with Aβ₄₂ as the immunogen generated antibodies, which crossed the blood–brain barrier and caused neuroencephalitis in a small number of patients because of which the trials were halted. Hope however has not dimmed as there is a possibility that protein engineering may be able to overcome this deficiency. In a similar manner, inhibition of α-synuclein using small molecule inhibitors has proved to be an attractive strategy for therapeutic intervention in PD. Screening for small molecules that inhibit fibrillization of α-synuclein generated molecules that were chiefly catecholamines related to dopamine. The inhibitory activity of dopamine and L-DOPA could be reversed by the addition of antioxidants, hinting at the oxidation of catechol.¹⁸ The formation of dopamine-α-synuclein adduct may provide a clue to the role of dopaminergic neurons in PD.

Conclusion

Despite the advantage of sharing common structural elements by all the protein misfolding diseases, a cure to any one of them would still require a gigantic effort from all the parties concerned. These include the research fraternity as well as the pharmaceutical companies. As the pathways of misfolding are better understood, they will open routes for rational design of drugs to tackle them. The hypothesis that the symptoms of many of these diseases appear with advancing age because of lower levels of chaperones during ageing also needs to be looked into.

Ipsita Roy

is an assistant professor at the National Institute of Pharmaceutical Education and Research (NIPER), Department of Biotechnology (India).

References

1. A.L. Goldberg, Nature, 426(6968), 895–899 (2003).

2. M.H. Glickman and A. Ciechanover, Physiol. Rev., 82(2), 373–428 (2002).

3. J.A. Johnston, C.L. Ward and. R. Kopito, J. Cell. Biol.143(7), 1883–1898 (1998).

4. A. Young, "Huntington disease and other trinucleotide repeat disorders," in J. Martin, Ed., Scientific American Molecular Neurology (Scientific American Inc., New York, NY, USA, 1998).

5. G.P. Bates, L. Mangiarini and S.W. Davies, Brain Pathology, 8(4), 699–714 (1998).

6. J. Hardy and D.J. Selkoe, Science, 297(5580), 353–356 (2002).

7. S. Sellarajah et al., J. Med. Chem., 47(22), 5515–5534 (2004).

8. M. Tanaka et al., Nat. Med., 10(2),148–154 (2004).

9. I. Sanchez, C. Mahlke and J. Yuan, Nature, 421(6921), 373–379 (2003).

10. K. Doh-Ura, T. Iwaki and B. Caughey, J. Virol., 74(10), 4894–4897 (2000).

11. G. P. Lim et al., J. Neurosci., 20, 5709–5714 (2000).

12. D.M. Skovronsky, V.M.Y. Lee and J.Q. Trojanowski, Annu. Rev. Pathol. Mech.1, 151–170 (2006).

13. D.S. Choi et al.,Proc. Natl. Acad. Sci. USA, 103, 8215–8220, (2006).

14. H. Geerts, Curr. Opin. Investig. Drugs, 5(1), 95–100 (2004).

15. R.A. Cherny et al.,J. Biol. Chem., 274(33), 23223–23228 (1999).

16. A.R. White et al.,J. Biol. Chem., 281(26), 17670–17680 (2006).

17. F. Bard et al.,Proc. Natl. Acad. Sci. USA, 100(4), 2023–2028 (2003).

18. K. A. Conway et al., Science, 294(5545), 1346–1349 (2001).