Design of a molecular support for cryo-EMstructure determinationThomas G. Martina, Tanmay A. M. Bharata,b, Andreas C. Joergera,c, Xiao-chen Baia, Florian Praetoriusd, Alan R. Fershta,Hendrik Dietzd,1, and Sjors H. W. Scheresa,1

aMedical Research Council Laboratory of Molecular Biology, Cambridge Biomedical Campus, Cambridge CB2 0QH, United Kingdom; bSir William DunnSchool of Pathology, University of Oxford, Oxford OX1 3RE, United Kingdom; cGerman Cancer Consortium (DKTK), Institute of Pharmaceutical Chemistry,Johann Wolfgang Goethe University, 60438 Frankfurt am Main, Germany; and dPhysik Department, Walter Schottky Institute, Technische UniversitätMünchen, 85748 Garching near Munich, Germany

Edited by Fred J. Sigworth, Yale University, New Haven, CT, and approved October 13, 2016 (received for review August 2, 2016)

Despite the recent rapid progress in cryo-electron microscopy(cryo-EM), there still exist ample opportunities for improvementin sample preparation. Macromolecular complexes may disassociateor adopt nonrandom orientations against the extended air–waterinterface that exists for a short time before the sample is frozen. Wedesigned a hollow support structure using 3D DNA origami to pro-tect complexes from the detrimental effects of cryo-EM samplepreparation. For a first proof-of-principle, we concentrated on thetranscription factor p53, which binds to specific DNA sequences ondouble-stranded DNA. The support structures spontaneously formmonolayers of preoriented particles in a thin film ofwater, and offeradvantages in particle picking and sorting. By controlling the posi-tion of the binding sequence on a single helix that spans the hollowsupport structure, we also sought to control the orientation ofindividual p53 complexes. Although the latter did not yet yieldthe desired results, the support structures did provide partial in-formation about the relative orientations of individual p53 com-plexes. We used this information to calculate a tomographic 3Dreconstruction, and refined this structure to a final resolution of∼15 Å. This structure settles an ongoing debate about the symme-try of the p53 tetramer bound to DNA.

cryo-EM | DNA-origami | single particle analysis | structural biology | p53

Cryo-electron microscopy (cryo-EM) structure determinationof biological macromolecules is undergoing rapid progress.

With the advent of efficient direct electron detectors and thedevelopment of powerful algorithms for image processing, nu-merous structures to near-atomic resolution have been reportedin the past few years (1, 2). In cryo-EM single-particle analysis,solutions of purified protein and/or nucleic acid complexes aretypically applied to a thin, amorphous carbon film with micro-meter-sized holes in it that is held in place by a metal grid. Excessliquid is then blotted away with filter paper, and the sample israpidly plunged in liquid ethane (3, 4). This procedure ideallyresults in the formation of a film of vitreous ice that is onlyslightly thicker than the macromolecular complex of interest.Keeping the frozen sample at liquid nitrogen temperatures allowsits insertion into the high vacuum of a transmission electron mi-croscope and limits the effects of radiation damage by the elec-trons that are used for imaging (5). Images taken through theholes of the carbon film ideally contain 2D projections of many,assumedly identical copies of the macromolecular complex ofinterest, which are often called particles. Projections from dif-ferent viewing directions can then be combined in a 3D re-construction of the scattering potential of the molecule (6). If theresulting map approaches a resolution of 3 Å, it allows buildingan atomic model of the molecules, from which useful in-formation about their function may be derived.A major hurdle in single-particle analysis is the need to re-

cover the relative viewing angles of the individual particles. Thisinformation is lost in the experiment because every particleadopts an uncontrolled orientation in the ice layer. The viewing

angles are, therefore, determined a posteriori by image-pro-cessing algorithms that match the experimental projection ofevery individual particle with projections of a 3D model (7).However, the projection-matching procedure is ultimately ham-pered by radiation damage. Because the electrons that are usedfor imaging destroy the very structures of interest (see ref. 8 for arecent review), one needs to limit carefully the number of elec-trons used for imaging. This procedure results in high levels ofexperimental noise, which in turn lead to errors in the a poste-riori determination of the viewing angles. These errors imposesevere limitations on the 3D reconstruction, in particular forsmaller complexes, because the signal-to-noise ratio in the im-ages decreases with the size of the particles. If one could ex-perimentally control the orientations of each particle in the icelayer, then, in principle, one could determine structures to higherresolution and of smaller complexes. This situation is illustratedby samples where the relative orientations of many molecules isset—for example, in 2D crystals or helical assemblies of proteinmolecules. In such cases, near-atomic-resolution reconstructionswere already achieved decades ago, and from much smallermolecules than currently possible with single-particle analysis (9–12). Both developments that triggered the recent revolution inattainable resolution of cryo-EM single-particle analysis directlyaddressed this same hurdle. Better detectors led to lower levels

Significance

As the scope of macromolecular structure determination bycryo-electron microscopy (cryo-EM) is expanding rapidly, it isbecoming increasingly clear that many biological complexesare too fragile to withstand the harsh conditions involved inmaking cryo-EM samples. We describe an original approach toprotect proteins from harmful forces during cryo-EM samplepreparation by enclosing them inside a three-dimensionalsupport structure that we designed using DNA origami tech-niques. By binding the transcription cofactor p53 to a specificDNA sequence, and by modifying the position of this sequencein our support structure, we also sought to control the relativeorientation of individual p53:DNA complexes.

Author contributions: T.G.M., A.R.F., H.D., and S.H.W.S. designed research; T.G.M., T.A.M.B.,and X.-c.B. performed research; A.C.J. and F.P. contributed new reagents/analytic tools;T.G.M., T.A.M.B., X.-c.B., and S.H.W.S. analyzed data; and T.G.M., A.C.J., A.R.F., H.D., andS.H.W.S. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.

Data deposition: The p53 reconstruction reported in this paper has been deposited in theElectron Microscopy Data Bank, https://www.ebi.ac.uk/pdbe/emdb (accession no. 3453).1To whom correspondence may be addressed. Email: scheres@mrc-lmb.cam.ac.uk ordietz@tum.de.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1612720113/-/DCSupplemental.

E7456–E7463 | PNAS | Published online November 7, 2016 www.pnas.org/cgi/doi/10.1073/pnas.1612720113

Dow

nloa

ded

by g

uest

on

Aug

ust 3

, 202

0

of experimental noise, and better image-processing algorithmsled to more accurate viewing angles.Another complication of cryo-EM structure determination lies

in the way the sample is prepared (see also refs. 13 and 14). Theexact physics of cryo-EM sample preparation is poorly described,but several factors may negatively affect its results. First, theblotting process itself may involve strong forces in the samplethat destroy fragile protein complexes. Second, during the shorttime between blotting and vitrification, the macromolecules arein a thin liquid film that extends for millimeters to the side, but isonly a few hundred angstroms thick. Brownian motion will causethe macromolecules to collide with the air–water interface>1,000 times per second (15). Biological macromolecules mayunfold when they hit the air–water interface (16), or they mayadsorb to this interface in a nonrandom manner—for example,by presenting their most hydrophobic patch to it. This interactionleads to an uneven distribution of viewing angles in the cryo-EMimages, and the corresponding lack of different views mayhamper 3D reconstruction. Despite the numerous successes ofsingle-particle analysis in recent years, preparing suitable cryo-EM samples therefore remains a nontrivial task. Often, estab-lishing optimal ice thickness and particle distribution in the icerequires careful optimization of blotting and freezing conditionsby an experienced experimentalist, and fragile complexes haveoften been observed to fall apart (17, 18).This work describes an attempt to address both the problems

of the unknown viewing angles and the unfavorable conditions ofsample preparation from an experimental perspective. To thisend, we set out to design a support structure that simultaneouslyexerts control over the orientations of individual particles, whilealso controlling the ice thickness and protecting the particlesfrom the air–water interface. We chose to use 3D DNA-origamitechniques to design such a support structure. DNA-origamiallows flexible and customizable design of 3D structures at thenanometer scale (19–21). The use of 2D arrays of DNA scaffoldto bind proteins for cryo-EM sample preparation has been pro-posed before (22). Many proteins naturally interact with nucleicacids like DNA or RNA. In the first instance, to limit the numberof technological challenges, we designed a support structure forproteins that naturally bind to double-stranded DNA in a se-quence-dependent fashion. This approach includes many pro-teins involved in the regulation of transcription, and we chose thetranscription factor p53 as a paradigm.p53 plays a central role in the cell cycle (23) and is best known

for its function in tumor suppression (24). The active form ofhuman p53 is a homotetramer of 4 × 393 amino acids. Its domainorganization consists of an intrinsically disordered N-terminaltransactivation domain, a proline-rich region, a structured DNA-binding domain (DBD), a tetramerization domain connected viaa flexible linker, and an intrinsically disordered C-terminal reg-ulatory domain (25). The structures of the DBD and the tetra-merization domain of human p53 have been solved by X-raycrystallography and NMR (26–33), as have tetrameric complexesof the DBD with fragments of DNA (34, 35). The structures offull-length and truncated mutants of human p53 bound to DNAhave been analyzed by negative-stain EM and small-angle X-raydiffraction (36, 37), and the DBDs have the same symmetry asfound in the crystallographic studies. Other cryo-EM studies onmouse p53 propose a different symmetry (38, 39). Because the Cterminus of p53 increases unspecific binding to DNA (40), wechose to use the human truncated p531–360 construct (with amolecular mass of ∼160 kDa for the tetramer) in our experiments.Below, we describe the rationale behind our approach, the resultsobtained with this p53 construct, and the opportunities that thistype of support structure offer for more general cryo-EM samplepreparation procedures.

ResultsSupport Structure Design. Fig. 1 illustrates the design of our pro-posed DNA-origami support structure. We designed a hollowpillar with a honey-combed motif of 82 parallel double-strandedDNA (dsDNA) helices with a height of 26 nm. Two parallelarrays of DNA helices create a central cavity of 13.2 × 13.6 nmand outer dimensions of 26.4 × 32.7 nm. A dsDNA helix with acentral p53-specific binding sequence spans the center of thehollow space. Ten parallel helices at the periphery, which we willcall the flag (Fig. 1 A and B, top left), make the structure asym-metric so that top and bottom views can be readily distinguished.Different aspects of this design aim to address four main ob-jectives of our support structure.First, we aimed to keep the target protein in the middle of a

sufficiently thin ice layer and away from the air–water interface.p53 binds preferentially to the sequence GGACATGTCCG-GACATGTCC. By including this sequence in the central dsDNAhelix (Fig. 1 A and B, red), our target protein would bind to theapproximate center of our support structure. By designing ssDNAoverhangs (T10) on both the top and bottom faces of the pillar(Fig. 1 A–C, blue), we aimed to expedite a preferred orientation ofthe support in the ice layer. The ssDNA overhangs have exposedaromatic bases, which make them more likely to interact with thehydrophobic air–water interface than the sides of the pillar, whichis more hydrophilic because of its exposed DNA backbones. Thisrationale was inspired by initial observations of a preferred ori-entation in the ice layer of a different DNA-origami object, wheresimilar overhangs were added to prevent end-to-end polymeriza-tion of the structures (41) (Fig. S1). If both faces of the pillarwould indeed interact with the air–water interface at the top andbottom of the sample, then the height of the pillar itself woulddetermine the thickness of the ice layer. By designing a distance of∼260 Å between the ssDNA overhangs at the top and the bottomof the pillar, the p53 complex, with an expected size of <120 Å,would then not interact with either of the air–water interfaces. Inaddition, by anchoring the protein in the middle of the ice layer,differences in defocus height that might lead to loss of resolutionin relatively thick ice layers (42) can be prevented.Second, by effectively enclosing the target protein inside the

support structure, we aimed to protect it from blotting forces,aggregation, or other harmful interactions. It is also conceivablethat interactions of the target protein with the holey carbon film,with itself (in the form of aggregation), or with the air–water interfacemay all influence optimal blotting conditions. Therefore, enclosingthe target protein inside a DNA-origami structure that is always thesame may also have a beneficial effect on the reproducibility of op-timal blotting conditions for different target proteins.Third, even though we enclosed the target protein inside the

support, we still aimed to minimize the overlap in projecteddensities of the target protein and the support, because thisdensity overlap would hamper alignment of the particle. Thecentral dsDNA helix traverses the middle of the hollow pillar,and the target protein binds to (approximately) the center ofthe pillar. Thereby, provided the support structure adopts theintended orientation in the ice, with its open top or bottom facesperpendicular to the electron beam, the projected density fromthe target protein may be separated from the projected density ofthe support by excising smaller subimages (also see Fig. 2).Lastly, we also sought to exert experimental control over the

orientation of our target protein, p53, inside the support struc-ture. Because of the helical character of the central dsDNA, wecan induce different relative orientations of our target proteinwith respect to the support structure. By translating the bindingsequence 1 bp upward, the p53 complex will rotate 34° along theaxis of the dsDNA (Fig. 1D). Thereby, by making five differentversions of our support structure, shifting the p53-bindingsequence one base pair at a time, we can generate different

Martin et al. PNAS | Published online November 7, 2016 | E7457

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

PNASPL

US

Dow

nloa

ded

by g

uest

on

Aug

ust 3

, 202

0

orientations that cover the entire 180° of a tomographic tilt series(Fig. 1D). Therefore, we also refer to the central dsDNA helix asthe tilt axis. The capability of rotating a single copy of our targetprotein complex in a controlled manner around the tilt axis turnsour design from a passive support into the nanoscale equivalentof a sample holder with a tilting stage. However, as we will dis-cuss in more detail below, in practice, it is difficult to achieveprecise control over the orientation of p53 along the tilt axis.

Synthesis and Imaging of the Support Structure. The designedDNA-origami structure was synthesized and purified with a yieldbetween 50% and 90% by using standard procedures (Materialsand Methods). We first used the support structure alone (i.e.,without the target protein) to determine suitable freezing condi-tions for cryo-EM grid preparation. We found conditions wherethe support structures adopt a pseudo 2D-crystalline arrangementin large areas of the grid, where most structures adopt theintended top-bottom orientation in the ice layer (Fig. 2A). Usingcryo-electron tomography, we confirmed that the ice layer is in-deed as thick as the designed DNA-origami support structures(Fig. 2B and Movie S1). Interestingly, in areas where the iceappeared thinner than the designed height, we observed no sup-port structures. In areas where the ice appeared thicker, we alsoobserved support structures in side-view orientations (Fig. S2).Subsequently, we made cryo-EM grids of the support structure

together with p53.Wemade five different samples, each with a differentposition of the p53 binding sequence on the tilt axis. As hypothesized,

we were able to use the same blotting conditions as for the emptysupport structures, although the grids exhibited fewer areas with theoptimal, near-crystalline arrangements of the support structures.Nevertheless, we could use the observation that support structurespreferentially adopt top-bottom views in regions with the desired icethickness to select suitable areas for data acquisition from relativelylow-magnification overviews in the microscope (Fig. S2). Data acquisi-tion was performed for each tilt axis setting separately on an FEI TitanKrios microscope at 300 kV with a K2 Summit detector (Table 1).

The 2D Image Preprocessing. The incorporation of the targetprotein inside a larger support structure not only providesexperimental information about its orientation, it also fa-cilitates the selection of individual particles from the mi-crographs. We use the template-based particle selectionprocedure in RELION for this process (43). By using a 2Dtemplate structure that corresponds to the top view of oursupport structure (Fig. 2 A, Inset), we automatically selected272,914 particles from the five different experiments andcombined all of these particles into a single dataset.To remove inadvertently picked side views, broken support

structures, or other false positives (Fig. 2C) from the dataset, weselected 177,287 particles in three rounds of reference-free 2Dclassification. Because the support structure may adopt either atop-up or a bottom-up orientation in the ice, we applied a mirroroperation to the 53% of these particles that contributed to 2Dclass averages with a flag on the top right side of the support

A

D

B C

6.6

nm

13.2

nm

26.4

nm

26.2

nm

13.6 nm

32.7 nm

9.8 nm

1bp

34º

1bp

34º

1bp

34º

1bp

34º

Fig. 1. Design of the support structure. (A) Perspective view of the support structure. Each dsDNA helix is shown as a white cylinder. The position of thespecific binding sequence on the central dsDNA helix is shown in red; ssDNA overhangs (T10) are shown in blue. (B) Top view of the support structure. Innerand outer dimensions of the support structure are shown in gray. The dimensions of the asymmetric feature (flag) are shown in green. (C) Side view of thesupport structure. (D) Illustration of five different settings for the p53-specific binding sequence on the central dsDNA helix. A surface representation of thetetrameric DBD of p53 is shown in red. In each representation, the p53-binding site is shifted one base upward from left to right. Because of the helical natureof the dsDNA, this shift also results in a rotation of the p53 complex.

E7458 | www.pnas.org/cgi/doi/10.1073/pnas.1612720113 Martin et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 3

, 202

0

structure (Fig. 2D and Table 1). The resulting set of intactsupport structures was aligned to a common reference (Fig. 2E),to allow extraction of the central part of the individual particleimages in a smaller box. An additional round of reference-free2D classification was used to discard particles for which no ob-vious p53 density was visible (Fig. 2F). In this calculation, weused a prior probability, or prior in short, on the in-plane rota-tions. In the empirical Bayesian approach to image processing inRELION (44), prior probabilities on orientational parametersare expressed as Gaussian functions centered on the expectedvalue for that orientation and with a SD that expresses the un-certainty in that expectation. In this case, we centered the prioron the in-plane rotations from the initial 2D alignment of theentire support structure and used a SD of 15° to allow for errorsin those assignments. At this point, 36,837 particle images wereselected. A final round of particle selection was based on iden-tifying particles that rotated too much in the p53-only re-alignment with respect to the support structure (Fig. 2G) and ledto a final dataset of 22,698 subimages containing p53.

The 3D Reconstruction. The 2D alignment parameters of thesupport structures, combined with the known setting of the

tilt axis, provide information about the five parameters that definethe orientation of every p53 complex. By using only 2D alignmentparameters, one can therefore calculate a tomographic 3D re-construction of p53. The two in-plane translations of the supportstructure, combined with the known offset of the binding sequence onthe tilt axis, can be used to center the target protein. In addition,three Euler angles describe the relative orientations of all p53 par-ticles with respect to a common 3D frame of reference. By using areference with the dsDNA tilt axis along the y axis, the first Eulerangle (“rot” in RELION) is expected to be ∼0°, because the tilt axis isdesigned to run perpendicular to the top–bottom axis of the supportstructure. The second Euler angle (“tilt”) is determined by the tilt-axis setting of the experiment. By defining the tilt angle for ourcentral tilt axis position (0 bp) as 90°, the 1- or 2-bp shifted po-sitions of the binding sequence on either side have expected tiltangles of 21.4° (−2 bp), 55.7° (−1 bp), 124.3° (+1 bp), and 158.6°(+2 bp). Finally, the third Euler angle (“psi”) is directly availablefrom the in-plane rotation determined in the 2D alignment of thesupport structure.To calculate the initial tomographic p53 reconstruction, we

first split the selected 22,698 subimages by their original tilt axissettings and performed five separate 2D classifications using a

A

B

D E F

C

G

Fig. 2. Image-processing strategy. (Scale bars: 20 nm.) (A) Part of a typical micrograph. A template for automated particle picking is shown in the top rightcorner. (B) Tomographic side view of a hole showing a monolayer of support structures. Near the edge of the hole (indicated with an arrow), the ice getsslightly thicker. (C) Examples of 2D classes that are discarded. (D) Examples of 2D classes of intact structures with the flag in the top left or top-right. A mirroroperation is applied to images with the flag on the top right. (E) The average of all intact particles with the correct orientation in ice, including the mirroredparticles, is used as a template for their alignment. (F) Subimages (cyan) with a width and height of 20 nm are extracted from the aligned particles andsubmitted to 2D classification with a prior on the in-plane rotation. A circular mask that is applied during this process is shown in yellow. Three types ofparticles are distinguished: supports without tilt axis (F, Top), supports with tilt axis but without a density for p53 (F, Middle), and supports with tilt axis andp53 density (F, Bottom). (G) Illustration of the final selection of particles, where the angle from a realignment of the p53-only subimage is compared with theangle of the entire support structure, and particles with large differences in these angles (e.g., Lower) are discarded.

Martin et al. PNAS | Published online November 7, 2016 | E7459

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

PNASPL

US

Dow

nloa

ded

by g

uest

on

Aug

ust 3

, 202

0

prior on the in-plane rotations of 15°. From these five runs, weselected 35 total 2D class averages with the best protein-likefeatures (Fig. 3A). We then used the expected Euler angles asdefined above to perform a tomographic reconstruction, directlyfrom the 2D class averages. A preliminary map without sym-metry (Fig. 3B) indicated the presence of C2 symmetry, whichwas subsequently imposed (Fig. 3C). Apart from the twofoldrotational symmetry, the reconstructed map also showed localtranslational pseudosymmetry along the dsDNA axis.The observations that particles from different tilt-axis settings

gave rise to similar 2D class averages, and that different 2D classaverages were observed within a single tilt-axis dataset, indicatethat the tilt axis is probably more flexible than anticipated. Inprinciple, the statistical framework of RELION is well-suited tomodel deviations from the expected orientations of each indi-vidual particle through the use of Gaussian priors on the trans-lations and the Euler angles. The SD of the Gaussian priors canbe used to tune the amount of expected deviations. Larger devi-ations from the expected orientations (e.g., because the attach-ment of the target protein to the DNA-origami structure is more

flexible than anticipated) will lead to less-informative priors. Tosome extent, deviations from the anticipated tilt angle are actuallybeneficial to the reconstruction process, because they will lead to amore uniform angular sampling along the tomographic tilt axis.Therefore, we first performed a rotational and translational

realignment of the 35 selected 2D class averages, where we useda prior with a SD of 10° on rot and psi, but tilt was left free. Theresulting reconstruction (Fig. 3D) was then used as an initialmodel in a set of 3D refinements, where we used the 9,271 p53particles that were assigned to the selected 2D classes. Withoutusing rotational priors on any of the Euler angles, or when usingonly a prior on rot or psi, the reconstructions looked worse (e.g.,density for the dsDNA disappeared) than when using a 5° prioron both rot and psi. As expected from the observation that ourtilt axis is more flexible than anticipated, imposing a prior on tiltmade the reconstruction worse (Fig. S3). When using pro-gressively less informative (i.e., broader, priors on rot and psi), thereconstructions also became worse (Fig. S4), whereas priors withSDs <5° are too narrow for the angular sampling rate used in therefinement. Consequently, our best refinement had a SD of 5° for

Table 1. Data acquisition statistics for each tilt axis setting

Tilt axissetting, bp Micrographs

Autopickedparticles

Intact top viewsupports (mirrored)

Emptysupports

Supports withonly DNA

Supports withp53-like density

p53 in expectedorientation

p53 in 2D classesused for initial-model

calculation

−2 604 69,214 36,727 (19,672) 11,018 15,439 10,270 5,821 2,363−1 415 51,830 40,824 (26,886) 15,916 17,082 7,826 4,831 2,1970 582 61,528 41,131 (19,122) 18,452 16,587 6,092 3,373 1,375+1 576 65,942 39,303 (19,526) 14,424 15,665 9,214 5,169 1,795+2 385 24,400 19,302 (8,988) 6,319 7,412 5,571 3,504 1,541Total 2,562 272,914 177,287 (94,194) 66,129 72,185 38,973 22,698 9,271

A B C D

E F

Fig. 3. Initial model generation and final maps. (A) Class averages used for the initial model generation sorted by tilt axis setting. An illustration of thedesigned orientation is shown on the left, and the applied tilt angle is shown in front of the class averages. (B) Initial model reconstructed with the expectedtilt angles in C1. The angles for rot and psi were all set to 0°. In RELION, the first rotation of the 3D reference object (rot) is around the z axis, which comes outof the xy plane of the figure; the second rotation (tilt) is around the new y axis, and the third rotation (psi) is around the new z axis. B, Inset shows a simplifiedexplanation of these rotations from the point of view of the experimental particles. In that case, psi is the in-plane rotation, tilt is the rotation around thecentral DNA axis, and rot describes out-of-plane rocking. (C) The same model as in B with C2 symmetry imposed. The twofold symmetry axis is along the z axisand is indicated with an oval. (D) Map after realignment of the class averages with 5° priors on the rot and psi angles and unrestricted tilt angles. (E) Differentviews of the final map generated from 9,271 particles. The estimated resolution is ∼15 Å. Protein Data Bank model 4HJE (35) of the DBD in blue. (F) Histogramof the refined tilt angles for each of the tilt axis settings.

E7460 | www.pnas.org/cgi/doi/10.1073/pnas.1612720113 Martin et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 3

, 202

0

the priors on psi and rot and left the tilt angle unrestrained.The resulting map, at an estimated resolution of ∼15 Å, shows thesame domain architecture as the known crystal structure of theDBD of p53 bound to dsDNA (35) and shows additional densitiesin both the front and the back of the complex (Fig. 3E). Thehistogram of the refined tilt angles confirms that the tilt axis ismore flexible than anticipated. Some information in the tilt angleis maintained, but angles ∼90° are somehow disfavored (Fig. 3F).

DiscussionWe present an approach to sample preparation for cryo-EMstructure determination of biological macromolecules. Using 3DDNA-origami, we designed a support structure with a definedsize and shape that binds specifically to a target protein of interest.The resulting structure is an artificial molecular support that ex-erts experimental control over the orientations of individual pro-tein molecules and protects them from aggregation or harmfulinteractions with the air–water interface. In addition, the supportstructure may facilitate the optimization of freezing conditionsand aid in the selection of suitable ice thickness for data acqui-sition. Nevertheless, as discussed below, this work does not yetrepresent a ready-to-use solution for high-resolution structure de-termination of a wide range of different target proteins, but shouldrather be considered as a proof-of-principle toward achieving thisambitious goal.By choosing a target protein that naturally binds to dsDNA in

a sequence-specific manner, our support structure was designedto act as the nanoscale equivalent of a goniometer that exertsexperimental control over the orientations of individual proteincomplexes. By using five different positions of the p53 bindingsequence on a central dsDNA helix in our structure, our struc-ture intended to provide five different views of a p53 tetramerbound to the support structure. Although the prior informationabout the five different orientations could indeed be used suc-cessfully to calculate an initial tomographic reconstruction of thep53 tetramer bound to DNA, further refinement of the orien-tations of 2D class averages or individual particle imagesrevealed a distribution in tilt angles that is much broader thanone would expect from a rigid dsDNA helix, and somehow seemsto disfavor tilt angles ∼90°. Because the complete structure ofthe p53 tetramer remains unknown, it could be that the supportstructure is too small to accommodate p53 in this orientation.Analysis of multiple 2D class average images from our data(Movie S2) revealed that the support structures are also not asrigid as one might need to exert precise orientation control.Moreover, despite previous attempts at optimizing the in-corporation of the central dsDNA helix in the support structure(45), it could be that our current design (Figs. S5 and S6) still leadsto incorrect attachments in a subset of the structures. Neverthe-less, despite the lack of complete control over the tilt angle, thesupport structure still provided information about the center ofthe particle, its in-plane rotation (the psi angle), and its out-of-planerocking (the rot angle), and this information could be used instatistical priors to improve the 3D reconstruction. The intendeddesign of our support structure as a molecular goniometer thatprovides experimental information about the orientations of in-dividual p53 complexes was, therefore, partially successful.Our experiments also revealed several challenges that will

need to be overcome in future studies. First of all, the finalnumber of 9,271 selected p53 particle images represents a lowyield from 2,562 selected micrographs. Approximately 42% ofour support structures did not incorporate the tilt axis, and 77%of the remaining support structures did not bind to p53. Thelatter was surprising, because at a concentration of 1 mM p53tetramers and 150 nM support structures, and with a dissociationconstant of ∼50 nM for dsDNA, we had expected almost allsupport structures to contain p53. Possibly, steric hindrance fromthe support or strain in the central dsDNA affected the binding

of p53. A further reduction in particle number came from ourimage-selection procedures: 38% of the particles that containeddensity for p53 had it bound in an unrealistic orientation in re-spect to the support structure, and from these only 41% con-tributed to the selected 2D classes of the subimages that showedthe best protein-like features. Second, although the low particleyield may have hampered reaching higher resolution, the se-lected particles still represented >18,000 asymmetric units, whichshould probably have yielded a higher resolution reconstruction(cf. ref. 46). One possibility could be that the central cavity of thesupport structure was too small for the defocus used. De-localization in the images because of defocusing may have led toa superposition of the signal from the support structure and thetarget protein, which may have limited the alignment of the in-dividual target proteins, and therefore the resolution of the finalreconstruction. This problem could in principle be circumventedby in-focus imaging through phase-plate technology (47). Inaddition, future designs of larger, more stable support structureswith a more accessible protein binding site, and with better in-corporation of the tilt axis may improve both particle yield andresolution. To create these larger supports with low defect rates,new assembly and purification strategies will need to be con-sidered and adopted (see, for example, refs. 48 and 49).Nevertheless, at a resolution of ∼15 Å, the resulting recon-

struction provides useful information about how p53 binds todsDNA. The observation that our reconstruction shows C2 sym-metry with an additional twofold translational component alongthe DNA axis, which is similar to the crystal structure of the tet-rameric core of p53 bound to dsDNA, indicates that p53 does notbind to DNA in a previously proposed arrangement with D2symmetry (38, 39). Instead, the presence of extra density in thefront and back of the complex (Fig. 3E) provides support for analternative model where the p53 tetramerization domain binds tothe opposite side of the dsDNA helix from the DBD (36).The design of our support structure also contains several

useful features that may inspire future experiments. By enclosingthe target protein in a hollow structure, it is protected from in-teractions with copies of itself or with the hydrophobic air–waterinterface during cryo-EM sample preparation. This protectionmay prevent the proteins from aggregation, or from unfolding oradopting preferred orientations against this air–water interface.Also, it might be that the optimization of freezing conditions willdepend more on the support structure than on the nature of theprotein inside, so that the optimal conditions would differ lessfor different proteins compared with current methods. We in-deed used similar conditions for the samples with only supportstructures and the mixture with p53, although the mixtureshowed fewer regions of near-crystalline arrangements of thesupport structures. We used approximately six times more p53tetramers than support structure in our mixture and did not at-tempt to remove unbound p53 tetramers through additionalpurification steps. It could therefore be that interactions of un-bound p53 tetramers with the air–water interface still had aneffect on the freezing conditions. In addition, the observationthat the support structures adopt monolayers in an ice layer withthe intended thickness is potentially useful. One might, for ex-ample, try to add support structures to standard cryo-EM sam-ples with the idea of using them as “spacers” (13) to control icethickness. Admittedly, it remains difficult to conclude whetherthe support structures maintain the desired ice thickness over agiven area or whether the monolayers just happen to form inthose areas where the ice thickness is ideal. However, becauseone can easily select areas with monolayers of support structuresfrom low-magnification overview images in the microscope, thesupport structures do facilitate the data acquisition process. Theinteractions of our support structure with the air–water interfacemay also have a beneficial effect on the local particle concen-tration. For a 300-Å-thick layer of a 150 nM solution, one would

Martin et al. PNAS | Published online November 7, 2016 | E7461

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

PNASPL

US

Dow

nloa

ded

by g

uest

on

Aug

ust 3

, 202

0

expect only a single support structure in every micrograph. Ourobservation of ∼100 structures per micrograph suggests that inter-actions of the support structure with the air–water interface, pos-sibly both at the top and the bottom, may lead to a strong localenrichment in concentration. For proteins, a similar enrichment hasbeen attributed to a sticky layer of denatured protein at the air–water interface (15). Because our samples without protein show asimilar—or even stronger—enrichment in DNA support structures,unbound p53 tetramers are probably not required for the enrich-ment. It is difficult to assess whether the interactions with the air–water interface lead to unfolding of the support structures them-selves. We did observe partial structures (Fig. 2 A and C), but thesecould also be explained by folding defects. It therefore remainsunclear whether the presence of denatured DNA material at theair–water interface is important for the enrichment effect.The experiments described here rely on the specific binding of

a target protein that naturally binds to dsDNA in a sequence-specific manner. Although several transcription factors areknown to do so, the design of a more broadly applicable supportstructure would be desirable. It would be relatively straightfor-ward to include a DNA mismatch in the central axis, which couldinclude target proteins from DNA-mismatch repair pathways.Alternatively, one could construct a central axis with a bubble orfrom DNA/RNA hybrids to include a range of different targetproteins related to replication, transcription, or DNA repair.However, to design a support that could bind to an arbitrarytarget protein, one would probably need to combine chemicalmodifications of some of the DNA staples within the supportwith a specific tag on the target protein. For example, one coulduse commercially available biotin-labeled DNA on the supportstructure to bind a counterpart to a protein tag attached to amonovalent streptavidin (50, 51). For the study of membraneproteins, one might even consider designs where nanodiscs (52)are attached to DNA-origami support structures or where thesestructures interact directly with patches of membranes (53). Suchmore generally applicable designs probably would retain evenless control over the orientation of the target protein than thedesign described in this work. Still, the support structure couldmaintain some of its other useful features, such as facilitatingparticle selection, preventing aggregation, and keeping proteinsaway from harmful interfaces. Apart from the beneficial effects oncryo-EM sample preparation, large support structures that specif-ically bind a target protein of interest might even play a role inprotein purification. Provided that the binding of the target proteinwould be tight enough, their large molecular mass (∼5 MDa)would allow relatively straightforward separation of the targetprotein from smaller contaminants through the use of, for example,gel filtration of sucrose gradients. Alternatively, one could try to fixthe DNA supports directly on the grid surface and use the gridsthemselves for on-grid affinity purification in a similar manner aswas done with nitrilotriacetic acid-modified lipid monolayers (54,55) or antibodies attached to carbon films (56, 57).In summary, this work provides an original approach to exert

experimental control over the orientations of individual proteincomplexes and protect them from harmful forces during cryo-EM sample preparation. As the field of cryo-EM structure de-termination keeps growing, the physics involved in its samplepreparation will become clearer, and new concepts in samplepreparation will continue to emerge. Our approach providesample possibilities for further developments and may thereby

contribute to tackling some of the outstanding challenges in thisrapidly changing technique.

Materials and MethodsOrigami Design, Synthesis, and Purification. DNA-origami design (Dataset S1)was performed in cadnano (Version 0.2) (58). The scaffold DNA (the 7,560-nt-long version of the M13mp18 phage genome) was prepared as described (59).DNA staple oligonucleotides, prepared by solid-phase chemical synthesis, wereordered from Eurofins MWG. The DNA oligos for the tilt axis were ordered toHPLC-grade purity; all other oligos (Table S1) were ordered to high-purity, salt-free grade. The supports containing different tilt-axis settings were synthe-sized individually in one-pot mixtures containing 50 nM scaffold DNA, 75 nMtilt axis DNA, and 200 nM staple DNA in a 10 mM Tris buffer at pH 7.6 with1 mM EDTA and 20mMMgCl2. The mixture was incubated at 65 °C for 15 min,then annealed from 60 °C to 45 °C over the course of 8 h, and stored at 25 °C.Purification from excess staple oligos was performed in five rounds of mo-lecular mass cutoff filtration by using 100-kDa Amicon filters (Millipore) in abuffer containing 20mM Tris base and 5 mMMgCl2. The final concentration ofthe support structures was adjusted to ∼250 nM.

P53 Expression and Purification. Production and purification of truncated p53variant (residues 1–360) lacking the C-terminal regulatory domain followedpublished protocols (28, 60). Briefly, the proteins were produced in Escher-ichia coli BL21(DE3) as a fusion protein with N-terminal 6× His-tag, Bacillusstearothermophilus lipoyl domain, and tobacco etch virus protease cleavagesite. They were then purified by using standard His-tag purification proto-cols, followed by tobacco etch virus protease cleavage, heparin affinitychromatography, and a final gel filtration step on a Superdex 200 16/60preparative gel filtration column (GE Healthcare) in 300 mM NaCl, 20 mMTris (pH 7.5), and 5 mM DTT. Protein samples were flash-frozen in liquidnitrogen and stored at −80 °C. The variant contained four stabilizing mu-tations, M133L/V203A/N239Y/N268D (28, 61, 62), in the DBD.

The purified p53 samplewasmixedwith the purifiedDNA-origami sample tofinal concentrations of ∼150 nM DNA origami and ∼4 μM p53 (monomer) in a20 mM Tris buffer containing 1.5 μM DTT, 45 mM NaCl, and 4 mM MgCl2. Themixture was incubated for 20 min at 4 °C before preparing cryo-EM grids.

Electron Microscopy. Cryo-EM grids for the support structures alone or for thesupport structures with p53 and the five different settings of the tilt axis wereprepared separately using similar procedures. Aliquots of 3 μL of sample wereincubated for 10 s on glow-discharged Quantifoil grids, blotted for 2 s, andplunge-frozen in liquid ethane by using a Vitrobot Mark 3 (FEI Company).Grids were transferred to an FEI Titan Krios microscope that was operated at300 kV. Images were recorded on a K2 detector using a Gatan energy filter(with a slit width of 20 eV). Movies with a total electron dose of ∼38 e−/Å2

were recorded in superresolution mode and subsequently downscaled to afinal pixel size of 1.76 Å in RELION. The defocus was varied between 1 and5 μm. Cryo-electron tomography data were collected by using describedprocedures (63).

Electron micrographs were manually evaluated for astigmatism and drift,and 2,562 micrographs were selected for further analysis. Beam-induced mo-tion correction was performed in UCSF MOTIONCORR (64); estimation ofcontrast transfer function parameter was performed in Gctf (65); and allsubsequent image-processing operations were performed in RELION-1.4 (44).

ACKNOWLEDGMENTS. We thank Christos Savva, Shaoxia Chen, TobyDarling, and Jake Grimmett for technical support, and Miriana Petrovichfor protein purification. This work was supported by the EuropeanMolecular Biology Organisation through a long-term postdoctoral fellow-ship (ALTF-1229-2013 to T.G.M.) and an advanced fellowship (aALTF-778-2015 to T.A.M.B.), and by European Commission Marie Skłodowska-Curiepostdoctoral fellowships (to T.G.M. and X.-c.B.). The project was furthersupported by European Research Council Starting Grant GA 256270 (to H.D.);by the Deutsche Forschungsgemeinschaft through grants provided within theSonderforschungsbereich SFB863, the Center for Integrated Protein ScienceMunich, and the Nano Initiative Munich; and UK Medical Research CouncilGrants MC_UP_A024_1010 (to A.R.F.) and MC_UP_A025_1013 (to S.H.W.S.).

1. Bai X-C, McMullan G, Scheres SHW (2015) How cryo-EM is revolutionizing structuralbiology. Trends Biochem Sci 40(1):49–57.

2. Cheng Y, Grigorieff N, Penczek PA, Walz T (2015) A primer to single-particle cryo-electron microscopy. Cell 161(3):438–449.

3. Adrian M, Dubochet J, Lepault J, McDowall AW (1984) Cryo-electron microscopy ofviruses. Nature 308(5954):32–36.

4. Dubochet J, et al. (1988) Cryo-electron microscopy of vitrified specimens. Q RevBiophys 21(2):129–228.

5. Taylor KA, Glaeser RM (1976) Electron microscopy of frozen hydrated biologicalspecimens. J Ultrastruct Res 55(3):448–456.

6. Frank J, Verschoor A, Boublik M (1981) Computer averaging of electron micrographsof 40S ribosomal subunits. Science 214(4527):1353–1355.

7. Penczek P, Radermacher M, Frank J (1992) Three-dimensional reconstruction of singleparticles embedded in ice. Ultramicroscopy 40(1):33–53.

8. Glaeser RM (2016) The resolution revolution: Recent advances in cryoEM. Methods inEnzymology, ed Crowther R (Academic, New York), Vol 579, pp 19–50.

E7462 | www.pnas.org/cgi/doi/10.1073/pnas.1612720113 Martin et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 3

, 202

0

9. Henderson R, et al. (1990) Model for the structure of bacteriorhodopsin based onhigh-resolution electron cryo-microscopy. J Mol Biol 213(4):899–929.

10. Kühlbrandt W, Wang DN, Fujiyoshi Y (1994) Atomic model of plant light-harvestingcomplex by electron crystallography. Nature 367(6464):614–621.

11. Nogales E, Wolf SG, Downing KH (1998) Structure of the alpha beta tubulin dimer byelectron crystallography. Nature 391(6663):199–203.

12. Gonen T, et al. (2005) Lipid-protein interactions in double-layered two-dimensionalAQP0 crystals. Nature 438(7068):633–638.

13. Glaeser RM, et al. (2016) Factors that influence the formation and stability of thin,cryo-EM specimens. Biophys J 110(4):749–755.

14. McDowall AW, Dobro MJ, Melanson LA, Jensen GJ (2010) Cryo-EM part A: Samplepreparation and data collection. Methods in Enzymology, ed Jensen GJ (Academic,New York), Vol 481, pp 63–82.

15. Taylor KA, Glaeser RM (2008) Retrospective on the early development of cryoelectronmicroscopy of macromolecules and a prospective on opportunities for the future.J Struct Biol 163(3):214–223.

16. Glaeser RM, Han B-G (2016) Opinion: Hazards faced by macromolecules when con-fined to thin aqueous films. Biophys Rep, 10.1007/s41048-016-0026-3.

17. Thompson RF, Walker M, Siebert CA, Muench SP, Ranson NA (2016) An introductionto sample preparation and imaging by cryo-electron microscopy for structural bi-ology. Methods 100:3–15.

18. Stark H, Chari A (2016) Sample preparation of biological macromolecular assembliesfor the determination of high-resolution structures by cryo-electron microscopy.Microscopy (Oxf) 65(1):23–34.

19. Douglas SM, et al. (2009) Self-assembly of DNA into nanoscale three-dimensionalshapes. Nature 459(7245):414–418.

20. Dietz H, Douglas SM, Shih WM (2009) Folding DNA into twisted and curved nanoscaleshapes. Science 325(5941):725–730.

21. Castro CE, et al. (2011) A primer to scaffolded DNA origami. Nat Methods 8(3):221–229.

22. Malo J, et al. (2005) Engineering a 2D protein-DNA crystal. Angew Chem Int Ed Engl44(20):3057–3061.

23. Lane D, Levine A (2010) p53 research: The past thirty years and the next thirty years.Cold Spring Harb Perspect Biol 2(12):a000893.

24. Joerger AC, Fersht AR (2016) The p53 pathway: Origins, inactivation in cancer, andemerging therapeutic approaches. Annu Rev Biochem 85:375–404.

25. Joerger AC, Fersht AR (2008) Structural biology of the tumor suppressor p53. AnnuRev Biochem 77:557–582.

26. Cañadillas JMP, et al. (2006) Solution structure of p53 core domain: Structural basisfor its instability. Proc Natl Acad Sci USA 103(7):2109–2114.

27. Wang Y, Rosengarth A, Luecke H (2007) Structure of the human p53 core domain inthe absence of DNA. Acta Crystallogr D Biol Crystallogr 63(Pt 3):276–281.

28. Natan E, et al. (2011) Interaction of the p53 DNA-binding domain with its N-terminalextension modulates the stability of the p53 tetramer. J Mol Biol 409(3):358–368.

29. Cho Y, Gorina S, Jeffrey PD, Pavletich NP (1994) Crystal structure of a p53 tumorsuppressor-DNA complex: Understanding tumorigenic mutations. Science 265(5170):346–355.

30. Jeffrey PD, Gorina S, Pavletich NP (1995) Crystal structure of the tetramerizationdomain of the p53 tumor suppressor at 1.7 angstroms. Science 267(5203):1498–1502.

31. Mittl PRE, Chène P, Grütter MG (1998) Crystallization and structure solution of p53(residues 326-356) by molecular replacement using an NMR model as template. ActaCrystallogr D Biol Crystallogr 54(Pt 1):86–89.

32. Clore GM, et al. (1995) Refined solution structure of the oligomerization domain ofthe tumour suppressor p53. Nat Struct Biol 2(4):321–333.

33. Lee W, et al. (1994) Solution structure of the tetrameric minimum transforming do-main of p53. Nat Struct Biol 1(12):877–890.

34. Kitayner M, et al. (2010) Diversity in DNA recognition by p53 revealed by crystalstructures with Hoogsteen base pairs. Nat Struct Mol Biol 17(4):423–429.

35. Chen Y, et al. (2013) Structure of p53 binding to the BAX response element revealsDNA unwinding and compression to accommodate base-pair insertion. Nucleic AcidsRes 41(17):8368–8376.

36. Tidow H, et al. (2007) Quaternary structures of tumor suppressor p53 and a specificp53 DNA complex. Proc Natl Acad Sci USA 104(30):12324–12329.

37. Melero R, et al. (2011) Electron microscopy studies on the quaternary structure of p53reveal different binding modes for p53 tetramers in complex with DNA. Proc NatlAcad Sci USA 108(2):557–562.

38. Okorokov AL, et al. (2006) The structure of p53 tumour suppressor protein reveals thebasis for its functional plasticity. EMBO J 25(21):5191–5200.

39. Aramayo R, et al. (2011) Quaternary structure of the specific p53-DNA complex re-veals the mechanism of p53 mutant dominance. Nucleic Acids Res 39(20):8960–8971.

40. Anderson ME, Woelker B, Reed M, Wang P, Tegtmeyer P (1997) Reciprocal in-terference between the sequence-specific core and nonspecific C-terminal DNAbinding domains of p53: Implications for regulation. Mol Cell Biol 17(11):6255–6264.

41. Bai X-C, Martin TG, Scheres SHW, Dietz H (2012) Cryo-EM structure of a 3D DNA-origami object. Proc Natl Acad Sci USA 109(49):20012–20017.

42. Jensen GJ (2001) Alignment error envelopes for single particle analysis. J Struct Biol133(2-3):143–155.

43. Scheres SHW (2015) Semi-automated selection of cryo-EM particles in RELION-1.3.J Struct Biol 189(2):114–122.

44. Scheres SHW (2012) RELION: Implementation of a Bayesian approach to cryo-EMstructure determination. J Struct Biol 180(3):519–530.

45. Martin TG (2014) Functional synthetic DNA nanostructures. PhD dissertation (Tech-nical University of Munich, Munich).

46. Fernández IS, et al. (2013) Molecular architecture of a eukaryotic translational initi-ation complex. Science 342(6160):1240585.

47. Danev R, Baumeister W (2016) Cryo-EM single particle analysis with the Volta phaseplate. eLife 5:e13046.

48. Gerling T, Wagenbauer KF, Neuner AM, Dietz H (2015) Dynamic DNA devices andassemblies formed by shape-complementary, non-base pairing 3D components.Science 347(6229):1446–1452.

49. Shaw A, Benson E, Högberg B (2015) Purification of functionalized DNA origaminanostructures. ACS Nano 9(5):4968–4975.

50. Fairhead M, et al. (2014) SpyAvidin hubs enable precise and ultrastable orthogonalnanoassembly. J Am Chem Soc 136(35):12355–12363.

51. Veggiani G, Zakeri B, Howarth M (2014) Superglue from bacteria: Unbreakablebridges for protein nanotechnology. Trends Biotechnol 32(10):506–512.

52. Bayburt TH, Grinkova YV, Sligar SG (2002) Self-assembly of discoidal phospholipidbilayer nanoparticles with membrane scaffold proteins. Nano Lett 2(8):853–856.

53. Kocabey S, et al. (2015) Membrane-assisted growth of DNA origami nanostructurearrays. ACS Nano 9(4):3530–3539.

54. Kelly DF, Abeyrathne PD, Dukovski D, Walz T (2008) The Affinity Grid: A pre-fabri-cated EM grid for monolayer purification. J Mol Biol 382(2):423–433.

55. Walz T, Kelly DF, Dukovski D (2010) Cryo-EM part A: Sample preparation and datacollection. Methods in Enzymology, ed Jensen GJ (Academic, New York), Vol 481, pp83–107.

56. Yu G, et al. (2014) Single-step antibody-based affinity cryo-electron microscopy forimaging and structural analysis of macromolecular assemblies. J Struct Biol 187(1):1–9.

57. Yu G, Li K, Jiang W (2016) Antibody-based affinity cryo-EM grid. Methods 100:16–24.58. Douglas SM, et al. (2009) Rapid prototyping of 3D DNA-origami shapes with cadnano.

Nucleic Acids Res 37(15):5001–5006.59. Douglas SM, Chou JJ, Shih WM (2007) DNA-nanotube-induced alignment of mem-

brane proteins for NMR structure determination. Proc Natl Acad Sci USA 104(16):6644–6648.

60. Veprintsev DB, et al. (2006) Core domain interactions in full-length p53 in solution.Proc Natl Acad Sci USA 103(7):2115–2119.

61. Nikolova PV, Henckel J, Lane DP, Fersht AR (1998) Semirational design of active tumorsuppressor p53 DNA binding domain with enhanced stability. Proc Natl Acad Sci USA95(25):14675–14680.

62. Joerger AC, Allen MD, Fersht AR (2004) Crystal structure of a superstable mutant ofhuman p53 core domain. Insights into the mechanism of rescuing oncogenic muta-tions. J Biol Chem 279(2):1291–1296.

63. Bharat TAM, Russo CJ, Löwe J, Passmore LA, Scheres SHW (2015) Advances in single-particle electron cryomicroscopy structure determination applied to sub-tomogramaveraging. Structure 23(9):1743–1753.

64. Li X, et al. (2013) Electron counting and beam-induced motion correction enable near-atomic-resolution single-particle cryo-EM. Nat Methods 10(6):584–590.

65. Zhang K (2016) Gctf: Real-time CTF determination and correction. J Struct Biol 193(1):1–12.

Martin et al. PNAS | Published online November 7, 2016 | E7463

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

PNASPL

US

Dow

nloa

ded

by g

uest

on

Aug

ust 3

, 202

0