1152 Biophysical Journal Volume 106 March 2014 1152–1163

Structural and Dynamic Properties of the Human Prion Protein

Wei Chen,† Marc W. van der Kamp,† and Valerie Daggett†*†Department of Bioengineering, University of Washington, Seattle, Washington

ABSTRACT Prion diseases involve the conformational conversion of the cellular prion protein (PrPC) to its misfolded patho-genic form (PrPSc). To better understand the structural mechanism of this conversion, we performed extensive all-atom,explicit-solvent molecular-dynamics simulations for three structures of the wild-type human PrP (huPrP) at different pH valuesand temperatures. Residue 129 is polymorphic, being either Met or Val. Two of the three structures have Met in position 129 andthe other has Val. Lowering the pH or raising the temperature induced large conformational changes of the C-terminal globulardomain and increased exposure of its hydrophobic core. In some simulations, HA and its preceding S1-HA loop underwent largedisplacements. The C-terminus of HB was unstable and sometimes partially unfolded. Two hydrophobic residues, Phe-198 andMet-134, frequently became exposed to solvent. These conformational changes became more dramatic at lower pH or highertemperature. Furthermore, Tyr-169 and the S2-HB loop, or the X-loop, were different in the starting structures but converged tocommon conformations in the simulations for the Met-129, but not the Val-129, protein. a-Strands and b-strands formed in theinitially unstructured N-terminus. a-Strand propensity in the N-terminus was different between the Met-129 and Val129 proteins,but b-strand propensity was similar. This study reveals detailed structural and dynamic properties of huPrP, providing insight intothe mechanism of the conversion of PrPC to PrPSc.

INTRODUCTION

Prion diseases are a group of fatal neurodegenerative disor-ders, including Creutzfeldt-Jakob disease, Gerstmann-Straussler-Scheinker syndrome, fatal familial insomnia,and kuru in humans; scrapie in sheep; bovine spongiformencephalopathy in cattle; and chronic wasting disease in cer-vids (1,2). According to the protein-only hypothesis, priondiseases are caused by the conformational conversion ofthe cellular prion protein (PrPC) to its misfolded pathogenicform (PrPSc) (3,4). Once formed, PrPSc acts as a template tocatalyze the conversion of PrPC to PrPSc, resulting in accu-mulation of PrPSc. Most cases of prion diseases in humansare sporadic, which means that PrPC spontaneously convertsto PrPSc in vivo or is transmitted in an unidentified manner(1,2).

Structures of PrPC have been solved for multiple species(5–15). These structures are very similar: the N-terminalportion is largely unstructured and the C-terminal regioncontains the folded, globular domain (Fig. 1 A). The C-ter-minal domain consists of three a-helices (HA, HB, and HC)and a short b-sheet with two strands (S1 and S2). PrPSc hasthe same primary structure as PrPC but a different fold (16).During the conversion of PrPC to PrPSc, the a-helix contentdecreases somewhat but the b-sheet content increasesgreatly (17,18). Also, whereas PrPC is protease sensitive,PrPSc contains a protease-resistant core of residues

Submitted July 23, 2013, and accepted for publication December 26, 2013.

*Correspondence: daggett@u.washington.edu

Marc van der Kamp’s present address is: Centre for Computational Chem-

istry, School of Chemistry, University of Bristol, Bristol BS8 1TS, UK.

Wei Chen’s present address is: Department of Pharmaceutical Sciences,

School of Pharmacy, University of Maryland, Baltimore, MD.

Editor: Martin Blackledge.

� 2014 by the Biophysical Society

0006-3495/14/03/1152/12 $2.00

~90–230 (19). An experimentally derived PrPSc structurehas been elusive, but several models of PrPSc have been pro-posed (20–24).

Molecular-dynamics (MD) simulations have been em-ployed to study the conversion of PrPC to PrPSc (25). Forinstance, growth of the native b-sheet of PrPC has beenobserved in MD simulations under acidic conditions(21,26–29). The converted structures were used to build spi-ral models of early aggregates of PrPSc (21,23). MD simula-tions of PrP mutants show that disease-linked pointmutations have varied effects on the structure of PrPC

(30–42). In simulations under harsh nonphysiological con-ditions (e.g., very high temperature and/or very low pH),the globular domain of PrPC unfolds and new b-sheetsform (43–45).

The structure of human PrPC (huPrPC) was first obtainedby NMR (8,46). Most previous MD simulations of huPrPC

were based on these NMR structures (28–30,32,35,36,38–44,47). Later, the structure of huPrPC was also solved byx-ray crystallography (48–50). Although the overall struc-ture of huPrPC in the crystals is similar to that derived byNMR, differences exist in local regions such as the X-loopbetween S2 and HB, which is part of an epitope for aPrPSc-specific monoclonal antibody (51) and hypotheticalprotein X binding partner (52) (Fig. 1). To further investi-gate the structural and dynamic properties of huPrPC, andhow these properties factor into the conversion of PrPC toPrPSc, we performed extensive MD simulations for twonewly solved crystal structures of huPrPC (Protein DataBank (PDB) codes 2W9E and 3HAK) (49,50) along withthe first NMR structure of huPrPC (PDB code 1QLX)(8). In all three cases, we added the missing N-terminal res-idues 90–127 to yield conversion-competent constructs.

http://dx.doi.org/10.1016/j.bpj.2013.12.053

Properties of huPrP 1153

Simulations were carried out at neutral pH (pH 7) or mildlyacidic pH (pH ~5) at 298 and 310 K. pH 5 is similar to thepH of endosomes (53), where the conversion of PrPC toPrPSc occurs (54,55). Simulations at 298 K were performedfor comparison with in vitro experiments, whereas 310 K isphysiological temperature. In the simulations, various partsof the C-terminal globular domain were unstable and under-went conformational changes, which was facilitated atlower pH or higher temperature. The X-loop was differentin the starting structures but converged to a common confor-mation in the simulations of the two structures withMet-129. Furthermore, b-strands and a-strands (variationsof b-strands in which there is a peptide plane flip that resultsin the carbonyl oxygens being aligned on one face and theNH groups on the other (38)) spontaneously formed in theN-terminal tail, a key step in the conversion of PrPC toPrPSc. These results shed light on the mechanism of the con-version of PrPC to PrPSc and show that the results are inde-pendent of the starting structure.

MATERIALS AND METHODS

One NMR structure (PDB code 1QLX) (8) and two crystal structures (PDB

codes 2W9E and 3HAK) (49,50) were used as the starting structures for

simulations of huPrP. Molecules other than huPrP (e.g., waters and/or anti-

body in PDB 2W9E) in these PDB files were excluded. The missing

N-terminal tail (residues 90–127) and C-terminus (residues 229–230) in

the 1QLX structure were modeled as described previously (29). Then, the

same N-terminal tail was added to the two crystal structures. Addition of

these N-terminal residues is necessary to yield a conversion-competent

construct. The missing C-terminus of HC in the two crystal structures

was added using the corresponding portion of the structure of PDB

1QLX. As a result, three structures of huPrP containing residues 90–230

were obtained (Fig. 1 C). All simulations reported here were for proteins

containing residues 90–230, which corresponds to the PrPSc fragments

associated with infectivity and isolated from brains. For simulations at

pH 7, the His residues were protonated on Nεexcept for His-140, which

was protonated on Nd because of its neighboring residues (56). For simula-

tions at pH 5, the His residues were double protonated on both Nd and Nε,

and all titratable side chains were charged.

FIGURE 1 Comparison of three structures (1QLX, 2W9E, and 3HAK) of h

RMSD of the backbone (N, Ca, C, and O atoms) of each residue between ea

(C) Starting structure of huPrP for simulations. The core region of the C-term

218) is highlighted. The hydrophobic core of the globular domain (residues 13

209, 210, and 213–215) is represented as a translucent surface, with every resid

We performed the MD simulations using our in-house-developed molec-

ular modeling package in lucem molecular mechanics (ilmm) (57). The

Levitt et al. (58) protein force field and the F3C water model (59) were

used. A force-shifted nonbonded cutoff of 10 A was used and the

nonbonded list was updated every two steps. The starting structures of

huPrP were first energy minimized by 1000 steps of the steepest-descent

method in vacuo. Then, the proteins were solvated in rectangular water

boxes preequilibrated to the appropriate density of the target temperature

(60). After that, 1000 steps of steepest-descent minimization, 500 steps

of MD, and 500 steps of minimization were performed sequentially for

only water, followed by 500 steps of minimization for only the protein.

Finally, production MD simulations were executed in the microcanonical

ensemble (constant volume, total energy, and number of particles) at pH

7 or 5 at 298 and 310 K. Three simulations were performed for each struc-

ture for each condition. Each simulation was carried out for 100 ns using a

2 fs time step, amounting to a total of 36 simulations and 3.6 ms of simula-

tion time.

We analyzed MD trajectories using ilmm (57) and VMD (61). The Ca

root mean-square deviation (RMSD) of the globular domain (residues

128–228) was measured after alignment on the globular domain. The Ca

RMSD of HA was calculated after alignment on the two disulfide-bonded

helices (HB and HC, residues 175–187 and 206–218, referred to as the

core region of the globular domain) (62). The solvent-accessible surface

area (SASA) was calculated with a probe of 1.4 A radius according to

the Lee and Richards (63) method and the van der Waals radii of Chothia

(64). The Dictionary of Secondary Structure of Proteins (DSSP) algorithm

(65) with additional definitions (23) was used to assign secondary struc-

tures. An atom-atom contact was tallied if the distance between two

nonbonded carbon atoms was %5.4 A or the distance between any other

pair of nonbonded nonhydrogen atoms was %4.6 A.

RESULTS AND DISCUSSION

We selected one NMR structure (PDB code 1QLX) (8) andtwo crystal structures of huPrP (PDB codes 2W9E and3HAK) for this study (49,50). In the 2W9E structure, huPrPis bound to an antibody, which was removed for the simula-tions. A single-nucleotide polymorphism at residue 129 inhumans yields either methionine or valine at that position,the nature of which affects an individual’s susceptibility toacquired prion diseases. The 1QLX and 2W9E structureshave Met-129, whereas the 3HAK structure has Val-129.

uPrP. (A) Structure alignment on the backbones of residues 128–223. (B)

ch pair of structures. Secondary structures are indicated in the bar above.

inal globular domain (portion of HB and HC, residues 175–187 and 206–

4, 137, 139, 141, 158, 161, 175, 176, 179, 180, 184, 198, 203, 205, 206,

ue shown in stick mode. To see this figure in color, go online.

Biophysical Journal 106(5) 1152–1163

1154 Chen et al.

The three huPrP structures are very similar (Fig. 1 A). Inmost regions, the backbone RMSD is <2 A (Fig. 1 B). Ingeneral, the a-helices and b-strands have lower backboneRMSD (<1 A) than the loops (1–2 A). The largest differ-ence is at the X-loop, where the backbone RMSD is as largeas 6 A for some residues. In fact, the X-loop adopts differentconformations in the three structures: it is unstructured inthe 1QLX structure, but contains a short 310 helix in the2W9E and 3HAK structures. The 310 helix in the 2W9Estructure is shorter than in the 3HAK structure, and residues166–170 are oriented differently in the two structures.

The N-terminal residues 90–230 of huPrP are critical forthe conversion of PrPC to PrPSc (19). Therefore, we modeledthe missing N-terminal tail (residues 90–127) as well as theC-terminus of HC in the three structures. As a result, we ob-tained three structures of huPrP containing residues 90–230(Fig. 1 C). Hereafter, we refer to the three structures contain-ing residues 90–230 by the PDB codes for the fragments:1QLX, 2W9E, and 3HAK. We then performed multipleall-atom, explicit-solvent MD simulations for the threestructures at two pH values (pH 7 and 5) and two tempera-tures (298 and 310 K). We observed conformational changesof different regions of the C-terminal globular domain,which became more dramatic at lower pH or higher temper-ature. We also observed the spontaneous formation of a- andb-strands in the N-terminal region, which is a hallmark ofthe conversion of PrPC to PrPSc.

FIGURE 2 Changes of the C-terminal globular domain of huPrP under

different conditions. (A and B) Ca RMSD of the globular domain (A; resi-

dues 128–228) and SASA of the side chains of the hydrophobic core of the

globular domain (B; residues 134, 137, 139, 141, 158, 161, 175, 176, 179,

180, 184, 198, 203, 205, 206, 209, 210, and 213–215) averaged over the

simulations for each starting structure performed under the same

conditions.

Conformational changes in the C-terminalglobular domain

The C-terminal globular domain (residues 128–228) ofhuPrP is fairly stable at pH 7 and 298 K, with an averageCa RMSD of ~2 A (Fig. 2 A). The Ca RMSD of the globulardomain increased at pH 5 compared with pH 7, indicatingthat instability of the globular domain was induced bymildly acidic pH. This observation is consistent with previ-ous experimental and computational studies (29,46,66). Onthe other hand, the Ca RMSD of the globular domain wasalso larger at 310 K than at 298 K, suggesting that increasingtemperature also destabilized the globular domain. Theglobular domain contains a hydrophobic core consisting ofresidues 134, 137, 139, 141, 158, 161, 175, 176, 179, 180,184, 198, 203, 205, 206, 209, 210, and 213–215 (Fig. 1 C)(67). In line with the changes of the Ca RMSD of the glob-ular domain, the SASA of the side chains of the hydropho-bic core increased at pH 5 (Fig. 2 B). Also, the SASA of thehydrophobic core was larger at 310 K than at 298 K, exceptfor 1QLX. Again, lowering the pH or raising the tempera-ture destabilized the globular domain.

Motions of HA and the S1-HA loop

By inspecting the conformational changes in the globulardomain, we noticed that HA and its preceding S1-HA

Biophysical Journal 106(5) 1152–1163

loop (which we have referred to as E4 in previous publica-tions) had large displacements in some simulations, whichwas the major cause of the increase of the Ca RMSD. Inthe starting structures, HA is docked onto the HB-HC sub-domain, and the S1-HA loop preceding HA interacts withHC through hydrophobic interactions between residuesPro-137, Ile-139, and Phe-141 on the S1-HA loop, and res-idues Met-205, Val-209, and Met-213 on HC (Fig. 3 C). Toquantify the movement of HA with respect to the HB-HCsubdomain, we measured its Ca RMSD after alignment tothe core region of the globular domain (residues 175–187and 206–218). By this metric, HA remained near its initialposition with its Ca RMSD of ~2 A in some simulations,whereas it moved dramatically in others (Fig. 3 A). Thelarge movement of HA occurred even at pH 7. For example,in simulation 3 of 1QLX at pH 7 and 310 K, HA movedaway from the HB-HC subdomain (Fig. 3 C). However,the magnitude of the HA displacement was different inthe three structures: the displacement of HA was largest in

FIGURE 3 Movements of HA and the S1-HA loop. (A and B) Ca RMSD of HA (A; residues 144–156) and SASA of the side chains of three hydrophobic

residues (B; Pro-137, Ile-139, and Phe-141) of the S1-HA loop averaged over entire trajectory of each simulation. Structures were aligned on the core region

of the globular domain before RMSD measurements. (C) Displacements of HA and the S1-HA loop in simulation 3 of 1QLX at pH 7 and 310 K. Snapshots

were taken at the indicated times. Residues 137, 139, 141, 205, 209, and 213 are represented in sticks. To see this figure in color, go online.

Properties of huPrP 1155

3HAKand smallest in 2W9E. In some simulations, themove-ment of HA pulled the S1-HA loop away from HC, causingexposure of the S1-HA loop to solvent (Fig. 3 C). However,the exposure of the S1-HA loop was not always associatedwith the HAmovement. For instance, although HA remainedattached to the HB-HC subdomain in simulation 2 of 2W9Eat pH 7 and 298 K, the side chain of Phe141 pointed to sol-vent, resulting in an increase of the SASA of the S1-HAloop without large-scale displacement of the loop (Fig. 3;Fig. S1 in the Supporting Material). Moreover, comparedwith pH 7 or 298 K, the movement of HA and the exposureof the S1-HA loop to solvent increased at pH 5 or 310K, indi-cating that lower pH and higher temperature promoted themovements of HA and the S1-HA loop (Fig. 3, A and B).

We also observed the movements of HA and the S1-HAloop away from HC in our previous MD studies(21,26,28,29). We proposed that exposure and structuringof the S1-HA loop may provide an aggregation site for theformation of PrPSc oligomers, as illustrated in our spiralmodels (21,23). Note that the S1-HA loop becomes anextended strand during conversion, which we call E4 or E5depending on the number of new N-terminal strands thatform (21,23,28). Experimental studies support our findings,as the sequence of the S1-HA loop has a high aggregation

propensity and is indeed an initiation site for aggregation(68). Experiments also support the MD-simulated conver-sion, with the separation of HA and the S1-HA loop fromHC being an essential early step in the conversion of PrPC

to PrPSc (24,69). In addition, a therapeutic antibody that in-hibits prion propagation binds to HA and the N-terminus ofHC of huPrPC, preventing the separation of HA from HC(49). Moreover, a compound that inhibits PrPSc productionlinks the HA-S2 loop to the C-terminus of HB, which alsowould appear to limit the movement of HA (70).

Instability of the C-terminus of HB

The C-terminus of HB was stable in its a-helical conforma-tion in most simulations at pH 7; however, it became dis-torted in some simulations of 2W9E (Fig. 4). Theinstability of the C-terminus of HB was accentuated atpH 5. In one simulation of 1QLX, it unwound and con-verted from a-helix to p-helix (Fig. 4 F). For 2W9E, theC-terminus of HB partially unfolded in one simulationand broke into two a-helices in another simulation(Fig. S2). These results indicate that mildly acidic pH de-stabilizes the C-terminus of HB, in agreement with previ-ous studies (43,71,72).

Biophysical Journal 106(5) 1152–1163

FIGURE 4 Instability of the C-terminus of HB. (A–C) Fraction of the a-helical conformation in time for each residue of the C-terminus of HB of 1QLX

(A), 2W9E (B), and 3HAK (C) over entire trajectory of each simulation. Secondary structures are defined according to the DSSP algorithm. (D and E) SASA

of the side chain of His-187 (D) and the number of atom-atom contacts between the side chains of His-187 and Pro-158/Met-206 (E) averaged over the entire

trajectory of each simulation. (F) Unfolding of the C-terminus of HB in simulation 1 of 1QLX at pH 5 and 310 K. Snapshots were taken at the times indicated.

His-187, Pro-158, and Met-206 are shown as sticks. To see this figure in color, go online.

1156 Chen et al.

His-187 is positioned at the C-terminus of HB. It is buriedand interacts with Pro-158 and Met-206 in the starting struc-tures (Fig. 4 F). In three simulations at pH 5, when theC-terminus of HB was partially unfolded, His-187 becameexposed to solvent (Figs. 4 and S2). The SASA of the sidechain of His-187 increased greatly (Fig. 4 D) (except thatin one simulation, His-187 interacted with the flexible N-ter-minal tail and maintained a lower SASA; Fig. S2 B) and itsinteractions with Pro-158/Met-206 were mostly lost (Fig. 4E). At pH 5, the side chain of His-187 was double proton-

Biophysical Journal 106(5) 1152–1163

ated and charged so as to affect the original packing, even-tually causing it to become exposed, in turn leading to thepartial unfolding of the C-terminus of HB. In support ofthe critical role of His-187, the H187R mutation of huPrPcauses Gerstmann-Straussler-Scheinker syndrome (73,74).

Exposure of Phe-198 and Met-134 to solvent

Phe-198 is located at the HB-HC loop and is part of thehydrophobic core of the globular domain. In the starting

Properties of huPrP 1157

structures, it is surrounded by other residues, includingArg-156, Tyr-157, His-187, Thr-188, Thr-191, Thr-192,Glu-196, Asp-202, Val-203, and Met-206 (Fig. 5 C).When this conformation was stable, the average SASAof the side chain of Phe-198 was <25 A2; however, insome simulations, Phe-198 became exposed to solventand its SASA increased to 50–150 A2 (Fig. 5 A). Forinstance, in simulation 1 of 2W9E at pH 7 and 298 K,the side chain of Phe-198 swung backward and its SASAincreased to 75 A2 (Fig. 5, A and C). These results suggestthat the packing of Phe-198 in the hydrophobic core in thestarting structures is not stable enough to remain in placeeven at pH 7. The exposure of Phe-198 was more frequentat pH 5. In an extreme case, in simulation 3 of 2W9E atpH 5 and 310 K, the side chain of Phe-198 became fullyexposed to solvent with an average SASA of ~140 A2

(Fig. 5 A). These results suggest that mildly acidic pHfurther destabilized the packing of Phe-198 in the hydro-phobic core, which is consistent with our previous MDstudy (29).

Met-134 is positioned at the C-terminus of S1 and isalso part of the hydrophobic core of the globular domain.It is surrounded by Val-161, Met-213, Thr-216, andGln-217 in the starting structures (Fig. 5 D). Similarlyto Phe-198, Met-134 became exposed to solvent insome simulations even at pH 7 (Fig. 5, B and D), which

FIGURE 5 Exposure of Phe-198 and Met-134 to solvent. (A and B) SASA o

trajectory of each simulation. (C) Exposure of Phe-198 to solvent in simulation 1

lation 1 of 2W9E at pH 7 and 310 K. Snapshots were taken at the indicated times

To see this figure in color, go online.

illustrates the unstable packing of Met-134 in the hydro-phobic core. However, the extent of the exposure ofMet-134 was different among the three structures.2W9E had the most frequent exposure of Met-134 andlargest SASA. Furthermore, the exposure of Met-134was more prevalent at pH 5 than at pH 7, indicatingthat mildly acidic pH further destabilized the packing ofMet-134 in the hydrophobic core.

The exposure of Phe-198 and Met-134 to solvent contrib-uted to destabilization of the hydrophobic core of the glob-ular domain, which in turn contributed to partial unfoldingof huPrP and the conversion of PrPC to PrPSc. The exposedhydrophobic side chains of Phe-198 and Met-134 also pro-vide possible aggregation sites. In the spiral model of prionoligomers, Met-134 is buried at the oligomeric interface andinteracts with residues at the N-termini of its own and neigh-boring monomeric units (Fig. S3 A), and thus plays animportant role in the formation of prion oligomers (21).On the other hand, Phe-198 is on the outside of the spiraloligomers and does not contact any residues from othermonomeric units (Fig. S3 B). However, the exposure ofPhe-198 might aid in the aggregation between oligomersto generate larger aggregates. In addition, the exposure ofMet-134 provides opportunities for its oxidation underoxidative conditions, which may facilitate the conversionof PrPC to PrPSc (75).

f the side chains of Phe-198 (A) and Met-134 (B) averaged over the entire

of 2W9E at pH 7 and 298 K. (D) Exposure of Met-134 to solvent in simu-

. Phe-198, Met-134, and their neighboring residues are represented in sticks.

Biophysical Journal 106(5) 1152–1163

1158 Chen et al.

Conformational changes of Tyr-169 and the X-loop

The X-loop adopts different conformations in the starting

structures (Fig. 6 A). Located at the X-loop, Tyr-169 is

initially in different conformations with different levels of

exposure to solvent (the least in 2W9E and the most in

3HAK; Figs. 6 A and S4). In some simulations for 1QLX

FIGURE 6 Conformational changes of Tyr-169 and the X-loop. (A) Starting str

(B) Alignment of the structures of Tyr-169 the X-loop at 100 ns from the simu

ulations in which the conformation of Tyr-169 was converged at 100 ns were s

backbones of residues 161–163 and 172–175. (C) RMSD of the backbone (N

2W9E. RMSD at 100 ns (open circle) was averaged over all pairs of selected sim

structures of 1QLX and 2W9E (open square). (D) Converged conformation of

1QLX at pH 7 and 298 K. Tyr-169 and surrounding residues are shown as sti

Tyr-169 in which it is buried. A representative structure was taken at 100 ns of sim

Biophysical Journal 106(5) 1152–1163

and 2W9E (five and 10 simulations for 1QLX and 2W9E,

respectively), the X-loop converged to a common conforma-

tion (Fig. 6 B). Initially, the backbone RMSD between

1QLX and 2W9E was as large as 5 A for some X-loop res-

idues (Fig. 6 C). At the end of the simulations, the backbone

RMSD decreased substantially (Fig. 6 C). In the converged

conformation of the X-loop, Tyr-169 hydrogen bonded to

uctures of Tyr-169 and the X-loop. Tyr-169 and Asp178 are shown as sticks.

lations of 1QLX to those from the simulations of 2W9E. Five and 10 sim-

elected for 1QLX and 2W9E, respectively. Structures were aligned on the

, Ca, C, and O atoms) of each residue of the X-loop between 1QLX and

ulations as in panel B. The initial RMSD was measured between the crystal

Tyr-169. A representative structure was taken at 100 ns of simulation 2 of

cks. Dashed lines indicate hydrogen bonds. (E) Different conformation of

ulation 1 of 3HAK at pH 7 and 298 K. To see this figure in color, go online.

Properties of huPrP 1159

Asp-178 and interacted with Phe-175 (Fig. 6 D). Thisconformation of Tyr-169 was highly populated and very sta-ble for 2W9E, so that Tyr-169 seldom sampled other con-formations (Fig. S4). A recent experimental studysuggested that a similar conformation of the X-loop inmouse PrP is dominant, with a >90% population (76).Only at pH 5 and 310 K did Tyr-169 adopt other confor-mations and become exposed to solvent, indicating thatmildly acidic pH and the increase in temperature destabi-lized the most populated conformation. In line with thisobservation, pathogenic mutations have been shown toinduce the exposure of Tyr-169 (39,42). In the spiral olig-omer model, Tyr-169 formed multiple contacts with resi-dues from neighboring monomeric units (Fig. S3 C),contributing to the formation of oligomers (77). In contrastto 1QLX and 2W9E, Tyr-169 rarely formed hydrogenbonds with Asp-178 in the simulations of 3HAK(Fig. S4). However, Tyr-169 frequently sampled anotherlargely buried conformation (low SASA), where it inter-acted with Tyr-163, Phe-175, Tyr-218, and Tyr-225(Fig. 6 E).

Formation of a- and b-strands in the N-terminaltail

The N-terminus of huPrP (residues 90–127) was unstruc-tured in the starting structures (Fig. 1 C). However, in thesimulations, we observed the formation of a- andb-strands in the N-terminus, which is a sign of the conver-sion of PrPC to PrPSc (Fig. 7). a-Strands, includinga-bridges, formed mainly in two regions: residues 90–100 and residues 110–127 (Fig. 7, A–C). For example,in simulation 2 of 1QLX at pH 5 and 298 K, an a-sheetcontaining two strands started to form in the first region at~20 ns and was stable through the rest of the simulation(Fig. S5, A and C). The population of a-strand in the firstregion was higher at pH 5 than at pH 7, but the oppositeoccurred in the second region. Compared with the twostructures with Met-129 (1QLX and 2W9E), a-strandseldom formed in the second region for 3HAK, indicatingthe effect of polymorphism at residue 129 (methionine for1QLX and 2W9E, and valine for 3HAK). Consistently, inprevious simulations of huPrP under strongly acidicconditions (pH < 4), a-sheet also formed in residues90–100 (28). a-Sheet has been observed in many simula-tions of amyloidogenic proteins, including PrP, and wehave postulated that it is an intermediate during the for-mation of amyloid and is associated with toxicity(21,28,38,78–84).

On the other hand, b-strands, including b-bridges, pre-dominantly formed over residues 110–127 (Fig. 7, D–F).For instance, in simulation 2 of 3HAK at pH 5 and 310K, a b-strand formed and added to the native S1/S2b-sheet, resulting in a three-stranded b-sheet (Fig. S5, Band D). This result is in agreement with our previous

MD simulations (21,28,29,40). Furthermore, the elongationof the native b-sheet is consistent with the increase of theb-sheet content in the conversion of PrPC to PrPSc

observed by experiments (17,18). b-Strands also formedin residues 90–100, but much more rarely than in residues110–127.

The spontaneous formation of a- and b-strands in theN-terminal tail observed in our simulations supports thatthe idea that the conversion of PrPC to PrPSc can occur spon-taneously, which presumably is a major factor in sporadicprion diseases. It also suggests that the N-terminal tail playsa critical role in the conversion of PrPC to PrPSc. Consis-tently, the spiral model predicts the N-terminus to be a crit-ical aggregation site (21,23). In addition, peptides derivedfrom the N-terminus have highly amyloidogenic propertiesand are able to inhibit the conversion of PrPC to PrPSc

(85,86). Finally, truncation removing the N-terminus ren-ders the protein benign (87), whereas removal of the lasttwo helices does not (88).

CONCLUSIONS

We performed extensive MD simulations for threedifferent structures of the wild-type huPrP. We observedconformational changes in different regions of theC-terminus globular domain in the simulations (Fig. 8).HA and the S1-HA loop underwent large displacementsaway from the HB-HC subdomain, an early step in theconversion of PrPC to PrPSc. The C-terminus of HB wasunstable and unfolded in some simulations. In addition,two residues in the hydrophobic core of the C-terminalglobular domain, Phe-198 and Met-134, were very flexibleand sometimes exposed to solvent. These conformationalchanges were facilitated at lower pH or higher tempera-ture. Furthermore, although they were different in thestarting structures, Tyr-169 and the X-loop converged tocommon, highly populated conformations for the twostructures with Met-129, but not the structure with Val-129. Finally, a- and b-strands formed in the N-terminaltail, a sign of the conversion of PrPC to PrPSc. a-Strandsformed mainly in two regions (residues 90–100 and 110–127) for the two structures with Met-129, but rarely in thesecond region for the structure with Val-129. b-Strandsformed predominantly in residues 110–127 for all threestructures. These results are consistent with many previ-ous experimental and computational studies. They alsosuggest that huPrPC is intrinsically unstable and canconvert spontaneously. The detailed structural and dy-namic properties of huPrP obtained in this study shedfurther light on the mechanism of the conversion ofPrPC to PrPSc. Furthermore, individuals with Met-129are more susceptible to prion diseases than those withVal at that position. These results hint at dynamic struc-tural differences between the polymorphs that may berelevant to disease susceptibility.

Biophysical Journal 106(5) 1152–1163

FIGURE 7 Formation of a- and b-strands in the N-terminal tail. (A–F) Fraction of a-strand (A–C; including a-bridge) or b-strand (D–F; including

b-bridge) in time for each residue of the N-terminus (residues 90–127) of 1QLX (A and D), 2W9E (B and E), and 3HAK (C and F) over all three simulations

(i.e., 300 ns) for each condition. Secondary structures are defined according to the DSSP algorithm. To see this figure in color, go online.

Biophysical Journal 106(5) 1152–1163

1160 Chen et al.

FIGURE 8 Schematic illustrating the major conformational changes

observed during MD mapped onto the (90–230) M129 human PrPC struc-

ture. The six major changes discussed in the text are denoted on the affected

regions of the protein by colors and labels. To see this figure in color, go

online.

Properties of huPrP 1161

SUPPORTING MATERIAL

Five figures are available at http://www.biophysj.org/biophysj/supplemental/

S0006-3495(14)00132-5.

This work was supported by the National Institutes of Health (GM 81407 to

V.D.). Computing resources were provided by the NIH through the Multi-

Tiered Proteomic Compute Cluster (NCRR 1S10RR023044-01) and the

National Energy Research Scientific Computing Center supported by the

Office of Science of the U.S. Department of Energy under contract No.

DE-AC02-05CH11231.

REFERENCES

1. Prusiner, S. B. 1998. Prions. Proc. Natl. Acad. Sci. USA. 95:13363–13383.

2. Aguzzi, A., and A. M. Calella. 2009. Prions: protein aggregation andinfectious diseases. Physiol. Rev. 89:1105–1152.

3. Aguzzi, A., C. Sigurdson, and M. Heikenwaelder. 2008. Molecularmechanisms of prion pathogenesis. Annu. Rev. Pathol. 3:11–40.

4. Cobb, N. J., and W. K. Surewicz. 2009. Prion diseases and theirbiochemical mechanisms. Biochemistry. 48:2574–2585.

5. Riek, R., S. Hornemann, ., K. Wuthrich. 1996. NMR structure of themouse prion protein domain PrP(121-231). Nature. 382:180–182.

6. James, T. L., H. Liu, ., F. E. Cohen. 1997. Solution structure of a142-residue recombinant prion protein corresponding to the infectiousfragment of the scrapie isoform. Proc. Natl. Acad. Sci. USA. 94:10086–10091.

7. Lopez Garcia, F., R. Zahn, ., K. Wuthrich. 2000. NMR structure ofthe bovine prion protein. Proc. Natl. Acad. Sci. USA. 97:8334–8339.

8. Zahn, R., A. Liu,., K. Wuthrich. 2000. NMR solution structure of thehuman prion protein. Proc. Natl. Acad. Sci. USA. 97:145–150.

9. Calzolai, L., D. A. Lysek, ., K. Wuthrich. 2005. Prion protein NMRstructures of chickens, turtles, and frogs. Proc. Natl. Acad. Sci. USA.102:651–655.

10. Gossert, A. D., S. Bonjour,., K. Wuthrich. 2005. Prion protein NMRstructures of elk and of mouse/elk hybrids. Proc. Natl. Acad. Sci. USA.102:646–650.

11. Lysek, D. A., C. Schorn, ., K. Wuthrich. 2005. Prion protein NMRstructures of cats, dogs, pigs, and sheep. Proc. Natl. Acad. Sci. USA.102:640–645.

12. Christen, B., D. R. Perez,., K. Wuthrich. 2008. NMR structure of thebank vole prion protein at 20 degrees C contains a structured loop ofresidues 165–171. J. Mol. Biol. 383:306–312.

13. Christen, B., S. Hornemann, ., K. Wuthrich. 2009. Prion proteinNMR structure from tammar wallaby (Macropus eugenii) shows thatthe b2-a2 loop is modulated by long-range sequence effects. J. Mol.Biol. 389:833–845.

14. Perez, D. R., F. F. Damberger, and K. Wuthrich. 2010. Horse prion pro-tein NMR structure and comparisons with related variants of the mouseprion protein. J. Mol. Biol. 400:121–128.

15. Wen, Y., J. Li,., D. H. Lin. 2010. Unique structural characteristics ofthe rabbit prion protein. J. Biol. Chem. 285:31682–31693.

16. Stahl, N., M. A. Baldwin,., S. B. Prusiner. 1993. Structural studies ofthe scrapie prion protein using mass spectrometry and amino acidsequencing. Biochemistry. 32:1991–2002.

17. Caughey, B. W., A. Dong, ., W. S. Caughey. 1991. Secondary struc-ture analysis of the scrapie-associated protein PrP 27-30 in water byinfrared spectroscopy. Biochemistry. 30:7672–7680.

18. Pan, K. M., M. Baldwin, J. Nguyen, M. Gasset, A. Serban, D. Groth, I.Mehlhorn, Z. Huang, R. J. Fletterick, F. E. Cohen., 1993. Conversionof a-helices into b-sheets features in the formation of the scrapie prionproteins. Proc. Natl. Acad. Sci. USA. 90:10962–10966.

19. Prusiner, S. B., D. F. Groth, ., L. E. Hood. 1984. Purification andstructural studies of a major scrapie prion protein. Cell. 38:127–134.

20. Wille, H., M. D. Michelitsch, ., S. B. Prusiner. 2002. Structuralstudies of the scrapie prion protein by electron crystallography. Proc.Natl. Acad. Sci. USA. 99:3563–3568.

21. DeMarco, M. L., and V. Daggett. 2004. From conversion to aggrega-tion: protofibril formation of the prion protein. Proc. Natl. Acad. Sci.USA. 101:2293–2298.

22. Cobb, N. J., F. D. Sonnichsen, ., W. K. Surewicz. 2007. Moleculararchitecture of human prion protein amyloid: a parallel, in-registerb-structure. Proc. Natl. Acad. Sci. USA. 104:18946–18951.

23. Scouras, A. D., and V. Daggett. 2008. Species variation in PrPSc proto-fibril models. J. Mater. Sci. 43:3625–3637.

24. Hafner-Bratkovic, I., R. Bester,., R. Jerala. 2011. Globular domain ofthe prion protein needs to be unlocked by domain swapping to supportprion protein conversion. J. Biol. Chem. 286:12149–12156.

25. van der Kamp, M. W., and V. Daggett. 2011. Molecular dynamics as anapproach to study prion protein misfolding and the effect of pathogenicmutations. Top. Curr. Chem. 305:169–197.

26. Alonso, D. O. V., S. J. DeArmond, ., V. Daggett. 2001. Mapping theearly steps in the pH-induced conformational conversion of the prionprotein. Proc. Natl. Acad. Sci. USA. 98:2985–2989.

27. Alonso, D. O. V., C. An, and V. Daggett. 2002. Simulations of biomol-ecules: Characterization of the early steps in the pH-induced conforma-tional conversion of the hamster, bovine and human forms of the prionprotein. Philos. Trans. A Math Phys. Eng. Sci. 360:1165–1178.

28. DeMarco, M. L., and V. Daggett. 2007. Molecular mechanism for lowpH triggered misfolding of the human prion protein. Biochemistry.46:3045–3054.

29. van der Kamp, M. W., and V. Daggett. 2010. Influence of pH on thehuman prion protein: insights into the early steps of misfolding.Biophys. J. 99:2289–2298.

30. el-Bastawissy, E., M. H. Knaggs, and I. H. Gilbert. 2001. Moleculardynamics simulations of wild-type and point mutation human prionprotein at normal and elevated temperature. J. Mol. Graph. Model.20:145–154.

31. Gsponer, J., P. Ferrara, and A. Caflisch. 2001. Flexibility of the murineprion protein and its Asp178Asn mutant investigated by molecular dy-namics simulations. J. Mol. Graph. Model. 20:169–182.

32. Okimoto, N., K. Yamanaka, ., T. Hoshino. 2002. Computationalstudies on prion proteins: effect of Ala(117)—>Val mutation.Biophys. J. 82:2746–2757.

Biophysical Journal 106(5) 1152–1163

1162 Chen et al.

33. Santini, S., J. B. Claude, ., P. Derreumaux. 2003. Impact of the tailand mutations G131Vand M129Von prion protein flexibility. Proteins.51:258–265.

34. Barducci, A., R. Chelli,., V. Schettino. 2005. Misfolding pathways ofthe prion protein probed by molecular dynamics simulations.Biophys. J. 88:1334–1343.

35. Shamsir, M. S., and A. R. Dalby. 2005. One gene, two diseases andthree conformations: molecular dynamics simulations of mutants ofhuman prion protein at room temperature and elevated temperatures.Proteins. 59:275–290.

36. Bamdad, K., and H. Naderi-Manesh. 2007. Contribution of a putativesalt bridge and backbone dynamics in the structural instability ofhuman prion protein upon R208H mutation. Biochem. Biophys. Res.Commun. 364:719–724.

37. Chebaro, Y., and P. Derreumaux. 2009. The conversion of helix H2 tob-sheet is accelerated in the monomer and dimer of the prion proteinupon T183A mutation. J. Phys. Chem. B. 113:6942–6948.

38. Chen, W., M. W. van der Kamp, and V. Daggett. 2010. Diverse effectson the native b-sheet of the human prion protein due to disease-associ-ated mutations. Biochemistry. 49:9874–9881.

39. Rossetti, G., G. Giachin, ., P. Carloni. 2010. Structural facets of dis-ease-linked human prion protein mutants: a molecular dynamic study.Proteins. 78:3270–3280.

40. van der Kamp, M. W., and V. Daggett. 2010. Pathogenic mutations inthe hydrophobic core of the human prion protein can promote structuralinstability and misfolding. J. Mol. Biol. 404:732–748.

41. Meli, M., M. Gasset, and G. Colombo. 2011. Dynamic diagnosis offamilial prion diseases supports the b2-a2 loop as a universal interfer-ence target. PLoS ONE. 6:e19093. http://dx.doi.org/10.1371/journal.pone.0019093.

42. Rossetti, G., X. J. Cong,., P. Carloni. 2011. Common structural traitsacross pathogenic mutants of the human prion protein and their impli-cations for familial prion diseases. J. Mol. Biol. 411:700–712.

43. Langella, E., R. Improta, and V. Barone. 2004. Checking the pH-induced conformational transition of prion protein by moleculardynamics simulations: effect of protonation of histidine residues.Biophys. J. 87:3623–3632.

44. Campos, S. R. R., M. Machuqueiro, and A. M. Baptista. 2010.Constant-pH molecular dynamics simulations reveal a b-rich form ofthe human prion protein. J. Phys. Chem. B. 114:12692–12700.

45. Chakroun, N., S. Prigent, ., H. Rezaei. 2010. The oligomerizationproperties of prion protein are restricted to the H2H3 domain.FASEB J. 24:3222–3231.

46. Calzolai, L., and R. Zahn. 2003. Influence of pH on NMR structure andstability of the human prion protein globular domain. J. Biol. Chem.278:35592–35596.

47. Zhang, Y., L. R. Dai, ., Z. C. Ou-Yang. 2006. Molecular dynamicsstudy on the conformational transition of prion induced by the pointmutation: F198S. Thin Solid Films. 499:224–228.

48. Knaus, K. J., M. Morillas, ., V. C. Yee. 2001. Crystal structure of thehuman prion protein reveals a mechanism for oligomerization. Nat.Struct. Biol. 8:770–774.

49. Antonyuk, S. V., C. R. Trevitt, ., J. Collinge. 2009. Crystal structureof human prion protein bound to a therapeutic antibody. Proc. Natl.Acad. Sci. USA. 106:2554–2558.

50. Lee, S., L. Antony, ., V. C. Yee. 2010. Conformational diversity inprion protein variants influences intermolecular b-sheet formation.EMBO J. 29:251–262.

51. Korth, C., B. Stierli,., B. Oesch. 1997. Prion (PrPSc)-specific epitopedefined by a monoclonal antibody. Nature. 390:74–77.

52. Kaneko, K., L. Zulianello, ., S. B. Prusiner. 1997. Evidence for pro-tein X binding to a discontinuous epitope on the cellular prion proteinduring scrapie prion propagation. Proc. Natl. Acad. Sci. USA. 94:10069–10074.

Biophysical Journal 106(5) 1152–1163

53. Lee, R. J., S. Wang, and P. S. Low. 1996. Measurement of endosome pHfollowing folate receptor-mediated endocytosis. Biochim. Biophys.Acta. 1312:237–242.

54. Arnold, J. E., C. Tipler,., R. J. Mayer. 1995. The abnormal isoform ofthe prion protein accumulates in late-endosome-like organelles inscrapie-infected mouse brain. J. Pathol. 176:403–411.

55. Godsave, S. F., H. Wille,., P. J. Peters. 2008. Cryo-immunogold elec-tron microscopy for prions: toward identification of a conversion site.J. Neurosci. 28:12489–12499.

56. Langella, E., R. Improta, ., V. Barone. 2006. Assessing the acid-baseand conformational properties of histidine residues in human prion pro-tein (125-228) by means of pK(a) calculations and molecular dynamicssimulations. Proteins. 64:167–177.

57. Beck, D. A. C., and V. Daggett. 2004. Methods for molecular dynamicssimulations of protein folding/unfolding in solution. Methods. 34:112–120.

58. Levitt, M., M. Hirshberg, ., V. Daggett. 1995. Potential-energy func-tion and parameters for simulations of the molecular-dynamics ofproteins and nucleic-acids in solution. Comput. Phys. Commun. 91:215–231.

59. Levitt, M., M. Hirshberg,., V. Daggett. 1997. Calibration and testingof a water model for simulation of the molecular dynamics of proteinsand nucleic acids in solution. J. Phys. Chem. B. 101:5051–5061.

60. Kell, G. S. 1967. Precise representation of volume properties of waterat 1 atmosphere. J. Chem. Eng. Data. 12:66–69.

61. Humphrey, W., A. Dalke, and K. Schulten. 1996. VMD: visual molec-ular dynamics. J. Mol. Graph. 14:33–38, 27–28.

62. Viles, J. H., D. Donne, ., P. E. Wright. 2001. Local structural plas-ticity of the prion protein. Analysis of NMR relaxation dynamics.Biochemistry. 40:2743–2753.

63. Lee, B., and F. M. Richards. 1971. The interpretation of protein struc-tures: estimation of static accessibility. J. Mol. Biol. 55:379–400.

64. Chothia, C. 1976. The nature of the accessible and buried surfaces inproteins. J. Mol. Biol. 105:1–12.

65. Kabsch, W., and C. Sander. 1983. Dictionary of protein secondarystructure: pattern recognition of hydrogen-bonded and geometrical fea-tures. Biopolymers. 22:2577–2637.

66. Swietnicki, W., R. Petersen, ., W. K. Surewicz. 1997. pH-dependentstability and conformation of the recombinant human prion proteinPrP(90-231). J. Biol. Chem. 272:27517–27520.

67. Riek, R., G. Wider, ., K. Wuthrich. 1998. Prion protein NMR struc-ture and familial human spongiform encephalopathies. Proc. Natl.Acad. Sci. USA. 95:11667–11672.

68. Ziegler, J., C. Viehrig,., S. Schwarzinger. 2006. Putative aggregationinitiation sites in prion protein. FEBS Lett. 580:2033–2040.

69. Eghiaian, F., T. Daubenfeld, ., H. Rezaei. 2007. Diversity in prionprotein oligomerization pathways results from domain expansion as re-vealed by hydrogen/deuterium exchange and disulfide linkage. Proc.Natl. Acad. Sci. USA. 104:7414–7419.

70. Kuwata, K., N. Nishida,., S. Katamine. 2007. Hot spots in prion pro-tein for pathogenic conversion. Proc. Natl. Acad. Sci. USA. 104:11921–11926.

71. Dima, R. I., and D. Thirumalai. 2002. Exploring the propensities of he-lices in PrP(C) to form b sheet using NMR structures and sequencealignments. Biophys. J. 83:1268–1280.

72. Dima, R. I., and D. Thirumalai. 2004. Probing the instabilities in thedynamics of helical fragments from mouse PrPC. Proc. Natl. Acad.Sci. USA. 101:15335–15340.

73. Cervenakova, L., C. Buetefisch,., L. G. Goldfarb. 1999. Novel PRNPsequence variant associated with familial encephalopathy. Am. J. Med.Genet. 88:653–656.

74. Colucci, M., F. J. Moleres, ., P. Gambetti. 2006. Gerstmann-Strauss-ler-Scheinker: a new phenotype with ‘curly’ PrP deposits.J. Neuropathol. Exp. Neurol. 65:642–651.

Properties of huPrP 1163

75. Wolschner, C., A. Giese,., N. Budisa. 2009. Design of anti- and pro-aggregation variants to assess the effects of methionine oxidation in hu-man prion protein. Proc. Natl. Acad. Sci. USA. 106:7756–7761.

76. Damberger, F. F., B. Christen, ., K. Wuthrich. 2011. Cellular prionprotein conformation and function. Proc. Natl. Acad. Sci. USA.108:17308–17313.

77. Armen, R. S., D. O. V. Alonso, and V. Daggett. 2004. Anatomy of anamyloidogenic intermediate: conversion of b-sheet to a-sheet structurein transthyretin at acidic pH. Structure. 12:1847–1863.

78. Armen, R. S., M. L. DeMarco, ., V. Daggett. 2004. Pauling andCorey’s a-pleated sheet structure may define the prefibrillar amyloido-genic intermediate in amyloid disease. Proc. Natl. Acad. Sci. USA.101:11622–11627.

79. Armen, R. S., B. M. Bernard,., V. Daggett. 2005. Characterization ofa possible amyloidogenic precursor in glutamine-repeat neurodegener-ative diseases. Proc. Natl. Acad. Sci. USA. 102:13433–13438.

80. Armen, R. S., and V. Daggett. 2005. Characterization of two distinctb2-microglobulin unfolding intermediates that may lead to amyloidfibrils of different morphology. Biochemistry. 44:16098–16107.

81. Daggett, V. 2006. a-sheet: The toxic conformer in amyloid diseases?Acc. Chem. Res. 39:594–602.

82. Steward, R. E., R. S. Armen, and V. Daggett. 2008. Different disease-causing mutations in transthyretin trigger the same conformationalconversion. Protein Eng. Des. Sel. 21:187–195.

83. Schmidlin, T., B. K. Kennedy, and V. Daggett. 2009. Structural changesto monomeric CuZn superoxide dismutase caused by the familialamyotrophic lateral sclerosis-associated mutation A4V. Biophys. J.97:1709–1718.

84. Tagliavini, F., F. Prelli, L. Verga, G. Giaccone, R. Sarma, P. Gorevic, B.Ghetti, F. Passerini, E. Ghibaudi, G. Forloni., 1993. Synthetic pep-tides homologous to prion protein residues 106–147 form amyloid-like fibrils in vitro. Proc. Natl. Acad. Sci. USA. 90:9678–9682.

85. Chabry, J., B. Caughey, and B. Chesebro. 1998. Specific inhibition ofin vitro formation of protease-resistant prion protein by synthetic pep-tides. J. Biol. Chem. 273:13203–13207.

86. Luhrs, T. T., R. Zahn, and K. Wuthrich. 2006. Amyloid formation byrecombinant full-length prion proteins in phospholipid bicelle solu-tions. J. Mol. Biol. 357:833–841.

87. Watzlawik, J., L. Skora, ., M. L. Kramer. 2006. Prion protein helix1promotes aggregation but is not converted into b-sheet. J. Biol. Chem.281:30242–30250.

88. Scouras, A. D., and V. Daggett. 2012. Disruption of the X-loop turn ofthe prion protein linked to scrapie resistance. Protein Eng. Des. Sel.25:243–249.

Biophysical Journal 106(5) 1152–1163

1

Supporting Material

Structural and Dynamic Properties of the Human Prion Protein

Wei Chen, Marc W. van der Kamp, and Valerie Daggett

Figure S1. Exposure of the S1-HA loop without HA displacement. Snapshot of 2W9E was taken at 100 ns of simulation 2 at pH 7 and 298 K. Residues 137, 139, 141, 205, 209, and 213 are represented in sticks.

Figure S2. Unfolding of the C-terminus of HB. Snapshots of 2W9E at 100 ns of simulation 1 (A) and 3 (B) at pH 5 and 298 K. The C-terminus of HB is in red and His187, Pro158, and Met206 are shown as sticks.

2

Figure S3. Interactions of critical residues in the spiral model of prion oligomers. (A) Met134, (B) Phe198, and (C) Tyr169 are displayed with their surrounding residues. Met134, Phe198, and Tyr169 are shown as spheres. Their surrounding residues are represented in sticks. Different monomeric units are in different colors.

3

Figure S4. Distribution of Tyr169 conformations over all three simulations of each of the three structures (1QLX, 2W9E, and 3HAK) under four conditions: pH 7/298 K, pH 7/310 K, pH 5/298 K, and pH 5/310 K. x axis is the SASA of the side chain of Tyr169 and y axis is the distance between the side chain oxygen (Oη) of Tyr169 and the center of mass of the side chain oxygens (Oδ1-Oδ2) of Asp178. Gray scales represent number of occurrence. Cross symbols indicate the starting points. Solid and open triangles indicate the respective conformations as in Fig. 6, D and E, of the main text.

4

Figure S5. Formation of α- and β-strands in the N-terminal tail. (A and B) DSSP plots of secondary structures of the N-terminus in simulation 2 of 1QLX at pH 5 and 298 K (A) and simulation 2 of 3HAK at pH 5 and 310 K (B). (C and D) Formation of α- and β-strands in simulation 2 of 1QLX at pH 5 and 298 K (A) and simulation 2 of 3HAK at pH 5 and 310 K (B), respectively. Snapshots were taken at indicated times. The N-terminal tail is in red while the globular domain is in gray. Backbones of the formed α- or β-sheets are represented in sticks. Dashed lines indicate hydrogen bonds.