Articleshttps://doi.org/10.1038/s41564-020-0758-1

Structures of capsid and capsid-associated tegument complex inside the Epstein–Barr virusWei Liu1,2,3,8, Yanxiang Cui   1,8, Caiyan Wang1,2,4, Zihang Li   1,2, Danyang Gong5,6, Xinghong Dai1,2,7, Guo-Qiang Bi   3, Ren Sun1,5 and Z. Hong Zhou   1,2 ✉

1California NanoSystems Institute, University of California, Los Angeles, Los Angeles, CA, USA. 2Department of Microbiology Immunology and Molecular Genetics, University of California, Los Angeles, Los Angeles, CA, USA. 3Center for Integrative Imaging, Hefei National Laboratory for Physical Sciences at the Microscale, and School of Life Sciences, University of Science and Technology of China, Hefei, China. 4International Institute for Translational Chinese Medicine, Guangzhou University of Chinese Medicine, Guangzhou, China. 5Department of Molecular and Medical Pharmacology, University of California, Los Angeles, Los Angeles, CA, USA. 6Present address: Department of Therapeutics Discovery, Amgen Research, Amgen Inc., Thousand Oaks, CA, USA. 7Present address: Department of Physiology and Biophysics, School of Medicine, Case Western Reserve University, Cleveland, OH, USA. 8These authors contributed equally: Wei Liu, Yanxiang Cui. ✉e-mail: Hong.Zhou@ucla.edu

SUPPLEMENTARY INFORMATION

In the format provided by the authors and unedited.

NatuRE MiCRoBioLoGY | www.nature.com/naturemicrobiology

http://orcid.org/0000-0002-8909-6576http://orcid.org/0000-0002-4625-9283http://orcid.org/0000-0001-8735-3818http://orcid.org/0000-0002-8373-4717mailto:Hong.Zhou@ucla.eduhttp://www.nature.com/naturemicrobiology

Supporting Information

Structures of Capsid and Capsid-Associated Tegument Complex inside the Epstein-Barr Virus

Wei Liu, Yanxiang Cui et al.

List of Content Supplementary Discussion

Supplementary Tables 1-3

Supplementary Figures 1-5

1

Supplementary Discussion Inter-MCP interactions and structural plasticity of the 20 MCP conformers

First recognized in HCMV1 and subsequently confirmed in HSV2-4, HHV-6B5, and

KSHV6, the herpesvirus capsid floor is characterized by either type I, II, or III MCP-MCP

interactions (Extended Data Fig. 3e-l). Type I interaction is intra-capsomeric while type

II and type III are inter-capsomeric. Types I, II, and III interactions are formed between

two or three adjacent MCP subunits by either β-strands in the N-lasso, E-loop in the

Johnson-fold and dimerization domains (Extended Data Fig. 3g,i-j), or α-helices in the

dimerization domains (Extended Data Fig. 3h) of the neighboring MCPs. The lasso

action (type III) occurs only between hexon-hexon MCPs (Extended Data Fig. 3i-j) but

not penton-hexon MCPs (Extended Data Fig. 3k,l); this is due to the flexibility of the

penton MCP N-terminus, which causes P6 MCP to refold into a conformation that

effectively eliminates its lassoing ability (Extended Data Fig. 3k). Additionally, the inter-

capsomeric interactions between α-helices in the dimerization domains (i.e., type II) also

only occur between hexon-hexon MCPs because the dimerization domain of penton

MCP adopts a configuration which makes it unable to form type II interactions with the

dimerization domain of P1 hexon MCP, rendering the P1 dimerization domain flexible

(Extended Data Fig. 3k,l).

In contrast to these conserved MCP domain organization and MCP-MCP

interaction schemes, the structures among individual MCP subunits of the capsid exhibit

an unprecedented level of variability which we refer to as structural plasticity

(Supplementary Video 8). We identified 20 unique conformations among the MCP

structures resolved in each EBV capsid (Fig. 3; Extended Data Fig. 4): 16 within an

icosahedral asymmetric triangle (Fig. 3e), 2 portal proximal (Extended Data Fig. 4a,

right panel), and 2 related to CATC binding (Extended Data Fig. 4a, middle panel).

These MCPs show structural plasticity at different degrees depending on their

surroundings. 16 MCPs within the asymmetric unit vary in their local structures

throughout multiple domains, particularly at their floor regions (Fig. 3e-p). Compared

with P1, P6 and Pen MCPs, the structures of C1-C6, E1-E3, and P2-P5 MCPs do not

have dramatic changes. The alignment of these 13 MCP models highlights not only the

2

conformational changes at their N-lasso domains (Fig. 3j) and two fragments (a.a. 124-

172 and a.a. 1072-1091) in their Johnson-fold domains (Fig. 3k), but also the orientation

switch (i.e., being “up” and “down”) of two MCP fragments (a.a. 1149-1169 and a.a.

1256-1267) in their buttress domains (Fig. 3h,i). Each triplex adopts a unique direction,

but still can be stabilized by interacting with three quasi-equivalent neighboring MCPs.

Different interacting sites on the heterotrimeric triplexes (Supplementary Figs. 4-5)

might be responsible for the multiple binding states found in the buttress sub-domains

(Fig. 3h,i).

Variations among MCP subunits can be observed in the floor regions of P1, P6

and Pen MCPs, particularly in N-lasso (Fig. 3o) and dimerization (Fig. 3n) domains. N-

lasso of P1 MCP folds similarly with that of C1 MCP, but is completely different from

those of P6 and Pen MCPs (Fig. 3o). Likewise, the fragment of a.a. 295-374 of P6 MCP

folds in conformity with that of C1 MCP while differing vastly compared with those of P1

and Pen MCPs (Fig. 3n). It is worth noting that the fragment a.a. 1277-1323 in the

buttress domain of Pen MCP folds into a long helix instead of two short helices in those

of other 15 MCPs (Fig. 3m). As shown in Fig. 3p, the spine helix in the Johnson-fold

domain of Pen MCP tilts downward ~40º compared with the 15 hexon MCPs.

There are also variations in the P1 (and P6) MCP structures in the CATC-absent

vertex (Extended Data Fig. 4a, left panel), CATC-binding vertex (Extended Data Fig. 4a,

middle panel), and portal vertex (Extended Data Fig. 4a, right panel). Among these

three, the portal vertex displays the greatest level of structural plasticity (P1 in Extended

Data Fig. 4b-e; P6 in Extended Data Fig. 4f,g). Such structural variations may result

from differences in capsid-bindings by CATC and portal proteins. For instance, fragment

a.a. 80-96 of P1 MCP in the portal vertex flips up rather than down as those in the P1

MCPs of CATC-absent and CATC-binding vertices (Extended Data Fig. 4d). Fragments

of a.a. 305-340 and a.a. 1146-1168 are missing in P1 MCP (Extended Data Fig. 4c,e)

while a.a. 26-63 is invisible in P6 MCP (Extended Data Fig. 4g) in the portal vertex

reconstruction. These variations in hexons MCPs, particularly P1 and P6 MCPs, and

penton MCP reveal the intrinsic plasticity of MCP structure in EBV; the folding of

individual domains is preserved with gradual structural changes spreading out across

3

different domains, rather than through an abrupt large-scale conformational switch (Fig.

3e-p). Despite remaining consistent with its role in organizing the three types of MCP-

MCP interactions (Extended Data Fig. 3e-l), the greatest structural changes were

observed in N-lasso areas. Taken together, the 20 unique conformers of MCP—16

MCPs within an asymmetric unit, 2 P1 and 2 P6 MCPs in CATC-binding penton and

portal vertexes—unveil a remarkable level of structural plasticity of MCP not previously

reported for any herpesviruses.

Supplementary Tables Supplementary Table 1. Statistics of cryoEM imaging, data processing and model refinement.

Supplementary Table 2. Model building statistics.

Supplementary Table 3. RMSD statistics for MCP and Tri1.

References 1. Yu, X., Jih, J., Jiang, J. & Zhou, Z.H. Atomic structure of the human cytomegalovirus

capsid with its securing tegument layer of pp150. Science 356, eaam6892 (2017). 2. Dai, X. & Zhou, Z.H. Structure of the herpes simplex virus 1 capsid with associated

tegument protein complexes. Science 360(2018). 3. Wang, J. et al. Structure of the herpes simplex virus type 2 C-capsid with capsid-vertex-

specific component. Nature Communications 9, 3668 (2018). 4. Yuan, S. et al. Cryo-EM structure of a herpesvirus capsid at 3.1 Å. Science 360(2018). 5. Zhang, Y. et al. Atomic structure of the human herpesvirus 6B capsid and capsid-

associated tegument complexes. Nature Communications 10, 5346 (2019). 6. Dai, X. et al. Structure and mutagenesis reveal essential capsid protein interactions for

KSHV replication. Nature 553, 521 (2018).

1000 Å

Viral envelope

Virus-like vesicles

Virion

Supplementary Figure 1. A representative cryoEM micrograph containing one EBV virionparticle. Two types of particles can be seen: virion containing a nucleocapsid (marked by a green box,middle) and virus-like vesicles39 (upper left). Zoomed-in in the right panel is the nucleocapsid showingfingerprint-like pattern, characteristic of dsDNA.

4

C6 hexon sub‐particle reconstruction

3Å

7Å

3Å

7Å

C5 portal vertex sub‐particle reconstruction (cut‐away view) 

CATC‐binding penton vertex  sub‐particle reconstruction

3Å

7Å

6Å

10Å

C12 portal vertex sub‐particle reconstruction

CATC‐absent penton vertex sub‐particle reconstruction C1 3‐fold sub‐particle 

reconstruction

3Å

7Å

3Å

7Å

C3 3‐fold sub‐particle reconstruction 

C2 2‐fold sub‐particle reconstruction 

C5 penton sub‐particle reconstruction

3Å

7Å

3Å

7Å

3Å

7Åa

b

FSC = 0.143

0.00.0 0.1 0.2 0.3

0.5

1.0

3.0Å3.4Å

4.0Å4.4Å

5.5 Å

Fourier she

ll correctio

n (FSC) coe

fficie

nt

Resolution (1/Å)

7.8 Å6.7 Å

0.4

C2 2‐fold sub‐particle reconstruction C3 3‐fold sub‐particle reconstruction 

CATC‐absent penton vertex sub‐particle reconstructionCATC‐binding penton vertex sub‐particle reconstructionC5 portal vertex sub‐particle reconstruction  

C6 hexon sub‐particle reconstruction

Icosahedral capsid reconstruction (bin2)C5 capsid reconstruction with portal vertex (bin2)

C5 penton sub‐particle reconstruction C1 3‐fold sub‐particle reconstruction

C12 portal vertex sub‐particle reconstruction  

3.5Å

Supplementary Figure 3. Resolution assessment of 3D reconstructions. (A) Local resolutionevaluation of sub-particle reconstructions by ResMap59. Color bar represents resolution range. (B)“Gold-standard” Fourier shell correlation (FSC) curves of the 3D reconstructions. The resolutions ofthe icosahedral reconstruction and C5 whole virus reconstruction are 5.6 Å and 7.8 Å, respectively;those for the other sub-particles are as follow: 3.0 Å (C6 hexon sub-particle reconstruction), 3.4 Å (C22-fold sub-particle reconstruction), 3.4 Å (C3 3-fold sub-particle reconstruction), 3.5 Å (C5 pentonvertex sub-particle reconstruction), 4.0 Å (CATC-binding sub-particles reconstruction), 4.1 Å (C1 3-fold sub-particles reconstruction), 3.5 Å (CATC-absent penton sub-particle reconstruction) and 4.4 Å(C5 portal vertex sub-particles reconstruction). All resolutions are based on the 0.143 FSC criterion 51.

6

C1C2C3C4

P1P2P3 E3P4

E1E2

PenC5C6

P5P6Pen

P1

P2

P3P4

P5

P6

C1

C2

C3C4

C5

C6E1

E2

E3

Tri1Tri2ATri2B

a.a., 1149-1169MCPa.a., 1256-1267

“up” in Fig. 3, H and I

“down” in Fig. 3, H and I

Un-modeled

E1 E2

E3

E1E2

E3

E1

E2

E3

P1

Pen

PenPen

Pen

P2

P3

P4P5

P6

Ta

Tc

Tb

Te

Tf

Td

Te

MCP within one asymmetric unit

MCP neighboring to one asymmetric unit

Triplex within one asymmetric unit

Triplex neighboring to one asymmetric unit

Supplementary Figure 4. Schematic diagram of plasticity of triplex interactions with MCP buttress domain shown inFigure 3h,i.

7

Tri2ATa

Tc

Tb Td

Te

Tri2BTri2A

Tri2BTri1

Tri2A

P2

Pen1

P1

P6

P5

C6

P4

C1E1

E2 C5

C4

C2

E5

P3

Triplex Ta

Pen1

P6P1Tri1

Tri2BTri2A

Tri1

Tri2B

P6Tri2B

Tri1

Tri2AP1

P1Tri2A

a.a. 1149-1169

a.a. 1256-1267Tri2B

Pen1

a.a. 1256-1267

a.a. 1256-1267

a

b

C6

P5

P2

Tri1

Tri2BTri2A

Triplex Tc

Tri1

Tri1

Tri2B

P2

a.a. 1149-1169

P2

a.a. 1256-1267

c

a.a. 1149-1169missing

Tri2A

P5

C6

Tri2B

a.a. 1256-1267

a.a. 1256-1267

P3

E1

C5

Tri1

Tri2BTri2A

Triplex Tb

dTri1Tri2B

C5Tri1

Tri1

Tri2B

Tri2A

E1

E1

Tri2AP3

Tri2B

a.a. 1149-1169

a.a. 1256-1267

a.a. 1256-1267

a.a. 1149-1169

a.a. 1256-1267

C5

Tri1

Tri2BTri2A

Triplex Td

E5

C1

P4

C1

C1

Tri1

Tri2B

Tri1e

a.a. 1149-1169

a.a. 1256-1267

E5

E5

Tri2A

Tri2B

a.a. 1149-1169

a.a. 1256-1267

P4

P4Tri2A

Tri2A

Tri2B

Tri1 a.a. 1149-1169

a.a. 1256-1267

Triplex Te

Tri1

Tri2BTri2A

C2

E2

C4

E2E2

Tri1

Tri1

Tri2Ba.a. 1149-1169

a.a. 1256-1267

Tri1

Tri2B Tri2A C4

C4Tri2A

a.a. 1256-1267

a.a. 1149-1169

a.a. 1256-1267

Tri2BTri2A

C2

missing

a.a.1149-1169

f

Tri1

Tri2ATri2B

Tri1

Tri2ATri2B

Tri1

Tri1

Tri2B

Supplementary Figure 5. Details of plasticity of triplex interactions with MCP buttress domain shown in Figure3h,i. (A) Distribution of triplexes Ta, Tb, Tc, Td, and Te in the MCP network. (B-F) Magnified profile views of the regionssurrounding triplexes Ta (B), Tb (D), Tc (C), Td (E), and Te (F) as viewed from the outside of the capsid. Away from thecapsid floor, triplex Ta is dominantly stabilized by two long loop-containing segments (a.a. 1149-1169 and a.a. 1256-1267) from the buttress domains of P1 MCP (purple), P6 MCP (orange), and Pen1 MCP (green) (B). Tb (D), Tc (C), Td(E), and Te (F) triplexes are stabilized by their surrounding MCPs in an architecturally similar but locally different manner.

8

SpringerNature_NatMicro_758_ESM.pdfReferencesSfig2-only-largesize.pdfSlide Number 2