STRUCTURAL BIOLOGY

Near-atomic model ofmicrotubule-tau interactionsElizabeth H. Kellogg1,2*, Nisreen M. A. Hejab2*, Simon Poepsel1, Kenneth H. Downing2,Frank DiMaio3,4, Eva Nogales1,2,5†

Tau is a developmentally regulated axonal protein that stabilizes and bundles microtubules(MTs). Its hyperphosphorylation is thought to cause detachment from MTs and subsequentaggregation into fibrils implicated in Alzheimer’s disease. It is unclear which tau residuesare crucial for tau-MT interactions,where tau binds onMTs, and how it stabilizes them.We usedcryo–electron microscopy to visualize different tau constructs on MTs and computationalapproaches to generate atomic models of tau-tubulin interactions.The conservedtubulin-binding repeats within tau adopt similar extended structures along the crest of theprotofilament, stabilizing the interfacebetween tubulin dimers.Our structures explain the effectof phosphorylation on MTaffinity and lead to a model of tau repeats binding in tandem alongprotofilaments, tethering together tubulin dimers and stabilizing polymerization interfaces.

Microtubules (MTs) are formed by the as-sembly of ab-tubulin dimers into proto-filaments (PFs) that associate laterallyinto hollow tubes. MTs are regulated byMT-associated proteins (MAPs), includ-

ing “classical”MAPs such as MAP-2, MAP-4, andtau that are critical to neuronal growth andfunction. Tau constitutes more than 80% of neu-ronal MAPs, stabilizes and bundles axonal MTs(1), and is developmentally regulated (2). Full-length adult tau is intrinsically disordered andincludes a projection domain, anMT-binding re-gion of four imperfect sequence repeats (R1 to R4),and a C-terminal domain (3) (Fig. 1A; the precisedefinition of the repeats has not always been con-sistent in the literature and the one displayed inthe figure is justified by the structural findings inthis study). Different repeats bind to and stabilizeMTs (4, 5), with affinity and activity increasingwiththe number of repeats (5, 6). Neurodegenerativetauopathies, includingAlzheimer's disease, devel-opwhenmutated (7,8) or abnormally phosphoryl-ated (9–11) tau loses affinity for MTs and formsfilamentous aggregates called neurofibrillary tan-gles.Whereas we know the structure of amyloidtau fibrils (12), the physiological conformationof MT-bound tau remains controversial (13–17).Here we present atomic models of MT-boundtau by using a combination of single-particlecryo–electronmicroscopy (cryo-EM) and Rosettamodeling.We used cryo-EM to visualize MTs in the pres-

ence of an excess of different tau constructs (figs.S1 and S2). The cryo-EM structure of dynamic

MTs (without stabilizing drugs or nonhydrolyzableguanosine 5′-triphosphate analogs) assembledwith full-length tau (overall resolution of 4.1 Å)(fig. S2C) shows tau as a narrow, discontinuousdensity along each PF (Fig. 1B), following theridge on theMT surface defined by the H11 andH12 helices of a- and b-tubulin and adjacent tothe site of attachment of the C-terminal tubulintails (Fig. 1C) that are important for tau affinity(18, 19). This location is consistentwith a previous,low-resolution cryo-EM study (16). To test an al-ternatively proposed tau-binding site on the MTinterior (17), we added tau to preformed MTs orto polymerizing tubulin, both in the absence ofTaxol, but never saw tau density on the MT lu-minal surface.We also examined two N- and C-terminally

truncated tau constructs including either all fourrepeats (4R) or just the first two (2R) (fig. S2, Aand B) (we refer to the construct containing fourrepeats as 4R,whereas the sequence of the fourthrepeat is referred to as R4). Both reconstructions(4.8 and 5.6 Å, respectively) (fig. S2C) and thatof full-length tau were indistinguishable at theresolutions obtained. In all three cases, the lengthof the tau density corresponds to an extendedchain of ~27 amino acids, with a weak connectingdensity that would accommodate another 3to 4 residues, adding up to the length of onerepeat (31 to 32 amino acids).The lower resolution for tau (4.6 to 6.5 Å) than

for tubulin (4.0 to 4.5 Å) in our reconstructions(fig. S3)may be due to substoichiometric binding(unlikely because of the excess of tau), flexibility,and/or differences between the repeats. Thus, wepursued the structure of a synthetic tau constructwith four identical copies of R1 (R1×4) (Fig. 2)and reached an overall resolution of 3.2 Å (Fig. 2Aand fig. S2C), with local resolution for taubetween 3.7 and 4.2 Å (fig. S3). In this mapthe best-resolved region of tau approached theresolution of surface regions of tubulin. Again,tau appears as regularly spaced segments sep-arated by more discontinuous density, as ex-

pected from the alternation of more tightlybound segments interspersed withmoremobileregions (5, 20). A polyalanine model accom-modating 12 residues was built into the best-resolved segment of the R1×4 tau density atthe interdimer interface.Given the lack of large side chains or second-

ary structure, we used Rosetta (21) as an unbiasedapproach to assess different tau repeat sequenceregisters (see methods in supplementary mate-rials). A single register and conformation werefavored, both energetically and on the basis of fitto density, and included amino acids 256 to 267(VKSKIGSTENLK), the most conserved segmentamong tau repeats (Fig. 2B). An alternative,moreaggressive refinement protocol (22), which intro-ducesmore structural variability during the refine-ment procedure, converged to the same sequenceand structure (fig. S4). This common solutioncorresponds to a 12-residue sequence containedwithin an 18–amino acid fragment that is suffi-cient to promote MT polymerization (4, 23). Fur-thermore, the inferred interactions are consistentwith sequence conservation among classicalMAPsbeyond tau (fig. S5), with conserved residuescontributing critical tubulin interactions withinour model (Fig. 2C). Ser258 and Ser262 form hy-drogen bonds with a-tubulin Glu434. Phospho-rylation of the universally conserved Ser262 (fig. S5)strongly attenuates MT binding (24) and is amarker of Alzheimer’s disease (25). Our structurenow explains how its phosphorylation disruptstau-tubulin interactions. Although Thr263 is alsopositioned to hydrogen-bond with Glu434, hydro-phobic substitutions are tolerated at this position(fig. S5), indicating that this interaction may notbe as essential. The conserved Lys259 is positionedto interact with an acidic patch on a-tubulinformed by Asp424, Glu420, and Glu423 (Fig. 2C).Ile260, conserved in hydrophobic character acrossall R1 sequences (fig. S5), is buried within a hy-drophobic pocket formed by a-tubulin residuesIle265, Val435, and Tyr262 at the interdimer inter-face (Fig. 2C). Asn265, universally conserved amongrepeats of classical MAPs (fig. S5), forms a stabi-lizing intramolecular hydrogen bond within thetype II′ b turn formed by residues 263 to 266.Lastly, Lys267 is positioned to interact with theacidic a-tubulin C-terminal tail, and its basic char-acter is conserved (Fig. 2C and fig. S5).The tau density beyond this 12-residue stretch

is very weak (Fig. 2A), indicating that residues242 to 255 lack ordered interactions with tubulin.This region of R1 is rich in prolines, but the cor-responding regions of R2 and R3 show a distinct,conserved hydrophobic pattern (Fig. 1A and fig.S5). As other tau repeats might form additionalinteractionswith theMT surface, we also obtaineda cryo-EM reconstruction by using a synthetic tauconstruct containing four copies of repeat R2(R2×4). TheR2×4-MT reconstruction, at an overallresolution of 3.9 Å (Fig. 3A and fig. S2C), was sim-ilar to R1×4, especially at the interdimer tubulininterface (Fig. 3A and fig. S6A), but had additionaltau density along the surface of b-tubulin. Wecould model a backbone stretch of 27 residuesinto the R2×4 tau density, spanning three tubulin

RESEARCH

Kellogg et al., Science 360, 1242–1246 (2018) 16 June 2018 1 of 4

1QB3 Institute and Department of Molecular and Cell Biology,University of California–Berkeley, Berkeley, CA 94720, USA.2Division of Molecular Biophysics and Integrated Bioimaging,Lawrence Berkeley National Laboratory, Berkeley, CA 94720,USA. 3Department of Biochemistry, University of Washington,Seattle, WA 98195, USA. 4Institute for Protein Design,Seattle, WA 98195, USA. 5Howard Hughes Medical Institute,University of California–Berkeley, Berkeley, CA 94720, USA.*These authors contributed equally to this work.†Corresponding author. Email: enogales@lbl.gov

on October 20, 2020

http://science.sciencem

ag.org/D

ownloaded from

monomers, with a length close to that of the tu-bulin dimer repeat in the MT lattice (~80 Å). Aswe found for R1, we discovered a single, stronglypreferred register and conformation for R2 withthe use of Rosetta (Fig. 3B), regardless of the cho-sen simulation parameters. Our analyses of R1 andR2 resulted in equivalent sequence registers andvirtually identical atomicmodels at the interdimerinterface (Fig. 3, C andD), with conserved residuesmaking critical contactswith tubulin. For two non-

conserved positions, Cys291(R2) versus Ile260(R1)and Lys294(R2) versus Thr263(R1) (Fig. 3C), thenature of the interactions is preserved [free cys-teines demonstrate strong hydrophobic character(26), and Lys294 likely interacts with the acidicC-terminal tail (Fig. 3, E and G)]. The identicalsequence register and atomic details from twoindependent maps underscore the robustness ofour solution and provide high confidence in theaccuracy of our atomic models. The sequence

assignment is further supported by previousgold-labeling experiments on the binding ofMAP-2 to MTs (fig. S6, B to D).OurR2×4model includes the peptide VQIINKK

(not conserved inR1), which as an isolated peptidecan bind MTs (6) and promote MT polymeri-zation (23). This R2 peptide localizes to theintradimer interface and is sufficiently close tointeract with the b-tubulin C-terminal tail (Fig. 3,E and F). Though tau and kinesin make distinct

Kellogg et al., Science 360, 1242–1246 (2018) 16 June 2018 2 of 4

A CB Full-length tau

R1R2R3R4

242274305336

273304335367

1 441100 200 300 400full-length tau

N-terminal proline-rich repeat domains

C-terminal

projection domainmicrotubule-binding

domains

intr

a-d

imer

inte

r-d

imer

tub

ulin

dim

er

*

*

*

*

R L Q T A P V PMP D L K - NV K S K I G S T E N L K HQ PGGGK VQ I I NK - K L D L S - NVQ S K CG S K DN I K HV PGGGS VQ I V Y K - P V D L S - K V T S K CG S L G N I HHK PGGGQ V E V K S E - K L DF K DRVQ S K I G S L DN I T HV PGGG

*

R1R1 R2R2 R3R3 R4R4R1 R2 R3 R4

Fig. 1. Tau binding to microtubules. (A) Schematic of tau domainarchitecture and assigned functions.The MT-binding domain of fourrepeats is defined as residues 242 to 367.The inset shows the sequencealignment of the four repeat sequences, R1 to R4, that make up therepeat domain. Ser262 is marked by the asterisk. Single-letterabbreviations for the amino acid residues are as follows: A, Ala;C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met;N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T,Thr; V, Val; W,Trp; and Y,Tyr.(B) Cryo-EM density map (4.1-Å overall resolution) of an MTdecoratedwith full-length tau.Tau (red) appears as a nearly continuous stretchof density along PFs (a-tubulin in green, b-tubulin in blue). The boxedregionmarks a tau repeat and is shown inmore detail in (C). (C) The footprint of a continuous stretch of tau spans over three tubulinmonomers, binding acrossboth intra- and interdimer tubulin interfaces (only one repeat of tau is shown for clarity).The positions of the C termini of tubulin are indicated with asterisks.

C

N

modeled

synthetic R1x4 tau conservation

100%0%

R1x4 tau197 400

R1 R1 R1 R1

R LQ T AP V PMP D L KNV KS K I GS T E N L KHQPGGG

E434

K402

F399

VKSKIGSTENLK

-tub tail

S258

S262

N265

I260

K267

**

D424

**

*

*

*

*

**

A B C

**

Fig. 2. Near-atomic-resolution reconstruction of synthetic R1×4 tau on microtubules. (A) In the 3.2-Åcryo-EM reconstruction of tau-bound MTs, the best-ordered segment of tau is bound at the interfacebetween tubulin dimers (boxed) (lower threshold density for tau is shown in transparency). (B) Rosettamodeling reveals a single energetically preferred sequence register (circled) for the best-ordered tau region,corresponding to a conserved (underlined) 12-residue stretch of residues within the R1 repeat sequence.All Rosetta simulations were repeated until convergence (100 models per register). The intensity of thecolored boxes indicates the extent of conservation among tau homologs. Ser262 is indicated with a redasterisk in (B) and (C). (C) Atomic model of tau and tubulin, with (left) and without (right) the density map,showing the interactions over the interdimer region. The C termini of tubulin are indicated with yellowasterisks in (A) and (C). a-tub, a-tubulin.

RESEARCH | REPORTon O

ctober 20, 2020

http://science.sciencemag.org/

Dow

nloaded from

contacts with tubulin, their MT-binding sitespartially overlap (fig. S7), which explains why taubinding interferes with kinesin attachment toMTs (27). Residues Val275, Ile277, Leu282, andLeu284 (Fig. 3F) are buried against the MT sur-face and tolerate only conservative hydrophobicsubstitutions in R2, R3, and R4 (fig. S8). Lys274

and Lys281 are crucial for tau-MT binding (6). Inour model, Lys274 is close to an acidic patch onthe MT formed by Asp427 and Ser423 in b-tubulin(Fig. 3, E and G), and Lys281 is well positioned tointeract with the b-tubulin C-terminal tail (Fig.3E). The highly conserved His299 in R2 is buriedin a cleft formed by b-tubulin residues Phe395

and Phe399 (Fig. 3, C and E).Although we could model most of the residues

in a tau repeat, we could not clearly visualize thehighly conserved PGGG motif, which would cor-respond to the region connecting the modeledsegments. Figure S9 shows a model of consecu-tive R1 andR2 binding, on the basis of the atomicmodels of both repeats connected by an extendedPGGG segment, placed into the full-length tau

experimentalmap. The similarity among repeats,especially R2 and R3, and the geometry of theirbinding to tubulin strongly support a tandembinding mode of the four repeats along a PF (fig.S8 shows the conserved character of hydropho-bic interactions between these tau repeats andtubulin). Comparison of our MT-bound tau struc-tures for R1 and R2 to that of fibrillary tau for R3and R4 shows that though they are globally verydifferent, they share some similarities in localstructure, especially in the conserved, hydropho-bic regions (fig. S10).OurMT-tau structures lead to amodel inwhich

each tau repeat has an extended conformationthat spans both intra- and interdimer interfaces,centered on a-tubulin and connecting threetubulin monomers. Extensive modeling on twoindependently determined reconstructions com-plemented the experimental map and led to aproposed model of the repeat structure and MTinteractions that is supported by sequence con-servation and explains previous biochemicalobservations. The universally conserved Ser262, a

major site of phosphorylation, is critically in-volved in tight contacts with tubulin near apolymerization interface, explaining how its mod-ification interferes with tubulin binding and MTstabilization. The major tau-binding site corre-sponds to the “anchor point” (28), a tubulinregion that is practically unaltered during thestructural changes accompanying nucleotide hy-drolysis or in a comparison of assembled anddisassembled states of tubulin (Fig. 4, right).This finding explains how tau promotes the for-mation of tubulin rings (29) and copurifies withMTs through tubulin assembly-disassembly cycles.Our structures suggest that all four tau repeatsare likely to associate with the MT surface intandem, through adjacent tubulin subunits alonga PF. This modular structure explains howalternatively spliced variants can have essentiallyidentical interactions with tubulin but differ-ent affinities according to the number of repeatspresent. The tandem binding of tau along a PFexplains how tau promotes both MT polymer-ization and stabilization by tethering multiple

Kellogg et al., Science 360, 1242–1246 (2018) 16 June 2018 3 of 4

C

10°

K294

C291

H299

R2x4 tau197 400

R2R2 R2R2 R2R2 R2R2R2 R2 R2 R2

KVQ I I NKK L D L S NVQS KCGS KDN I KHV PGGG

B

modeled

conservation

100%0%synthetic R2x4

tau

A

*

*

*

*

*

*

*

*

R2R2VQSKCGSKDNIKHV

R1R1VKSKIGSTENLK

**-tub tailR

2x

4 m

od

ele

d

45°

D

R1x4 m

od

eled

KVQII

NKKLDLSNVQSKCGSKDNI

KHV

Q276

K274

K281

H299

**

**

S423

D427

F399

F395

mo

del

ed R

1 le

ng

th

90°

90°

K294

K298 **

V275

I277

L282

L284

E F

G

Fig. 3. High-resolution reconstruction of synthetic R2×4 tau onmicrotubules. (A) The (3.9-Å) cryo-EM reconstruction of R2×4 is highlysimilar to that of R1×4 but reveals a longer stretch of ordered density fortau along the MT surface (lower threshold density for tau is shown intransparency). The boxed region marks the footprint of one R2 repeat.(B) Rosetta modeling supports a sequence register (circled) for theR2 sequence binding to tubulin equivalent to that for R1 shown inFig. 2. All Rosetta simulations were repeated until convergence(100 models per register). (C) Major tau-tubulin interactions at theinterdimer cleft are highly similar between the R1 (shown in orange foreasier visualization) and R2 (purple) sequences. (D) Extending the modelto account for the additional density reveals an almost entire repeat

of tau, spanning three tubulin monomers (centered on a-tubulin andcontacting b-tubulin on either side), with an overall length of ~80 Å(the approximate length of a tubulin dimer). (E) Atomic model ofthe R2 repeat, shown along with the corresponding R2 sequence. Theposition of the previously studied gold-labeled residue in MAP-2 at theequivalent position according to homology (16) is indicated with anarrowhead and is in very good agreement with our model (see also fig. S6).Boxed-out regions show (F) the hydrophobic packing of tau residues onthe MT surface (see also fig. S8) and (G) the positioning of two R2 lysineswith potential interactions with the a-tubulin acidic C-terminal tail.The C termini of tubulin are indicated with yellow asterisks in (A), (C),(E), and (G).

RESEARCH | REPORTon O

ctober 20, 2020

http://science.sciencemag.org/

Dow

nloaded from

tubulin dimers together across longitudinal in-terfaces (Fig. 4, left). This mode of interactionexplains the reduced off-rate of tubulin dimers inthe presence of tau (30) that results in reducedcatastrophe frequencies (transitions from assem-bly to disassembly of MTs) and the loss of dy-namicity caused by tau and other classicalMAPs.Our study does not discard the possibility thatother structural elementswithin tau are involvedin additional tubulin interactions, especially ifengaging with the unstructured C-terminal tubu-lin tails. Such potential contactsmay contribute tointer-PF and/or inter-MT interactions.

REFERENCES AND NOTES

1. C. Conde, A. Cáceres, Nat. Rev. Neurosci. 10, 319–332 (2009).2. K. S. Kosik, L. D. Orecchio, S. Bakalis, R. L. Neve, Neuron 2,

1389–1397 (1989).

3. A. Himmler, D. Drechsel, M. W. Kirschner, D. W. Martin Jr., Mol.Cell. Biol. 9, 1381–1388 (1989).

4. D. J. Ennulat, R. K. Liem, G. A. Hashim, M. L. Shelanski, J. Biol.Chem. 264, 5327–5330 (1989).

5. K. A. Butner, M. W. Kirschner, J. Cell Biol. 115, 717–730(1991).

6. B. L. Goode, S. C. Feinstein, J. Cell Biol. 124, 769–782(1994).

7. M. Hutton et al., Nature 393, 702–705 (1998).8. M. Hong et al., Science 282, 1914–1917 (1998).9. A. C. Alonso, T. Zaidi, I. Grundke-Iqbal, K. Iqbal, Proc. Natl.

Acad. Sci. U.S.A. 91, 5562–5566 (1994).10. A. C. Alonso, I. Grundke-Iqbal, K. Iqbal, Nat. Med. 2, 783–787

(1996).11. N. Gustke et al., FEBS Lett. 307, 199–205 (1992).12. A. W. P. Fitzpatrick et al., Nature 547, 185–190 (2017).13. D. Eliezer et al., Biochemistry 44, 1026–1036 (2005).14. X. H. Li, J. A. Culver, E. Rhoades, J. Am. Chem. Soc. 137,

9218–9221 (2015).15. H. Kadavath et al., Angew. Chem. Int. Ed. Engl. 54,

10347–10351 (2015).

16. J. Al-Bassam, R. S. Ozer, D. Safer, S. Halpain, R. A. Milligan,J. Cell Biol. 157, 1187–1196 (2002).

17. S. Kar, J. Fan, M. J. Smith, M. Goedert, L. A. Amos, EMBO J. 22,70–77 (2003).

18. M. F. Chau et al., Biochemistry 37, 17692–17703 (1998).19. L. Serrano, E. Montejo de Garcini, M. A. Hernández, J. Avila,

Eur. J. Biochem. 153, 595–600 (1985).20. H. Kadavath et al., Proc. Natl. Acad. Sci. U.S.A. 112, 7501–7506

(2015).21. Y. Song et al., Structure 21, 1735–1742 (2013).22. R. Y. Wang et al., eLife 5, e17219 (2016).23. D. Panda, B. L. Goode, S. C. Feinstein, L. Wilson, Biochemistry

34, 11117–11127 (1995).24. J. Biernat, N. Gustke, G. Drewes, E. M. Mandelkow,

E. Mandelkow, Neuron 11, 153–163 (1993).25. M. Hasegawa et al., J. Biol. Chem. 267, 17047–17054

(1992).26. N. Nagano, M. Ota, K. Nishikawa, FEBS Lett. 458, 69–71

(1999).27. B. Trinczek, A. Ebneth, E. M. Mandelkow, E. Mandelkow,

J. Cell Sci. 112, 2355–2367 (1999).28. R. Zhang, G. M. Alushin, A. Brown, E. Nogales, Cell 162,

849–859 (2015).29. F. Devred et al., Biochemistry 43, 10520–10531 (2004).30. B. Trinczek, J. Biernat, K. Baumann, E. M. Mandelkow,

E. Mandelkow, Mol. Biol. Cell 6, 1887–1902 (1995).31. A. E. Prota et al., Science 339, 587–590 (2013).

ACKNOWLEDGMENTS

We thank A. Chintangal and P. Tobias for computational support. Wealso thank P. Grob and D. Toso for EM support. We thank D. Dynermanand N. Grigorieff for help with implementing phase plate support infrealign. We thank B. Greber for help with initial model building.Funding: This work was funded by a BWF collaborative research travelgrant (008185) (E.H.K.) and NIHGMS grants K99GM124463 (E.H.K.),GM123089 (F.D.), and GM051487 (E.N.). E.N. is a Howard HughesMedical Institute Investigator. Author contributions: Authorcontributions were as follows: conceptualization, E.H.K., N.M.A.H., andE.N.; methodology, E.H.K., N.M.A.H., F.D., and E.N.; investigation, E.H.K.,N.M.A.H., F.D., and E.N.; writing (original draft), E.H.K. and E.N.; writing(review and editing), all authors; funding acquisition, E.N.; resources,S.P.; and supervision, E.N. and K.H.D.Competing interests:The authorsdeclare no competing interests.Data andmaterials availability:Atomicmodels are available through the Protein Data Bank (PDB) withaccessions codes 6CVJ (R1×4) and 6CVN (R2×4); all cryo-EMreconstructions are available through the EMDB with accession codesEMD-7520 (2R tau), EMD-7523 (4R tau), EMD-7522 (full-length tau),EMD-7769 (R1×4 tau), and EMD-7771 (R2×4 tau).

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/360/6394/1242/suppl/DC1Materials and MethodsFigs. S1 to S10Table S1References (32–46)

1 February 2018; accepted 30 April 2018Published online 10 May 201810.1126/science.aat1780

Kellogg et al., Science 360, 1242–1246 (2018) 16 June 2018 4 of 4

PGGG

proline-rich domain

PGGG

R1

R2

straight microtubule lattice depolymerizing microtubule

curved tubulin oligomer

anchor point

Fig. 4. Model of full-length tau binding to microtubules and tubulin oligomers. Our structuraldata lead to a model of tau interaction with MTs in which the four repeats bind in tandem alonga PF. We did not observe strong density for the region that would correspond to the PGGGmotif, which is modeled in gray for illustrative purposes and must be highly flexible. Tau bindingat the interdimer interface, interacting with both a- and b-tubulin, promotes association betweentubulin dimers. The tau-binding site is also the location of the previously identified “anchorpoint” [rightmost box; orange is bent tubulin (PDB code 4I4T) (31), blue and green are straighttubulin (PDB code 3JAR) (28), and purple corresponds to the tightly bound region of a taurepeat]; thus tau-tubulin interactions are unlikely to change substantially with PF peelingduring disassembly (center) or with binding to small, curved tubulin oligomers (top right).

RESEARCH | REPORTon O

ctober 20, 2020

http://science.sciencemag.org/

Dow

nloaded from

Near-atomic model of microtubule-tau interactionsElizabeth H. Kellogg, Nisreen M. A. Hejab, Simon Poepsel, Kenneth H. Downing, Frank DiMaio and Eva Nogales

originally published online May 10, 2018DOI: 10.1126/science.aat1780 (6394), 1242-1246.360Science 

, this issue p. 1242Scienceinteraction and tau aggregation.corresponds to a clinically relevant site of tau phosphorylation, explaining the competition between microtubuleand thus stabilizing the polymer. A key tau amino acid within the tightly bound segment between tubulin subunits

subunitsmolecular modeling to show how tau interacts with the outer surface of the microtubule, stapling together tubulin electron microscopy and− used cryoet al.microtubules. Microtubule-tau interactions have been mysterious. Kellogg

Tangle formation is preceded by phosphorylation events that cause tau to dissociate from its native binding partner,accumulation of neurofibrillary tangles composed of tau, a protein important for neuronal development and function.

Alzheimer's disease is a major cause of death in the elderly. Disease progression is associated with theTackling microtubule-tau interactions

ARTICLE TOOLS http://science.sciencemag.org/content/360/6394/1242

MATERIALSSUPPLEMENTARY http://science.sciencemag.org/content/suppl/2018/05/09/science.aat1780.DC1

REFERENCES

http://science.sciencemag.org/content/360/6394/1242#BIBLThis article cites 46 articles, 14 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Science. No claim to original U.S. Government WorksCopyright © 2018 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of

on October 20, 2020

http://science.sciencem

ag.org/D

ownloaded from