Atomic structures of amyloid cross-b
spines reveal varied steric zippers
Michael R. Sawaya1, Shilpa Sambashivan1, Rebecca Nelson1, Magdalena I. Ivanova1, Stuart A. Sievers1,
Marcin I. Apostol1, Michael J. Thompson1, Melinda Balbirnie1, Jed J. W. Wiltzius1, Heather T. McFarlane1,
Anders Ø. Madsen2,3, Christian Riekel3& David Eisenberg1
Amyloid fibrils formed from different proteins, each associated with a particular disease, contain a common cross-b spine.
The atomic architecture of a spine, from the fibril-forming segment GNNQQNY of the yeast prion protein Sup35, was
recently revealed by X-ray microcrystallography. It is a pair of b-sheets, with the facing side chains of the two sheets
interdigitated in a dry ‘steric zipper’. Here we report some 30 other segments from fibril-forming proteins that form
the PrP prion protein, insulin, islet amyloid polypeptide (IAPP), lysozyme, myoglobin, a-synuclein and b2-microglobulin,
suggesting that common structural features are shared by amyloid diseases at the molecular level. Structures of 13 of these
microcrystals all reveal steric zippers, but with variations that expand the range of atomic architectures for amyloid-like
fibrils and offer an atomic-level hypothesis for the basis of prion strains.
Amyloid diseases are accompanied by the deposition of elongated,
unbranched protein fibrils. For pathologists to designate a disease as
amyloid, the fibrils must be deposited extracellularly, and must bind
Alzheimer’s disease and some 24 others have been found to satisfy
this stringent definition1,2. In addition, many other proteins have
been found to form amyloid-like fibrils with biophysical properties
in common with amyloid fibrils. These properties include an elon-
stituent protein molecules with cooperative, nucleation-dependent
kinetics4, and the so-called cross-b X-ray diffraction pattern5–7. This
pattern consists of an X-ray reflection at ,4.8A˚resolution along the
fibril direction, and another X-ray reflection at ,8–11A˚resolution
perpendicular to the fibril direction5,8,9. The pattern reveals that the
fibrils contain b-sheets parallel to the fibril axis, with their extended
protein strands perpendicular to the axis.
Finding atomic-level structures for cross-b spines has been
impeded by the fibrillar nature of amyloid, but the corner has
been turned. Recent progress has included models for fibrillar seg-
ments constrained by chemical labelling, scanning proline mutagen-
esis, electron paramagnetic resonance, NMR, H/D exchange and
X-ray fibre diffraction data10–21, and two atomic-resolution micro-
crystal structures of the fibril-forming segments GNNQQNY and
NNQQNY22from the yeast prion protein Sup35. Here we extend
atomic-resolution crystallographic studies to other fibril-forming
segments taken from disease-related proteins, and we relate the seg-
ments to the fibrils formed by their parent proteins.
Peptide microcrystals and fibrils
NNQQNY22–24showed that short peptides can themselves form
both fibrils and closely related microcrystals, the latter capable of
revealing atomic structures. From these structures, it was evident
how short segments form fibrils: the cross-b spine consists of a pair
hydrogen-bonding up and down the sheet to identical molecules, all
perpendicular to the axis of the fibril. Two sheets mate tightly at a
completely dry interface. At this interface, the residue side chains
The pair-of-sheets motif, with its dry, steric-zipper interface repeated
along the entire length of the needle-shaped crystal, accounts for the
elongated shape of the crystal and presumably of the fibril. In fibrils
formed from full proteins, the extra-spine regions may remain on the
periphery of the spine, in some cases in native-like conformation25,26.
These initial results encouraged us to identify fibril-forming seg-
ments in other disease-related and fibril-forming proteins, and to
of proteins, we have identified one or more such segments in every
combination of bioinformatic and experimental procedures23,27–29,
guided in some cases by the published work of others. So far we have
identified some 30 such segments from 14 different proteins (see
Supplementary Table 1). Most of these fibril-forming segments also
form needle-shaped microcrystals, varying in length but rarely larger
than 232mm in cross section (Fig. 1). Several control segments,
predicted not to form fibrils, did not form fibrils29. The wide variety
of protein sequences that form fibrils ranges from highly polar (for
example, NNQQ) to highly apolar (for example, MVGGVV), and
from small side chains (for example, SSTSAA) to large (for example,
FYLLYY). Despite their variation, these sequences have a property in
common: their self-complementary binding.
Fibrils related to microcrystals
Although the 13 segment structures are known with high accuracy
(resolutions between 0.85 and 2.0A˚and R-factors between 0.07 and
0.24), there is lingering uncertainty as to how reflective these struc-
tures are of amyloid fibrils. Three types of evidence suggest that the
crystal structures of segments we report here are related to structures
of the fibrils formed by the same segment, and also to fibrils formed
from the entire proteins from which the segments are taken.
Chemistry, University of Copenhagen, Universitetsparken 5, DK-2100 KBH, Denmark.3European Synchrotron Radiation Facility, BP 220, F-38043 Grenoble Cedex, France.
Vol 447|24 May 2007|doi:10.1038/nature05695
fibrils often grow under the same conditions, and are sometimes
found together in solution. Some fibrils (Fig. 1b, c) appear to grow
from tips of microcrystals. In all of the segment crystals, the segment
are the principal diffracting feature. The strands extend perpendic-
ular to the long axis of the crystal and of the fibrils. Also, the inter-
sheet spacings and interstrand spacings of our crystals are in general
accord with the spacings found in amyloid fibrils by X-ray fibre
diffraction7,30. We show this systematically for microcrystals of seg-
ments from Sup35 and amyloid-b protein in Supplementary Fig. 1.
This figure compares X-ray fibril diffraction patterns of the full pro-
by cylindrical averaging of the single microcrystal X-ray data. The
comparison shows a good fit of the principal diffraction features,
suggesting that the principal structural features of the protein fibrils
are closely similar to those within the microcrystals.
Nucleation experiments. Nucleated growth is one of the hallmarks
of amyloid fibrils31, and amyloid nucleation requires equivalence of
the molecular structures of the nucleus and the fibril32. To study
nucleated growth, we prepared crystalline seeds from the insulin
structure of VEALYL. These seeds shorten the characteristic lag time
of the entire insulin molecule (Supplementary Fig. 2), whereas a
non-fibril-forming segment of insulin fails to shorten the lag time.
This is strong evidence for the involvement of LVEALYL in insulin
fibrils. We also found that full-length Sup35 accelerates GNNQQNY
peptide aggregation23. Both studies suggest a structural similarity
between peptide segments and fibrils of their parent proteins.
Mutations in fibril-forming segments affect fibril formation of
whole proteins. If the segments of our structures are in fact in the
amyloid spines of their parent proteins, we would expect that muta-
fact, as detailed in Supplementary Table 2, work of others has estab-
lished that mutants within eight of these segments either slow or
diminish fibril formation. Other studies have found that three of
our segments lie in regions of fibrils that show protection to proton
In summary, diffraction patterns show that the principal diffracting
features of protein fibrils and the corresponding microcrystals are
closely similar; the ability of the protein segments to seed fibrils of
their parent proteins shows that these segments are involved in pro-
tein fibril formation and probably have similar atomic arrangements
in microcrystals and protein fibrils; and the fact that mutant
sequences of the microcrystalline segments affect fibril formation
of their parent proteins suggests that these segments are involved
in protein fibril formation. Although none of these experiments
proves that the structure of the microcrystal is the same as that of
Eight classes of steric zippers
Although varied in sequence, all of our 11 new high-resolution
microcrystal structures (Figs 2, 3; Supplementary Table 3; and Sup-
containing extended protein strands that are perpendicular to the
needle axes and organized into standard Pauling–Corey b-sheets.
Because every one of the structures is built around the steric zipper
that wefound previously in GNNQQNY and NNQQNY22, the struc-
tures suggest that dry, steric-zipper interfaces between b-sheets are a
general principle of protein complementation in amyloid structures.
Other examples of extended protein or peptide chains forming such
steric-zipper interfaces between protein chains are essentially absent
from the PDB (Protein Data Bank) and are rare in the CSD
(Cambridge Structural Database), supporting the idea that these
interfaces are the defining molecular property of the amyloid state.
Despite their fundamental similarity, the reported structures dis-
play variations of the basic steric-zipper structure and thereby
expand our understanding of amyloid structure. The structures to
date fall into five classes (Figs 2, 3), distinguished by (1) whether
their sheets (that is, the strands in their sheets) are parallel or anti-
parallel, (2) whether sheets pack with the same (‘face-to-face’) or
different (‘face-to-back’) surfaces adjacent to one another, and (3)
whether the sheets are oriented parallel (‘up–up’) or antiparallel
(‘up–down’) with respect to one another. To distinguish this third
type of orientation from the first, we refer to the relative sheet
orientations in terms of a given sheet edge facing ‘up’ or ‘down.’
Combinations of these three structural arrangements give eight
theoretically possible classes of steric zippers, shown in Fig. 4.
Examples of classes 1, 2, 4, 7 and 8 are represented in our 13 micro-
crystal structures (Figs 2, 3).
The steric zippers of class 1 share a basic unit of two parallel, in-
register b-sheets with their same sides facing each other (‘face-to-
by a 21axis parallel to the needle axis of the crystal, which rotates a
sheet by 180u about the axis and moves it along the axis by half the
the side chains of one sheet to nestle between layers of side chains on
a simple translation of one sheet onto its neighbour, resulting in a
Microcrystalline clusters Single microcrystals
Figure 1 | Amyloid fibrils and microcrystals. a, Electron micrographs of
representative amyloid-like fibrils (left), and magnified images of
microcrystalline clusters (middle) and single microcrystals mounted for
X-ray diffraction (right) of four segments identified from fibril-forming
and LYQLEN from insulin. b, c, Microcrystals of fibril-forming segments of
amyloid-forming proteins, appearingto have fibrilsgrowing fromtheir tips.
Shown are negatively stained transmission electron micrographs of a
microcrystal of segment NFLVHSS from IAPP (b), and of VEALYL from
insulin (c). In both panels, the microcrystals are on the right.
NATURE|Vol 447|24 May 2007
‘face-to-back, up–up’ packing. SNQNNF, from the human prion
protein, is thus far the only example of a class 2 zipper. In class 4,
each other (‘face-to-back’). However, neighbouring sheets of class 4
are oriented with one sheet’s edge facing ‘up’ and its neighbours’
edges ‘down’. GGVVIA, from the carboxy terminus of amyloid-b,
adopts this orientation. Thesheets of classes 7 and 8,like those of the
‘parallel’ classes 1–4, contain b-strands in register, but within each
sheet, adjacent strands run in opposite directions. Antiparallel
b-sheets in amyloid-like fibrils have been anticipated from previous
Figure 3 | 3D views of representative steric zipper structures of classes 1,
2, 4 and 7, showing the front sheet in silver and the rear sheet in purple.
Oxygen atoms are red; nitrogen atoms are blue. Black lines show
symmetry. The value of the shape complementarity parameter46, SC, for
GGVVIA (SC50.92) is the largest value we have found for any protein
b(1242) than (1–40).
Face = back
Up = down
MVGGVV (forms 1 and 2)
Not observed yet
Not observed yet
GGVVIA, NNQQ (form 1)
Not observed yet
GNNQQNY (forms 1 and 2), NNQQNY,
SSTSAA, VQIVYK, NNQQ (form 2)
Class 2Class 1
Figure 4 | The eight classes of steric zippers. Two identical sheets can be
classified by: the orientation of their faces (either ‘face-to-face’ or ‘face-to-
back’), the orientation of their strands (with both sheets having the same
edge of the strand ‘up’, or one ‘up’ and the other ‘down’), and whether the
strands within the sheets are parallel or antiparallel. Both side views (left)
and yellow arrows show translational symmetry. Below each class are listed
protein segments that belong to that class.
Face = Back
MVGGVV Form 2
MVGGVV Form 1
NNQQ Form 1 Sup35
SNQNNF Prion protein
NNQQ Form 2 Sup35
GNNQQNY Form 1 Sup35
GNNQQNY Form 2 Sup35
Figure 2 | Thirteen atomic-resolution structures for peptide segments of
fibril-forming proteins. See text for details of nomenclature. A two-sheet
showing only the top members of ,105stacked segments in each crystalline
sheet. A dry, steric-zipper interaction is evidenced by the interdigitation of
side chains between sheets. Carbon atoms are shown as purple or white,
nitrogen as blue, and oxygen as red. Water molecules are shown as yellow
spheres. NNQQNY also contains zinc acetate. Zippers are grouped by class
(1, 2, 4, 7, 8); see text for details. Previouslyreported Sup35zippers22belong
are polymorphic pairs (forms 1 and 2; see text for details). The red arrows
point to the 90u bend in the upper sheet of MVGGVV form 2.
NATURE|Vol 447|24 May 2007
X-ray diffraction studies33–36. ‘Face-to-back’ and ‘Face5back’
arrangements (classes 2, 4, 6 and 8) lead naturally to adhesion of
further b-sheets, favouring macroscopic tubes and sheets (for
example, LYQLEN, Fig. 1a), as well as the fibrils usually seen with
the pair-of-sheets motif of class 1 (Fig. 1a). To date, we have not
encountered microcrystals with steric zippers of classes 3, 5 or 6.
Multiple segments and prion strains
In identifying fibril-forming segments within known fibril-forming
proteins (Supplementary Table 1), we found that several proteins
contain more than one fibril-forming segment. Examples are Sup35,
tau, amyloid-b, b2-microglobulin, insulin, IAPP, a-synuclein and
PrP. To date, we have determined structures of different segments
of both insulin and amyloid-b, finding significantly different struc-
tures for segments from the same protein. For example, GGVVIA
from amyloid-b forms parallel sheets, whereas the overlapping seg-
ment MVGGVV forms antiparallel sheets. Thus fibrils formed from
entire proteins could conceivably contain more than a single type of
paired sheets, requiringmore elaboratemodels(see, forexample, refs
37, 38). Alternatively, protein fibrils may contain sheets built from
more than a single type of protein segment. Yet another possibility is
polymorphic fibrils of the same protein, each with its own steric-
will require structural studiesofthesemore complexsheet structures.
Three of the segments shown in Fig. 2 each form two polymorphs,
offering a glimpse of the possible molecular basis of the phenomena
of prion strains and amyloid polymorphism39–45. In prion strains, a
single protein sequence encodes different phenotypes, attributed
recently to ‘distinguishable, self-propagating structural features’44
or different ‘prion conformations’45. A possible molecular explana-
tion for these effects is alternative steric zippers formed by the same
sequence segment. For example, in Fig. 2, there are two polymorphic
structures of NNQQ. In form 2 (class 1), the sheets are ‘face-to-face’,
whereas they are ‘face-to-back’ in form 1 (class 4). We suggest that
fibrils built on these two arrangements would be distinctly different
in structure, and probably different in properties. Notice that the
two polymorphs of GNNQQNY show essentially identical steric zip-
pers, but the packing of the steric zippers within their crystals (Sup-
plementary Fig. 3)differ. This suggests that crystal-packing forces do
not substantially distort the basic structure of the steric zipper,
although it is likely that the flatness of the sheets is influenced by
In several of these crystal structures (as shown in Supplementary
which offers further possibilities for polymorphic structures. This is
in contrast to the structures of GNNQQNY form 1, NNQQNY and
VQIVYK, in which a single type of steric-zipper interaction accounts
for the majority of the buried surface area between sheets (Sup-
plementary Fig. 3). In the cases of MVGGVV form 1, LYQLEN and
SSTSAA, the amount of buried surface area on a given b-sheet face is
split roughly equally between two interfaces. In the example of
NNQQ form 2, there are two different ‘face-to-face’ steric zippers.
together, suggesting that sheets of the full-length protein could pack
together in multiple ways to form polymorphic fibrils.
and the 13 microcrystal structures in Fig. 2 are considered together,
they reinforce the hypothesis that microcrystal structures reveal fun-
damental structural features of amyloid fibrils. First, within each
one segment of 4–12 residues which itself forms fibrils in isolation
from the rest of the protein chain. That every fibril-forming protein
contains a fibril-forming segment suggests that this segment may
drive fibril formation of the protein. Second, as related above, the
fibril-forming peptide segments are linked to fibrils of their parent
the crystal structures reveal essential features of fibril structure,
although there is no definitive proof that the atomic interactions
On the basis of the crystal structures reported here, as well as of
earlier work by others, we summarize our observations of amyloid
fibrils. (1) A 4–7-residue segment of sequence is sufficient to form a
unit of amyloid-like fibrils is a steric zipper, formed by two tightly
metries with multiple steric zippers. That is, the ,30 microcrystal-
line, fibril-forming segments of Supplementary Table 1 and the dry
steric-zipper structures of Fig. 2 together suggest that amyloid dis-
level. (3) On the basis of the discovery of steric zippers in several
disease-associated proteins, it seems likely that the process of fibril
formation starts by the unmasking of zipper-forming segments in
several identical protein molecules, permitting them to stack into
b-sheets and the sheets to interdigitate. Recruitment of monomers
into pre-formed fibrils is expected to be more rapid than nucleation,
because recruitment requires only one molecule at a time to unmask
its fibril-forming sequence, but formation of the steric-zipper nuc-
leus requires several molecules to unmask their zipper-forming
segments simultaneously. That is, the common feature of all these
acteristics of fibril formation31. (4) The finding of distinct crystalline
level hypothesis for amyloid polymorphism and prion strains, which
awaits verification or disproof. (5) There are potentially eight classes
of steric zippers, five of which have been experimentally confirmed.
A full description of methods is given in Supplementary Information. Briefly,
fibrils and microcrystals of the peptides were grown through the dissolution of
lyophilized, synthetic peptide in water or buffers. Crystals were generally grown
by the hanging drop method using standard crystallization screens. X-ray data
were collected at ESRF beamline ID13, processed and scaled using standard
software, and phased by molecular replacement with idealized b-strands.
Thioflavin T fluorescence at 482nm wavelength was used to monitor insulin
fibril formation. Equal volumes of the various seeds were added to identical
reactions, and each experiment was recorded for six replicates.
Received 30 November 2006; accepted 19 February 2007.
Published online 29 April 2007.
1. Westermark, P. Aspects on human amyloid forms and their fibril polypeptides.
FEBS J. 272, 5942–5949 (2005).
Westermark, P. et al. Amyloid: toward terminology clarification. Report from the
Cohen, A. S. & Calkins, E. Electron microscopic observations on a fibrous
component in amyloid of diverse origins. Nature 183, 1202–1203 (1959).
Rochet, J. C., Lansbury, P. T. Jr., Amyloid fibrillogenesis: themes and variations.
Curr. Opin. Struct. Biol. 10, 60–68 (2000).
Astbury, W. T., Dickinson, S. & Bailey, K. The X-ray interpretation of
denaturation and the structure of the seed globulins. Biochem. J. 29, 2351–2360
Geddes, A. J., Parker, K. D., Atkins, E. D. & Beighton, E. ‘‘Cross-b’’ conformation in
proteins. J. Mol. Biol. 32, 343–358 (1968).
Sunde, M. & Blake, C. The structure of amyloid fibrils by electron microscopy and
X-ray diffraction. Adv. Protein Chem. 50, 123–159 (1997).
Eanes, E. D. & Glenner, G. G. X-ray diffraction studies on amyloid filaments.
J. Histochem. Cytochem. 16, 673–677 (1968).
Sunde, M. et al. Common core structure of amyloid fibrils by synchrotron X-ray
diffraction. J. Mol. Biol. 273, 729–739 (1997).
10. Ritter, C. et al. Correlation of structural elements and infectivity of the HET-s
prion. Nature 435, 844–848 (2005).
11.Sikorski, P. & Atkins, E. New model for crystalline polyglutamine assemblies and
their connection with amyloid fibrils. Biomacromolecules 6, 425–432 (2005).
12. Petkova, A. T. et al. A structural model for Alzheimer’s b-amyloid fibrils based on
experimental constraints from solid state NMR. Proc. Natl Acad. Sci. USA 99,
NATURE|Vol 447|24 May 2007
fibril determinedbymagicanglespinningNMRspectroscopy. Proc. NatlAcad. Sci.
USA 101, 711–716 (2004).
14. Krishnan, R. & Lindquist, S. L. Structural insights into a yeast prion illuminate
nucleation and strain diversity. Nature 435, 765–772 (2005).
15. Makin, O. S. & Serpell, L. C. Structures for amyloid fibrils. FEBS J. 272, 5950–5961
16. Lu ¨hrs,T.etal.3DstructureofAlzheimer’samyloid-b(1–42)fibrils.Proc.NatlAcad.
Sci. USA 102, 17342–17347 (2005).
17. To ¨ro ¨k, M. et al. Structural and dynamic features of Alzheimer’s Ab peptide in
amyloid fibrils studied by site-directed spin labeling. J. Biol. Chem. 277,
10–40 of the Alzheimer’s b-amyloid peptide. Biophys. J. 90, 4618–4629 (2006).
19. Williams, A. D. et al. Mapping Ab amyloid fibril secondary structure using
scanning proline mutagenesis. J. Mol. Biol. 335, 833–842 (2004).
20. Kheterpal, I., Zhou, S., Cook, K. D. & Wetzel, R. Ab amyloid fibrils possess a core
structure highly resistant to hydrogen exchange. Proc. Natl Acad. Sci. USA 97,
21. Ferguson, N. et al. General structural motifs of amyloid protofilaments. Proc. Natl
Acad. Sci. USA 103, 16248–16253 (2006).
22. Nelson, R. et al. Structure of the cross-b spine of amyloid-like fibrils. Nature 435,
23. Balbirnie, M., Grothe, R. & Eisenberg, D. S. An amyloid-forming peptide from the
yeast prion Sup35 reveals a dehydrated b-sheet structure for amyloid. Proc. Natl
Acad. Sci. USA 98, 2375–2380 (2001).
24. Diaz-Avalos, R. et al. Cross-b order and diversity in nanocrystals of an amyloid-
forming peptide. J. Mol. Biol. 330, 1165–1175 (2003).
a central core fiber. J. Biol. Chem. 278, 43717–43727 (2003).
26. Sambashivan, S., Liu, Y., Sawaya, M. R., Gingery, M. & Eisenberg, D. Amyloid-like
fibrils of ribonuclease A with three-dimensional domain-swapped and native-like
structure. Nature 437, 266–269 (2005).
27. Thompson, M. J. et al. The 3D profile method for identifying fibril-forming
segments of proteins. Proc. Natl Acad. Sci. USA 103, 4074–4078 (2006).
28. Ivanova, M. I., Sawaya, M. R., Gingery, M., Attinger, A. & Eisenberg, D. An
amyloid-forming segment of b2-microglobulin suggests a molecular model for
the fibril. Proc. Natl Acad. Sci. USA 101, 10584–10589 (2004).
29. Ivanova, M. I., Thompson, M. J. & Eisenberg, D. A systematic screen of b2-
30. Fa ¨ndrich,M.&Dobson,C.M.Thebehaviourofpolyaminoacidsrevealsaninverse
sidechain effect inamyloidstructure formation. EMBO J.21,5682–5690 (2002).
and scrapie: mechanistic truths and physiological consequences of the time-
dependent solubility of amyloid proteins. Annu. Rev. Biochem. 66, 385–407
32. Jarrett, J. T., Lansbury, P. T. Jr., Amyloid fibril formation requires a chemically
discriminating nucleation event: studies of an amyloidogenic sequence from the
bacterial protein OsmB. Biochemistry 31, 12345–12352 (1992).
33. Sikorski, P., Atkins, E. D. & Serpell, L. C. Structure and texture of fibrous crystals
formed by Alzheimer’s Ab(11–25) peptide fragment. Structure (Camb.) 11,
34. Makin, O. S., Atkins, E., Sikorski, P., Johansson, J. & Serpell, L. C. Molecular basis
for amyloid fibril formation and stability. Proc. Natl Acad. Sci. USA 102, 315–320
35. Halverson, K., Fraser, P. E., Kirschner, D. A., Lansbury, P. T. Jr., Molecular
determinants of amyloid deposition in Alzheimer’s disease: conformational
studies of synthetic b-protein fragments. Biochemistry 29, 2639–2644 (1990).
36. Serpell, L. C. & Smith, J. M. Direct visualisation of the b-sheet structure of
synthetic Alzheimer’s amyloid. J. Mol. Biol. 299, 225–231 (2000).
37. Kajava, A. V., Baxa, U., Wickner, R. B. & Steven, A. C. A model for Ure2p prion
filaments and other amyloids: the parallel superpleated b-structure. Proc. Natl
Acad. Sci. USA 101, 7885–7890 (2004).
38. Kajava, A. V., Aebi, U. & Steven, A. C. The parallel superpleated b-structure as a
model for amyloid fibrils of human amylin. J. Mol. Biol. 348, 247–252 (2005).
39. King, C. Y. & Diaz-Avalos, R. Protein-only transmission of three yeast prion
strains. Nature 428, 319–323 (2004).
40. Tanaka, M., Chien, P., Naber, N., Cooke, R. & Weissman, J. S. Conformational
variations in an infectious protein determine prion strain differences. Nature 428,
41. Petkova, A. T. et al. Self-propagating, molecular-level polymorphism in
Alzheimer’s b-amyloid fibrils. Science 307, 262–265 (2005).
42. Chien, P. & Weissman, J. S. Conformational diversity in a yeast prion dictates its
seeding specificity. Nature 410, 223–227 (2001).
43. Jones, E. M. & Surewicz, W. K. Fibril conformation as the basis of species- and
strain-dependent seeding specificity of mammalian prion amyloids. Cell 121,
44. Diaz-Avalos, R., King, C. Y., Wall, J., Simon, M. & Caspar, D. L. Strain-specific
morphologies of yeast prion amyloid fibrils. Proc. Natl Acad. Sci. USA 102,
45. Tanaka, M., Collins, S. R., Toyama, B. H. & Weissman, J. S. The physical basis of
how prion conformations determine strain phenotypes. Nature 442, 585–589
46. Lawrence, M. C. & Colman, P. M. Shape complementarity at protein/protein
interfaces. J. Mol. Biol. 234, 946–950 (1993).
Supplementary Information is linked to the online version of the paper at
M. Graf and K. Wu ¨thrich for discussions, and the NSF, the NIH and the HHMI for
National Research Service Award.
Author Contributions M.R.S., S.S., R.N. and M.I.I contributed equally to this work.
Author Information The 11 new structures shown in Fig. 2, and their structure
factors, have been deposited in the Protein Data Bank with accession codes as
follows: GNNQQNY form 2, 2OMM; NNQQ form 1, 2ONX; NNQQ form 2, 2OLX;
VEALYL, 2OMQ; LYQLEN, 2OMP; VQIVYK, 2ON9; GGVVIA, 2ONV; MVGGVV
form 1, 2ONA; MVGGVV form 2, 2OKZ; SSTSAA, 2ONW; SNQNNF, 2OL9. In
addition, at http://www.doe-mbi.ucla.edu/,sawaya/chime/xtalpept/ we offer,
for each microcrystal structure, coordinates of the asymmetric unit, the unit cell,
the steric zipper, polar zippers, crystal packing and water exclusion. Reprints and
permissions information is available at www.nature.com/reprints. The authors
declare no competing financial interests. Correspondence and requests for
materials should be addressed to D.E. (firstname.lastname@example.org).
NATURE|Vol 447|24 May 2007