62

“Fatal Attractions” of Proteins

A Comprehensive Hypothetical MechanismUnderlying Alzheimer’s Disease and Other Neurodegenerative Disorders

JOHN Q. TROJANOWSKIa AND VIRGINIA M.-Y. LEE

The Center for Neurodegenerative Disease Research, Department of Pathology and Laboratory Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104-4283, USA

ABSTRACT: Abnormal protein–protein interactions that result in the for-mation of intracellular and extracellular aggregates of proteinaciousfibrils are common neuropathological features of many, albeit diverse,neurodegenerative disorders, such as sporadic and familial Alzheimer’sdisease, Parkinson’s disease, amyotrophic lateral sclerosis, and prion en-cephalopathies. Indeed, increasing evidence suggests that abnormal pro-tein–protein interactions and/or the lesions that result from theaggregation of pathological protein fibrils could play a mechanistic role inthe dysfunction and death of neurons or glial cells in neurodegenerativediseases. Here we propose that “fatal attractions” between brain proteinsare the key pathological events underlying Alzheimer’s disease and a largenumber of other seemingly diverse neurodegenerative disorders. This hy-pothesis predicts that the abnormal interaction between normal brain pro-teins alters their conformation and promotes the assembly of thesepathological conformers into filaments that progressively accumulate asintracellular or extracellular fibrous deposits in the central nervous sys-tem. Further, the transformation of the normal proteins into pathologicalconformers is predicted to result in losses of critical functions, and the dis-ease proteins or their progressive accumulation into filamentous aggre-gates are predicted to acquire neurotoxic properties, all of which culminatein the dysfunction and death of affected brain cells. Thus, the “fatal attrac-tions” hypothesis describes a plausible unifying mechanism that accountsfor the onset/progression of Alzheimer’s disease and a large number of oth-er seemingly unrelated neurodegenerative disorders characterized neuro-pathologically by filamentous brain lesions formed by different proteins.

KEYWORDS: Tauopathies; Synucleinopathies; Filamentous inclusions;Protein aggregates

aAddress for correspondence: J.Q. Trojanowski, M.D., Ph.D., or V. M.-Y. Lee, Ph.D., Centerfor Neurodegenerative Disease Research, Department of Pathology and Laboratory Medicine,University of Pennsylvania School of Medicine, HUP, Maloney Building, 3rd floor, Philadelphia,Pennsylvania 19104-4283. Voice: 215-662-6399 or -6427; fax: 215-349-5909.

trojanow@mail.med.upenn.edu [or] leevmy@mail.med.upenn.edu

63TROJANOWSKI & LEE: “FATAL ATTRACTION” OF PROTEINS

Recognition of a common mechanistic theme shared by Alzheimer’s dis-ease (AD) and a number of other seemingly unrelated neurodegenerative dis-orders began to emerge towards the close of the 20th century as the pace ofresearch on these disorders continued to accelerate in the last decades of thisepoch (for recent reviews of AD and a select group of other neurodegenera-tive conditions linked by this theme, see References 1–12). As a result, thepessimistic notion that each of the numerous phenotypically and genotypical-ly distinct degenerative disorders of the aging brain would require a different,disease-specific therapeutic intervention has begun to dissipate.7,8,12–14 In-deed, since a large number of these disorders are characterized neuropatho-logically by intracellular and/or extracellular aggregates of proteinaciousfibrils that are implicated in progressive brain degeneration (TABLE 1), thesedisorders may share similar targets for drug discovery. Thus, despite differ-ences in the molecular composition of the structural elements of these fila-mentous lesions as well as the brain regions and cell types they affect, agrowing body of evidence adds increasing support to the hypothesis that sim-ilar pathological mechanisms may underlie all of these disorders.8,12 Morespecifically, the onset and/or progression of brain degeneration in AD andother neurodegenerative disorders may be linked mechanistically to abnor-mal interactions (“fatal attractions”) between brain proteins that lead to theirassembly into filaments and the aggregation of these filaments within braincells or in the extracellular space as fibrous inclusions or plaque-like depos-its, respectively (TABLE 1).

These filamentous lesions are exemplified by intranuclear neuronal inclu-sions formed by diverse proteins harboring mutant or abnormally expandedpolyglutamine tracts in hereditary trinucleotide repeat disorders, intracyto-plasmic neurofibrillary tangles (NFTs) as well as extracellular amyloid or se-nile plaques (SPs) in sporadic and familial AD (FAD), and by prion proteindeposits in sporadic or genetic forms of spongiform encephalopathy.1–12 Al-though many of these filamentous lesions are recognized as diagnostic hall-marks of specific disorders (TABLE 1), sporadic AD and FAD illustrate someof the complex and poorly understood overlap among these neurodegenera-tive diseases. For example, the heterogeneous dementing disorders classifiedas AD overlap with a large group of distinct neurodegenerative disorders(known as tauopathies) that are characterized by prominent tau-rich tanglepathology throughout the brain as well as with another diverse group of dis-orders (known as synucleinopathies) that are characterized by prominent sy-nuclein brain pathology (TABLE 1). Thus, while the diagnostic hallmarks ofAD are numerous SPs composed primarily of Aβ fibrils and intraneuronalNFTs formed by aggregated tau filaments, NFTs are similar to the filamen-tous tau inclusions characteristic of neurodegenerative tauopathies, many ofwhich do not show neuropathological evidence of other diagnostic disease-specific lesions. Notably, tau gene mutations have been shown to cause famil-ial frontotemporal dementia and parkinsonism linked to chromosome 17 (FT-

64 ANNALS NEW YORK ACADEMY OF SCIENCES

TABLE 1. “Fatal attractions” of brain proteins: a disease mechanismunderlying diverse neurodegenenerative disorders characterized by filamentousCNS lesions

Disease Lesion/components Location

ADa,b,c SPs/AβNFTs/PHFtau

ExtracellularIntracytoplasmic

ALSa Spheroids/NF subunits, SOD1 Intracytoplasmic

DLBc LBs/α-synuclein Intracytoplasmic

DSa,b,c SPs/AβNFTs/PHFtauLBs/α-synuclein

Extracellular IntracytoplasmicIntracytoplasmic

NBIA 1c LBs/α-synucleinGCIs/α-synuclein

IntracytoplasmicIntracytoplasmic

LBVAD (AD+DLB)c SPs/AβNFTs/PHFtauLBs/α-synuclein

Extracellular Intracytoplasmic Intracytoplasmic

MSAc GCIs/α-synuclein Intracytoplasmic

NIID Inclusions/expanded poly-glutamine tracts

Intranuclear

PDa,c LBs/α-synuclein Intracytoplasmic

Prion diseasesa Amyloid plaques/prions Extracellular

Tauopathiesa.b Tangles/abnormal tau Intracytoplasmic

Trinucleotide repeatdiseases

Inclusions/expanded poly-glutamine tracts

Intranuclear andIntradendritic

NOTE: This table lists hereditary and sporadic neurodegenerative disorders of the CNS char-acterized neuropathologically by prominent filamentous lesions. Most of these lesions arisewithin one or more compartments (i.e., nuclei, cell bodies, processes) of one or more cell typesof the CNS (neurons, astrocytes, oligodendroglia), but some are extracellular deposits of aggre-gated filaments.

ABBREVIATIONS: Aβ = beta-amyloid peptides; AD = Alzheimer’s disease; ALS = amytrophiclateral sclerosis; DLB = dementia with Lewy bodies; DS = Down’s syndrome; GCIs = glial cyto-plasmic inclusions; NBIA 1 = neurodegeneration with brain iron accumulation type 1 (Haller-vorden-Spatz disease); LBs = Lewy bodies; LBVAD = Lewy body variant of Alzheimer’sdisease; MSA = multiple system atrophy; NF = neurofilaments; NFTs = neurofibrillary tangles;NIID = neuronal intranuclear inclusion disease; PD = Parkinson’s disease; PHFtau = paired heli-cal filament tau; SOD1 = superoxide dismutase 1; SPs = senile plaques.

aBoth hereditary and sporadic forms of these disorders are known to occur.bNeurodegenerative diseases with prominent tau pathology are grouped together and referred

to as tauopathies. Well-known examples of tauopathies are: progressive supranuclear palsy,Pick’s disease, corticobasal degeneration, hereditary frontotemporal dementia with parkinsonismlinked to chromosome 17 (FTDP-17) and Guam amyotrophic lateral sclerosis/parkinsonismdementia complex. AD can be included among the tauopathies because of the prominent taupathology in almost all sporadic and familial forms of this disorder, but LBVAD is one of themost common subtypes of sporadic AD and the distinctive lesions in this form of AD are abun-dant cortical α-synuclein–positive LBs, while similar lesions are common in the brains of manypatients with familial AD and DS.

cNeurodegenerative diseases with prominent synuclein pathology are referred to assynucleinopathies.

65TROJANOWSKI & LEE: “FATAL ATTRACTION” OF PROTEINS

DP-17) in many kindreds, and, not surprisingly, some of these kindreds werepreviously thought to have FAD prior to the discovery of tau gene mutationspathogenic for FTDP-17.4 Moreover, while Lewy bodies (LBs) are regardedas hallmark intracytoplasmic neuronal inclusions of Parkinson’s disease(PD), they also occur in the most common subtype of AD, known as the LBvariant of AD (LBVAD), and numerous cortical LBs are the defining brainlesions of dementia with LBs (DLB), which is similar to AD clinically, butdistinct from AD pathologically.2,3 Further, α-synuclein gene mutationscause familial PD, albeit in rare kindreds, and these mutations may be patho-genic by altering the functions or properties of α-synuclein, thereby promot-ing the formation of α-synuclein filaments that aggregate into LBs. However,it is now known that FAD mutations and trisomy 21 lead to abundant accu-mulations of LBs composed of α-synuclein filaments in the brains of mostFAD and elderly Down’s syndrome (DS) patients, respectively, but it is un-clear how these genetic abnormalities promote the formation of LBs fromwild type α-synuclein proteins that are encoded by a normal gene withoutpathogenic mutations. Nonetheless, the accumulation of α-synuclein into fil-amentous inclusions now appears to play a mechanistic role in the pathogen-esis of a number of progressive neurological disorders including PD, DLB,DS, FAD, LBVAD, sporadic AD, multiple system atrophy (MSA), neurode-generation with brain iron accumulation type 1 (NBIA 1) and other synucle-inopathies.2,3,12

Thus, the aggregation of brain proteins into potentially toxic lesions isemerging as a common mechanistic theme in a diverse group of neurodegen-erative diseases that share the following enigmatic symmetry: missense mu-tations in the gene encoding the disease protein cause a familial variant of thedisorder as well as its hallmark brain lesions; but the same brain lesions alsocan be formed by the corresponding wild-type brain protein in sporadic formsof the disease.7,8,12 For this reason, clarification of this enigmatic symmetryin any one of these disorders is likely to have a profound impact on under-standing the mechanisms that underlie all of these disorders as well as on ef-forts to develop novel therapies to treat them. For example, compounds havebeen identified that prevent the conversion of normal proteins into abnormalconformers or variants with conformations that predispose the pathologicalproteins to form potentially toxic filamentous aggregates,13 and it is plausibleto speculate that some of these agents may have therapeutic efficacy in morethan one disorder listed in TABLE 1. Indeed, the Alzheimer Research Forumwebsite (http://www.alzforum.org) lists a number of submitted and/or ap-proved patents on potential AD therapies that target the disruption of filamen-tous lesions in the AD brain (i.e., SPs and NFTs) or are designed to preventthe formation of these filamentous lesions. Moreover, novel therapeutic ap-proaches that utilize peptide building blocks of the abnormal fibrils that formbrain deposits of amyloid in AD as vaccines to prevent or reverse AD

66 ANNALS NEW YORK ACADEMY OF SCIENCES

amyloidosis14 could be extended to treat other of the disorders reviewed here.As the full implications of insights into abnormal protein–protein interactionscoalesce from research advances that emerged at the twilight of the 20th cen-tury,1–12 prospects for the discovery of new and better therapies for neurode-generative dementias of the elderly early in the 21st century will bebrightened and the likelihood will be increased that the next few decades holdthe very real promise for effective treatments for AD and for other devastat-ing neurodegenerative disorders caused by abnormal filamentous aggregatesof different brain proteins.

ACKNOWLEDGMENTS

We thank members of our laboratory and our collaborators within and out-side the University of Pennsylvania for their important contributions to thestudies reviewed here. We also express our appreciation to the families of themany patients studied by our group over the past decade, who have made itpossible to pursue many of the research advances discussed here. The studiessummarized here from our laboratory were supported by grants from the Na-tional Institute on Aging of the National Institutes of Health, the Dana Foun-dation, and the Alzheimer’s Association. Additional information on theneurodegenerative diseases reviewed here also can be obtained by visiting theCenter for Neurodegenerative Disease Research (CNDR) website (http://www.med.upenn.edu/cndr).

REFERENCES

1. ARNOLD, S.E. 2000. Cellular and molecular pathology of the frontal dementias.In Neurodegenerative Dementias: Clinical Features And Pathological Mecha-nisms. C.M. Clark and J.Q.Trojanowski, Eds.: 291–300. McGraw-Hill. NewYork.

2. DICKSON, D.W. 2000. Alzheimer-Parkinson disease overlap: neuropathology.In Neurodegenerative Dementias: Clinical Features And Pathological Mecha-nisms, C.M. Clark and J.Q. Trojanowski, Eds.: 247–260, McGraw-Hill. NewYork.

3. GIASSON, B.I., J.E. GALVIN, V.M.-Y.LEE & J.Q.TROJANOWSKI. 2000. The cel-lular and molecular pathology of Parkinson’s disease. In NeurodegenerativeDementias: Clinical Features And Pathological Mechanisms. C.M. Clark andJ.Q. Trojanowski, Eds.: 219–228. McGraw-Hill. New York.

4. HONG, M., J.Q. TROJANOWSKI & V.M.-Y. LEE. 2000. Tau-based neurofibrillarylesions. In Neurodegenerative Dementias: Clinical Features And PathologicalMechanisms. C.M. Clark and J.Q. Trojanowski, Eds.: 161–175. McGraw-Hill. New York.

67TROJANOWSKI & LEE: “FATAL ATTRACTION” OF PROTEINS

5. IRONSIDE, J.W. 2000. General features of prion diseases. In NeurodegenerativeDementias: Clinical Features And Pathological Mechanisms. C.M. Clark andJ.Q.Trojanowski, Eds.: 329–340, McGraw-Hill. New York.

6. IWATSUBO, T. 2000. Amyloid deposits and plaque formation. In Neurodegener-ative Dementias: Clinical Features And Pathological Mechanisms. C.M.Clark and J.Q. Trojanowski, Eds.: 147–156. McGraw-Hill. New York.

7. LANSBURY, P.T. 1999. Evolution of amyloid: what normal protein folding maytell us about fibrillogenesis and disease. Proc. Natl. Acad. Sci. USA 96:3342–3344.

8. LEE, V.M.-Y., J.Q. TROJANOWSKI, L. BUEE & Y. CHRISTEN, EDS. 2000. FatalAttractions–Protein Aggregates in Neurodegenerative Diseases. Springer.Berlin.

9. MASLIAH, E. & F. LICASTRO. 2000. Neuronal and synaptic loss, reactive glio-sis, microglial response and induction of the complement cascade in Alzhe-imer’s disease. In Neurodegenerative Dementias: Clinical Features AndPathological Mechanisms. C.M. Clark and J.Q. Trojanowski, Eds.: 131–146,McGraw-Hill. New York.

10. PARCHI, P. 2000. Creutzfeldt-Jakob disease and fatal familial insomnia. In Neu-rodegenerative Dementias: Clinical Features And Pathological Mechanisms.C.M. Clark and J.Q. Trojanowski, Eds.: 341–360. McGraw-Hill. New York.

11. PRICE, D.L., P.C. WONG, D.R. BORCHELT, E. KOLIATSOS & J.D. ROTHSTEIN.2000. Motor neuron diseases. In Neurodegenerative Dementias: Clinical Fea-tures And Pathological Mechanisms. C.M. Clark and J.Q. Trojanowski, Eds.:373–388. McGraw-Hill. New York.

12. TROJANOWSKI, J.Q., M. GOEDERT, T. IWATSUBO & V.M.-Y LEE. 1998. Fatalattractions: abnormal protein aggregation and neuron death in Parkinson’sdisease and Lewy body dementia. Cell Death & Differentiation 5: 832–837.

13. PRIOLA, S.A., A. RAINES & W.S. CAUGHEY. 2000. Porphyrin and phthalocya-nine antiscrapie compounds. Science 287: 1503–1506.

14. SCHENK, D., R. BARBOUR, W. DUNN, G. GORDON, H. GRAJEDA, T. GUIDO, K.HU, J. HUANG, K. JOHNSON-WOOD, K. KHAN, D. KHOLODENKO, M. LEE , Z.LIAO, I. LIEBERBURG, R. MOTTER, L. MUTTER, F. SORIANO, G. SHOPP, N.VASQUIEZ, C. VANDEVERT, S. WALKER, M. WOGULIS, T. YEDNOCK, D. GAMES

& P. SLEUBERT. 1999. Immunization with amyloid-β attenuates Alzheimerdisease-like pathology in the PDAPP mouse. Nature 400: 173–177.