nature structural & molecular biology

Article https://doi.org/10.1038/s41594-023-01142-0

Structural insight into the assembly and

working mechanism of helicase-primase D5
from Mpox virus

Received: 4 March 2023 Yaning Li1,2,3,4, Jing Zhu1,4, Yingying Guo 1

& Renhong Yan 1

Accepted: 27 September 2023

Published online: xx xx xxxx The Mpox pandemic, caused by the Mpox virus (or monkeypox virus, MPXV),
has gained global attention. The D5 protein, a putative helicase-primase
Check for updates
found in MPXV, plays a vital role in viral replication and genome uncoating.
Here we determined multiple cryo-EM structures of full-length hexameric
D5 in diverse states. These states were captured during ATP hydrolysis
while moving along the single-stranded DNA (ssDNA) track. Through
comprehensive structural analysis combined with the helicase activity
system, we revealed that when the primase domain is truncated or the
interaction between the primase and helicase domains is disrupted, the
double-stranded DNA (dsDNA) unwinds into ssDNA, suggesting a critical
regulatory role of the primase domain. Two transition states bound with
ssDNA substrate during unwinding reveals that two ATP molecules were
consumed to drive DNA moving forward two nucleotides. Collectively, our
findings shed light on the molecular mechanism that links ATP hydrolysis to
the DNA unwinding in poxviruses.

The Mpox pandemic outbreak has confirmed more than 88,000 cases,
which have been reported in more than 110 countries and territories
since May 2022 (refs. 1–3). With the eradication of the Variola virus,
MPXV has emerged as a notable threat to public health as one of the
virulent orthopoxviruses capable of the human-to-human transmis-
sion through close contact with skin lesions, droplets or contaminated
objects4–6. The World Health Organization declared the MPXV outbreak
as a global health emergency at the end of July 2022, which has raised
concerns regarding the development of therapeutics and vaccines to
combat this virus3,7.
MPXV, belonging to the Poxviridae family, is a kind of large, envel-
oped, double-stranded DNA (dsDNA) virus with the whole life cycle
occurring entirely within the cytoplasm8–11. Similar to other members
in the Poxviridae family, such as the vaccinia virus (VACV), MPXV pos-
sesses its own replication machinery and does not rely on the host
cell’s nucleus for replication, ensuring efficient virus propagation12,13.
In VACV, five essential viral proteins are currently known to assemble
into viral replication apparatus, which contains the DNA polymerase
holoenzyme (E9, the DNA polymerase; A20, the processivity factor; D4,
the uracil-DNA glycosylase), the single-stranded DNA (ssDNA)-binding
protein I3 and the helicase-primase D5, which correspond to F8, A22,
E4, I3 and D5 in MPXV, respectively14–18. Inhibition of the replication
process of poxviruses is a crucial strategy for the treatment of viral
diseases19,20. However, the precise molecular mechanism of assembly
and the specific working stages of the complete replication machinery
from the Poxviridae family has remained poorly understood. In this
Article, we focus on the helicase-primase D5 of MPXV that exhibits
a high protein sequence identity of 99 and 98% to that of VACV and
Variola virus, respectively (Extended Data Fig. 1), which can be used
for the design of broad-spectrum anti-poxvirus drugs21–23.

1
Department of Biochemistry, School of Medicine, Key University Laboratory of Metabolism and Health of Guangdong, Institute for Biological Electron
Microscopy, Southern University of Science and Technology, Shenzhen, China. 2Center for Infectious Disease Research, Westlake Laboratory of Life
Sciences and Biomedicine, Key Laboratory of Structural Biology of Zhejiang Province, School of Life Sciences, Westlake University, Hangzhou, China.
3
Beijing Advanced Innovation Center for Structural Biology, Tsinghua-Peking Joint Center for Life Sciences, School of Life Sciences, Tsinghua University,
Beijing, China. 4These authors contributed equally: Yaning Li, Jing Zhu. e-mail: guoyingnba@163.com; yanrh@sustech.edu.cn

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

a Walker A493–516 Walker B544–559

1 238 323 404 706 785

Primase domain Helicase domain

RRM Zn-binding motif Collar domain AAA+ helicase domain C-terminal motif

b c

e
on
e
800

on

th

th

al
5

ss 785
ng

ng
n
UV280 absorbance (mAU)

78

78

78
al

A
le

le
ei

8–

3–

8–

3–

DN
rk

ll

ll
ot

23

32

23

32
fu

fu
D5

D5

D5

D5

D5

D5
Fo
Pr
600
Protein (µM) 0 0.8 0.8 0.8 0.8 0.8 0.8 0
Fork DNA (µM) 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0
400
ATP + + + + – – – +

200
ADP – – – – + + + –
Lane 1 2 3 4 5 6 7 8
0 Fork DNA
6 8 10 12 14 16 18
Elution volume (ml)
ssDNA
MWM 13 ml 14 ml 15 ml
kDa MWM
130 kDa
100 D5full length
100 D5
70 70
D5238–785
55 D5323–785

d Apo
D5A D5F
ATP
RRM RRM
ADP
D5B

90° 90°
ATP

D5E
D5C
ATP D5B
D5E
ATP D5A D5F
D5D
ssDNA

Fig. 1 | Biochemical characteristics of D5 of MPXV. a, Domain organization of
D5 from MPXV. b, Representative SEC purification of the D5 in milli-absorbance
unit (mAU). SDS–PAGE was visualized using Coomassie blue staining. This
experiment was repeated independently more than three times with similar
results, and the representative result is shown. c, Helicase assay performed with
purified D5 and its truncation proteins. The purified full-length D5 lacks helicase
activity, while truncations D5238–785 and D5323–785 show helicase activity. Three
independent replicates were performed and a representative gel is shown here.
d, The surface presentation of cryo-EM structures of D5 from MPXV is shown
here. Protomers A–F are colored light salmon, orange, yellow, green, blue and
purple, respectively. The RRM that cannot be assigned to protomers A–F is
colored gray. ssDNA is colored red. The ATP and ADP are represented with sticks.

Several temperature-sensitive D5 mutants (for example, Lys509,
Arg619 and Arg620 in ATP binding pocket) in VACV have shown a dra-
matic reduction in DNA synthesis and virus production18,24. An RNA
interference screen of early poxvirus genes identified the D5 as the
virus genome-uncoating factor, whose ATPase activity is essential for
this process25. The open reading frame of the OPG117 gene in the MPXV
genome encodes a 785-amino acid (aa) D5 protein comprising two
major parts: the N-terminal primase domain (residues 1 to 322 aa) and
C-terminal helicase domain (residues 323 to 785 aa)24,26–28 (Fig. 1a). The
primase of D5 originates from the superfamily of archaeo-eukaryotic
primases29, which harbors an RNA recognition motif (RRM) (residues
1 to 237 aa) and a Zn-binding motif (residues 238 to 322 aa)30,31. The
purified D5 of VACV could synthesize ribonucleotide oligomers on a
ssDNA template in vitro, confirming primase activity32. The helicase
domain of D5 comprises core ATPases associated with diverse cellu-
lar activities (AAA+) helicase domain (residues 404 to 705 aa), which
belongs to the helicase superfamily III (SF-III) that contains the Walker A
(residues 493 to 516 aa), Walker B (residues 544 to 559 aa) and Arginine
finger (residues 619 and 620 aa) consensus motifs that are responsible
for nucleoside triphosphate (NTP) binding and hydrolysis. A cysteine
cluster region forming the ring-shape oligomerization motif (residues
323 to 403 aa, also called the collar domain) and a short C-terminal
motif (residues 706 to 785 aa)26,33. The SF-III helicase usually proces-
sively translocates ssDNA in the 3′–5′ direction34–36. The D5 protein of
VACV may be recruited to the replication fork through its interaction
with the processivity factor A20 and this interaction is proposed to be
transient37. The oligomerization of D5 was needed for NTP hydrolysis
and there is no notable difference between ATP, CTP, GTP and UTP
used for the hydrolysis26. For clarity, we will simply use ATP instead of
these NTP because of the similar EM densities of these basic groups.
Although the ATPase activity of D5 is well characterized, the helicase
activity has not been confirmed so far.
Despite extensive research on VACV structures, including some
cryo-EM structures related to E9/A20/D4 polymerase holoenzyme and
some low-resolution structures of D5 (refs. 26,31,38–41), the detailed
working model of D5 protein in DNA unwinding at the replication fork
and the coordination of primer synthesis remains elusive. Here, we
confirm that the primase domain plays a critical regulatory role in the

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

Table 1 | Cryo-EM data collection, refinement and validation statistics

No. 1 ATP-ADP-apo-ssDNA form IS1 No. 2 ATP-ADP-apo-ssDNA form IS2

(EMDB-35051), (PDB 8HWA) (EMDB-35052), (PDB 8HWB)

Data collection and processing

Magnification 81,000
Voltage (kV) 300
Electron exposure (e−/Å2) 50
Defocus range (μm) −1.4–−1.8
Pixel size (Å) 1.095
Symmetry imposed C1
Initial particle images (no.) 1,368,211
Final particle images (no.) 85,148 67,228
Map resolution (Å) 3.7 3.9
FSC threshold 0.143 0.143
Map resolution range (Å) 3.7–5.0 3.9–5.2
Refinement
Initial model used (PDB code) N/A N/A
Model resolution (Å) 3.6 3.6
FSC threshold 0.5 0.5
Model resolution range (Å) 3.6–5.8 3.6–5.8
Map sharpening B factor (Å2) −90 −90
Model composition
Nonhydrogen atoms 25,740 25,740
Protein residues 3,147 3,147
Nucleotides 6 6
Ligands ATP 5; ADP 1; MG 4 ATP 5; ADP 1; MG 4
B factors (Å )
2

Protein 86.76 109.75

Nucleotide 103.33 137.72
Ligand 79.46 93.72
R.m.s. deviations
Bond lengths (Å) 0.003 0.003
Bond angles (°) 0.584 0.640
Validation
MolProbity score 1.66 1.76
Clashscore 7.39 9.79
Poor rotamers (%) 0.35 0.42
Ramachandran plot
Favored (%) 96.23 96.26
Allowed (%) 3.68 3.71
Disallowed (%) 0.09 0.03
No. 3 D5-apo form No. 4 D5-ADP form No. 5 D5-ATP-ADP form
(EMDB-35053), (PDB 8HWC) (EMDB-35054), (PDB 8HWD) (EMDB-35055), (PDB 8HWE)

Data collection and processing

Magnification 81,000 81,000 81,000
Voltage (kV) 300 300 300
Electron exposure (e−/Å2) 50 50 50
Defocus range (μm) −1.4–−1.8 −1.4–−1.8 −1.4–−1.8
Pixel size (Å) 1.095 1.072 1.0773
Symmetry imposed C1 C1 C1
Initial particle images (no.) 2,328,501 2,967,620 551,900

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

Table 1 (continued) | Cryo-EM data collection, refinement and validation statistics

No. 3 D5-apo form No. 4 D5-ADP form No. 5 D5-ATP-ADP form

(EMDB-35053), (PDB 8HWC) (EMDB-35054), (PDB 8HWD) (EMDB-35055), (PDB 8HWE)

Final particle images (no.) 779,993 621,428 321,782

Map resolution (Å) 3.3 3.3 3.3
FSC threshold 0.143 0.143 0.143
Map resolution range (Å) 3.3–4.6 3.3–4.6 3.3–4.6
Refinement
Initial model used (PDB code) N/A N/A N/A
Model resolution (Å) 4.1 4.1 3.7
FSC threshold 0.5 0.5 0.5
Model resolution range (Å) 4.1–5.6 4.1–5.6 3.7–6.0
Map sharpening B factor (Å2) −90 −90 −90
Model composition
Nonhydrogen atoms 14,943 15,029 17,247
Protein residues 1,836 1,834 2,112
Ligands 0 ADP 6 ATP 5; ADP1; MG 5
B factors (Å2)
Protein 65.46 77.08 69.78
Ligand N/A 116.76 77.10
R.m.s. deviations
Bond lengths (Å) 0.007 0.004 0.005
Bond angles (°) 0.732 0.759 0.733
Validation
MolProbity score 1.96 2.15 2.03
Clashscore 10.25 13.09 10.03
Poor rotamers (%) 0.37 0.56 0.36
Ramachandran plot
Favored (%) 93.52 91.56 91.48
Allowed (%) 6.48 8.44 8.52
Disallowed (%) 0.00 0.00 0.00
No. 6 D5-ADP-ssDNA form No. 7 D5-ATP-γS-ADP-ssDNA form No. 8 D5-apo-ssDNA form
(EMDB-35056), (PDB 8HWF) (EMDB-35057), (PDB 8HWG) (EMDB-350558), (PDB 8HWH)

Data collection and processing

Magnification 81,000 81,000 81,000
Voltage (kV) 300 300 300
Electron exposure (e−/Å2) 50 50 50
Defocus range (μm) −1.4–−1.8 −1.4–−1.8 −1.4–−1.8
Pixel size (Å) 1.072 1.095 1.095
Symmetry imposed C1 C1 C1
Initial particle images (no.) 6,149,582 1,108,611 149,303
Final particle images (no.) 60,551 310,318 56,953
Map resolution (Å) 3.3 3 3.6
FSC threshold 0.143 0.143 0.143
Map resolution range (Å) 3.3–4.6 3–4.0 3.6–5.0
Refinement
Initial model used (PDB code) N/A N/A N/A
Model resolution (Å) 3.7 3.1 4.7
FSC threshold 0.5 0.5 0.5
Model resolution range (Å) 3.7–5.2 3.1–4.2 4.7–6.8
Map sharpening B factor (Å2) −90 −90 −90

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

Table 1 (continued) | Cryo-EM data collection, refinement and validation statistics

No. 6 D5-ADP-ssDNA form No. 7 D5-ATP-γS-ADP-ssDNA form No. 8 D5-apo-ssDNA form

(EMDB-35056), (PDB 8HWF) (EMDB-35057), (PDB 8HWG) (EMDB-350558), (PDB 8HWH)

Model composition
Nonhydrogen atoms 18,441 18,110 16,949
Protein residues 2,251 2,209 2,079
Nucleotides 6 6 6
Ligands ADP: 6; MG: 3 AGS: 4; ADP: 2; MG: 6 N/A
B factors (Å2)
Protein 86.01 55.20 124.51
Nucleotide 90.23 72.02 159.31
Ligand 94.85 62.10 N/A
R.m.s. deviations
Bond lengths (Å) 0.006 0.012 0.005
Bond angles (°) 0.696 1.067 0.735
Validation
MolProbity score 1.78 2.32 1.89
Clashscore 8.46 8.38 10.56
Poor rotamers (%) 0.19 4.08 0.26
Ramachandran plot
Favored (%) 95.35 93.94 95.02
Allowed (%) 4.65 6.02 4.98
Disallowed (%) 0.00 0.05 0.00

helicase activity of D5. We also provide the comprehensive structural
characterization of full-length D5 hexamer of MPXV in apo and the
presence of ADP, ATP, ATP analog adenosine 5′-O-(3-thio) triphosphate
(ATP-γS), ssDNA, a mixture of ATP-γS and ssDNA as well as ADP with
ssDNA by single-particle cryo-EM analyses with a resolution ranging
from 3.0 to 3.9 Å.
Results
Structural determination of D5 protein in seven states
We first purified the recombinantly expressed D5 protein from the
MPXV 2022 West African strain (GenBank ID ON563414.3) using human
embryonic kidney 293F (HEK293F) cells. The full-length protein exhib-
ited stability and homogeneity was determined by size-exclusion chro-
matography (SEC) (Fig. 1b). Subsequently, we introduced an in vitro
assay system to assess the helicase activity of D5. Our results revealed
that the full-length D5 protein did not exhibit detectable persistent
helicase activity (Fig. 1c). Given that the D5 protein comprises both
the primase and helicase domains in a single protomer, we hypoth-
esized that the primase domain might regulate the helicase activity.
To test this hypothesis, we truncated the primase domain (D5238–785 and
D5328–785) and, as a result, we observed immediate unwinding of dsDNA
into ssDNA, thus confirming our hypothesis (Fig. 1c).
To gain insights into the working mechanism of D5, we conducted
cryo-EM studies to determine the structures of D5 in various states.
Detailed information regarding cryo-EM sample preparation, data
acquisition, processing and model building can be found in the Meth-
ods, Table 1, Extended Data Figs. 2–5 and Supplementary Table 1. Analy-
sis of the two-dimensional (2D) class average revealed the hexameric
form of D5 (Extended Data Fig. 2a). Initially, the structures of purified
D5 protein were determined at ADP- and ATP-ADP-bound states both
at an overall resolution of 3.3 Å (Extended Data Fig. 3d,e). While the
helicase domain was clearly resolved, the primase domain exhibited
some flexibility (Extended Data Fig. 3d,e).
To obtain a broader range of snapshots and to investigate whether
the primase domain could be stabilized, we further examined the
D5 protein in nucleotide-free (apo) state and in different conforma-
tions bound with various substrates (ATP, ADP and DNA). To trap D5
in the apo state, apyrase, an ATP-hydrolyzing enzyme was used to
degrade endogenously ATP and ADP to AMP, thereby releasing the
substrate from the binding pocket42. Subsequently, we determined the
cryo-EM structures of D5 protein under various conditions: apyrase
treated, supplemented with Mg2+, supplemented with ATP or ATP-γS,
supplemented with complementary dsDNA or fork DNA, alone or
combined. The detailed sample preparation and structural deter-
mination can be found in the Methods section. In total, we success-
fully solved seven cryo-EM structures of the D5 protein under these
conditions, representing the ATP-ADP-apo-ssDNA-bound state
(Fig. 1d), apo state (Fig. 2a), ADP-bound state (Fig. 2b), ATP-ADP-
bound state (Fig. 2c), ADP-ssDNA-bound state (Fig. 2d), ATP-γS-ADP-
ssDNA-bound state (Fig. 2e) and apo-ssDNA-bound state (Fig. 2f).
The resolutions of these structures range from 3.0 to 3.9 Å. In the
ATP-ADP-apo-ssDNA structure, we observed additional electron
densities at the N-terminal region of D5, positioned on top of the
adjacent ring-shaped helicase. These densities corresponded to
three RRMs and two zinc-binding motifs from the primase domain
(Figs. 1d and 3a).
The EM maps were well resolved for most of the helicase domain
in ADP and ATP-ADP state (Fig. 2). In all DNA-bound states, six helical
nucleotides are presented. Because of the strong base stacking, the
individual density of basic groups is not so clear, thus we build them
as oligo dT. The direction of ssDNA is defined according to the density
of the 3′ deoxyribose group and 5′ phosphate group (Extended Data
Fig. 5b). Besides, the C-terminal motif and some primase regions are
not good enough for model building. The low threshold indicated
the C-terminal motif might participate in the intersubunit interac-
tion (Extended Data Fig. 5f). The resolved regions of D5 in different

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

a Apo form

Apo Apo

Apo Apo

90° 90°

Apo Apo

b ADP form
ADP ADP

ADP
ADP

90° 90°

ADP ADP

c ATP-ADP form
ADP
ATP

ATP
ATP
90° 90°

ATP
ATP

d ADP-ssDNA form
ADP
ADP

ADP ADP

90° 90°

ADP
ADP

e ATPγS-ADP-ssDNA form
ATPγS ADP

ADP

ATPγS 90° 90°

ATPγS
ATPγS

f Apo-ssDNA form
Apo Apo

Apo
Apo 90° 90°

Apo Apo

Apo form ADP form ATP form ATP-γS form ssDNA

Fig. 2 | Cryo-EM analysis of D5 in different ligands and DNA-binding form. yellow respectively. ssDNA is colored red. a–f, The apo form (a), ADP form (b),
From left to right is the surface styled model of D5 complex viewed from the ATP-ADP form (c), ADP-ssDNA form (d), ATP-γS-ADP-ssDNA form (e) and apo-
bottom and top, alongside the corresponding cryo-EM map. The protomers in ssDNA form (f).
apo form or binding with ADP, ATP, ATP-γS are colored purple, blue, green and

structures are described in Supplementary Table 1. For clarity, the Overall, the structure of D5 could be clearly divided into four dis-
six subunits of D5 are defined as D5A–D5F protomers, respectively tinct layers in the ATP-ADP-apo-ssDNA state (Extended Data Fig. 6a).
(Fig. 1d). The first layer consists of the primase domains, followed by the collar

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

D5F Zn-binding motif

D5A RRM D5F RRM

90°

D5B
RRM

RRM
D5A Zn-binding
D5E motif
D5A D5F

b
RRM
Regulated
motif
D88

Collar domain
R387 N163
D5F E79

R128
D5D

E382

D5A

D5C

D5B

c
D5full length (apyrase)

D5full length (apyrase)

D5238–785 (apyrase)

D5238–785 (apyrase)
E79K/D88K/R128E

E79K/D88K/R128E
(apyrase)

(apyrase)

D5full length (apyrase)

ssDNA alone
Fork alone

D5238–785 (apyrase)
Protein

E79K/D88K/R128E
(apyrase)
NS NS
Protein (µM) 0 0.8 0.8 0.8 0.8 0.8 0.8 0
Fork DNA (µM) 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0 30

ADP – – – – + + + –
ssDNA ratio (%)

Fork DNA 20

10
ssDNA
MWM
kDa 0
100 D5full length
– ADP + ADP
70
D5238–785

Fig. 3 | Structural analysis of N-terminal domain of D5. a, A cartoon mutant and primase domain truncated form of the D5 protein. The helicase
presentation of the cryo-EM structures of D5. Protomers A–F are colored light activity comparison of full-length D5, E79K/D88K/R128E mutant and N-terminal
salmon, orange, yellow, green, blue and purple, respectively. The insets show truncations D5238–785 shows that the N-terminal truncations dramatically increase
a cartoon presentation of domain-colored cryo-EM structures of N-terminal the helicase ability, and the E79K/D88K/R128E mutant presents almost the same
primase domain that are visible. RRM of protomers A and F are colored red and helicase activity as D5238–785. Three replications were taken independently and
dark magenta. The RRM that cannot be assigned to protomers A–F is colored the ssDNA was delivered by quantitative analysis. Three replications were taken
gray. Zn-binding motifs of protomers A and F are colored light salmon and independently for data analysis and the data are presented as mean ± standard
purple. b, The regulated motif (dark red) in protomer A blocks the ssDNA exit site deviation. NS means no significance. For multiple comparisons, P values were
shaped by six collar domains colored light blue. The inset shows the interaction derived from ordinary one-way analysis of variance with Šídák’s multiple
network of the regulated motif. The residues outlined with the box and shaded comparisons test. The default parament settings were applied to multiple
in red, yellow and green are of the D5A, D5C and D5D protomers, respectively. comparisons. P < 0.05 (two sided) was considered significant. Graphs were
c, Helicase activity of the apyrase-digested full-length, E79K/D88K/R128E prepared in GraphPad prism (v.9.0).

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

r.m.s.d. 0.823
ATP bound ADP bound Apo

G508
D5A R514
D5A
T511
L655 K509
Mg
D656
R620
ATP
I464
D5B T507
ATP bound 90° A506 D652
ADP bound D467 R619
F630

D5B
ATP

L655
G508
R514 D5A
D656 D5A
K509 T511

R620

D5B
90° A506
D652
R619
T507
D5B ADP
ADP bound
Apo

D5A
D5A
G508 R514
D5B
K509
r.m.s.d. 5.837 I464
R620
L655 T511
T507
F630 90°
D5B A506
D467
R619
D652

Fig. 4 | Comparison of D5 bound to ATP, ADP or no ligands. The structural
comparison shows there are no conformation change among three collar
domains and no difference between AAA+ helicase domains D5 tightly bound
to ATP and ADP, but dramatic shifts in AAA+ helicase domain in apo form.
ATP-bound, ADP-bound and apo form are colored green, blue and purple,
respectively. An r.m.s.d. between 378 Cα pairs (323 to 700 aa) of the chain bound
ATP and ADP is 0.823 Å and that between 378 Cα pairs (323 to 700 aa) of the chain
bound apo and ADP is 5.837 Å. In the insets, the top panel shows the interaction
network of ATP in D5-ATP form, the middle panel shows the interaction network
of ADP in D5-ADP tight form and the bottom panel shows ATP or ADP-binding
pocket in apo form.

domain. The third layer comprises an asymmetric spiral staircase and
the ATPase domain, while the fourth layer corresponds to the short
C-terminal domain. Six protomers of D5 are arranged in a circular
manner, forming stacked rings within the structure. This arrangement
creates a positively charged central funnel-shaped chamber, which is
crucial for DNA binding (Fig. 1d). The structure provides a clear depic-
tion of how the six putative helicase subunits in the C terminus of D5
assemble into these stacked rings (Fig. 1d).
The assembly of primase
The primase activity of the N-terminal of the D5 protein, which includes
the RRM and a Zn-binding subdomain, has been extensively studied
and characterized30. In the ATP-ADP-apo-ssDNA intermediate state,
we identified three protomers that encompass the primase domain
(Fig. 3a). The overall structure of the primase domain exhibits a cata-
lytic palm structural core and shares similarities with the RRM fold of
the archaeo-eukaryotic primase superfamily (Fig. 3a,b). Within the
primase domain, a regulated motif spanning amino acids 76 to 130, as
well as an adjacent Zn-binding motif, potentially obstruct the channel
enclosed by the collar domain (Fig. 3b). Notably, Glu79 and Arg128
of the regulated motif form salt-bridge interactions with Arg387 and
Glu382 of the collar domain, respectively. Additionally, Asp88 likely
contributes to stabilizing the regulated motif by interacting with the
RRM. Hence, these three residues (Glu79, Asp88 and Arg128) are pro-
posed to play a critical role in the interaction between primase and
helicase domains (Fig. 3b).
The primase domain of D5 closely resembles PrimPol, a human
DNA polymerase with primase activity, displaying root mean square
deviation (r.m.s.d.) values as 3.4 Å, as determined through align-
ment using the Dali server43. When comparing the structure of D5
with that of PrimPol, an elongated density is observed in the nucle-
oside binding site of D5, which can be well fitted with an ATP mol-
ecule (Extended Data Fig. 6b)44. In contrast to PrimPol, the current
primase domains of D5 do not enclose a dsDNA substrate, suggest-
ing an open conformation (Extended Data Fig. 6b). Notably, despite
solving three out of the six primase domains in the D5 hexamer, the
remaining primase domains may show flexibility. Based on the archi-
tecture of primase domain, it is plausible to suggest that RNA primers
could be synthesized independently for the initiation of poxvirus
DNA replication.

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

a e Aligned N-terminal
c
D5AIS1 primase domain
D5FIS1 ssDNAIS2 D5AIS2
ssDNAIS1
D5BIS2 D5FIS2
D5BIS1 r.m.s.d. 4.768

90° 90°

D5EIS1 D5EIS2
D5CIS1 D5CIS2

D5DIS1
D5DIS2
D5 IS1 D5 IS2 90°

D5 IS1 D5 IS2
b f d
βB-hairpin βA-hairpin
R585

R585 R585
R585 F588 βF-hairpin ssDNAIS2
F588
F588
F588
F588 F588
R585 R585 R585
F588 R585
F588
ssDNAIS1
F588 F588
βc-hairpin F588 F588
120° R585
R585 R585
R585
βD-hairpin
βE-hairpin

90° 180° 90° 180°

F588
F588
3’ F588
R585 3’

F588 F588
R585
R585 R585
F588
F588 R585 R585

F588
F588
F588
R585
R585 F588 R585
F588
ssDNAIS1 ssDNAIS2
R585 R585 5’
5’ R585

Fig. 5 | The structural comparison of two intermediate states of D5 in ATP-
ADP-apo-ssDNA. a, The cartoon presentation of cryo-EM structures of D5 IS1.
Protomers A–F are colored light salmon, orange, yellow, green, blue and purple,
respectively. ssDNA is colored red. b, The inset shows the DNA-binding network
in IS1. c, The cartoon presentation of cryo-EM structures of D5 IS2. Protomers
A–F are colored light salmon, orange, yellow, green, blue and purple,
respectively. ssDNA is colored blue. d, The inset shows the DNA-binding
network in IS2. e, When we aligned the N-terminal primase domain of IS1 and
IS2, ssDNAs bound to different protomers. An r.m.s.d. between 3,139 Cα pairs of
D5 IS1 and D5 IS2 is 4.768 Å. f, D5 IS1 are colored pink, orange and red. D5 IS2 are
colored light blue, orange and blue. The top panel is the bottom view of e and the
bottom panel shows that when the IS1 structure is rotated 120° counterclockwise,
the ssDNA of IS1 (red) and IS2 (blue) can overlap.

To investigate the influence of ATP and ADP binding on the helicase
activity of D5, we used apyrase to degrade the γ and β phosphate groups
of ATP or ADP molecules, effectively removing endogenous ATP or ADP
in the purified D5 protein and its mutants. The treated protein samples
were labeled as ‘protein name (apyrase)’ for convenience. We observed
a faint ssDNA band in the lane of D5full-length (apyrase) when ATP and fork
dsDNA were added (Extended Data Fig. 7a,e) but not with ADP and fork
dsDNA (Extended Data Fig. 7b,e). This suggests that full-length D5 exhibits
weak helicase activity after apyrase treatment. This finding contrasts
with the results obtained without apyrase treatment, where we observed
helicase activity only when the primase domain was removed (Fig. 1c).
Since ATP and ADP can bind to the primase domain and helicase domain
simultaneously, and we could only locate the primase domain in the
ATP-ADP-apo-ssDNA state, we speculate that the apo state and the ATP and/
or ADP loading state may lead to different organization patterns between
the primase domain and helicase domain, which in turn affects the
helicase activity.
To test this hypothesis, we incubated the apyrase-digested D5
protein with ADP first, and then added fork dsDNA and ATP to initial
helicase activity (Extended Data Fig. 7c,e), the D5full-length (apyrase)
sample showed very low helicase activity in this condition. We also
attempted to added fork dsDNA first, followed by ADP and finally ATP

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

D5 C terminus
ATP
D C
E ADP

F B Arg finger
A
F588
D5 N terminus Fork substrates

Apo state
D
E
ATP binding
F C
3’
B
5’ A
Apo-ssDNA state
ADP state ATP-ADP state
D
E
E D E D
C
F
ATP binding F F
B C C
3’
A
5’
5’ A B A B
D
E

D C
E 3’ F

C A B
F

A B 5’
Translocation cycle

3’ 3’

5’ ADP-ssDNA state
3’ 3’

ATP-ADP-ssDNA state 5’
D E
...
E
D
ATP hydrolysis
F
C
F
...
ATP hydrolysis C
A
A B 5’

B
5’

3’
3’
IS1 IS2
ATP-ADP-Apo-ssDNA state
Fig. 6 | A proposed mechanism of ssDNA translocation from D5. Schematic purple). The basic elements, ATP, ADP, Arg finger (arginine finger) and F588 (the
of the MPXV D5 helicase domain dynamic revolving with the ssDNA substrate. phenylalanine residue at the 588 amino acid position) of helicase are illustrated in
Six protomers are shown in different colors (red, orange, yellow, green, blue and the top panel. The red arrow indicates the proposed direction of DNA translocation.

to initial the helicase activity (Extended Data Fig. 7d,e), In this case,
the D5full-length (apyrase) sample exhibited higher helicase activity than
the previous condition but was still lower than the truncation form
(D5238–785). These results indicated a partial regulatory role of ADP and
ATP in the helicase activity of D5full-length, although they may not play a
dominant role.
To investigate the influence of specific residues involved in the
interaction between the primase and helicase domains, we introduced a
triple point mutation (E79K/D88K/R128E) in the D5 protein. These resi-
dues were previously identified as critical for the interaction between
the primase and helicase domains (Fig. 3b). The results showed that the
triple mutation (E79K/D88K/R128E) remarkably enhanced the helicase
activity of D5, similar to the primase truncation form (Fig. 3c). This sug-
gests that the precise spatial placement of the primase domain affects
the unwinding function of the D5 protein.
Distinct nucleotide-binding states
In the presented structures of the D5 protein, we have observed
three different nucleotide states in the ATP/ADP-binding pocket of
all protomers: ATP-bound, ADP-bound and apo-like status (Figs. 2
and 4 and Extended Data Fig. 5). Notably, all three states are simul-
taneously present in ATP-ADP-apo-ssDNA states (Figs. 1d and 2c–e).
Among the protomers, protomers A to D exhibit strong ATP-like densi-
ties in the nucleotide-binding pocket, while protomer E shows ADP-like

Nature Structural & Molecular Biology

Article
density, and protomer F adopts an apo-like status (Extended Data
Fig. 5c–e). This 4ATP-1ADP-1apo binding mode of D5 differs notably
from other SF-III members, such as E1 of papillomavirus, which exhibits
a 3ATP-2ADP-1apo binding pattern45. Furthermore, in the structure of
the ADP-bound state, all nucleotide-binding pockets from six protom-
ers have well-resolved ADP molecules. In line with expectations, we did
not find nucleotide-like density in any chain of the apyrase-treated apo
state (Figs. 2a and 4).
The ATP- and ADP-binding pockets are located in the middle of
the AAA+ helicase domain. The adenine ring, ribose moiety and α, β
phosphate of ATP and ADP are associated with the nucleotide-binding
pocket in a similar manner. The Walker A motif, Walker B motif and an
associated Mg2+ ion are involved in the interaction with ATP and ADP
(Fig. 4). The adenine ring of ATP is stacked with Phe630 and interacts
with Asp467 through a π–π interaction and hydrogen bond interac-
tion, respectively. The ribose moiety is stabilized by the main chain
of Asp652 and the side chain of Asp656. Thr507, Gly508 and Thr511
are involved in α phosphate binding, and β phosphate interacts with
Ala506, Thr507 and Lys509. The γ phosphate is coordinated by Mg2+
and the arginine finger motif (Arg619 and Arg620) is located in an
adjacent protomer (Fig. 4). When the ATP molecule is hydrolyzed, the
γ phosphate is released, and the arginine finger is no longer in con-
tact with ADP, similar to the apo status. We refer to this ADP-binding
mode as ‘ADP loose form’. These interactions are confirmed by the
ATPase activity system with the mutant R619/R620 and K509A/T511A/
R514A/F630A/R656A (M5t mutant) that harbor notably reduced ATP
hydrolysis activity (Extended Data Fig. 8a). Besides, ATP-γS could also
significantly inhibit the ATP hydrolysis activity (Extended Data Fig. 8a).
This binding pattern might directly lead to conformational changes of
two neighboring protomers.
An asymmetric spiral staircase engages the ssDNA substrate
The DNA-binding properties are revealed in the structures of
ATP-ADP-apo-ssDNA (Fig. 5a,c), ADP-ssDNA (Fig. 2d), ATP-ADP-ssDNA
(Fig. 2e) and apo-ssDNA (Fig. 2f). As the visible parts of the ssDNA
substrates in these states are very similar to each other, the
ATP-ADP-apo-ssDNA form is chosen as a representative to describe the
DNA-binding pattern. Owing to the flexibility of the ssDNA substrate,
only the six nucleotides in the ssDNA substrate binding to protein are
visible. These six nucleotides form a helical array and interact with
five β-hairpins located in the third stack layer of the hexameric heli-
case domain from D5A to D5E (βA-hairpin to βE-hairpin) (Fig. 5b). The
βA-hairpin to βE-hairpin form a spiral-like arrangement from high to
low in turn, and the loosely βF-hairpin is located at the opposite side
of the ssDNA interface (Fig. 5b). The staircase residues Phe588 from
βA-hairpin to βE-hairpin play a crucial role in binding with the basic
groups of ssDNA. Arg585 also makes a notable contribution to binding
with the phosphate backbone of ssDNA (Fig. 5b). Mutations in these
key residues, such as R585A and F588A, lead to a notable decrease in
DNA-binding activity (Extended Data Fig. 8b).
Notably, two distinct intermediate states are further classified
from the particles in the ATP-ADP-apo-ssDNA state data set. On aligning
the N-terminal primase domains, we observed a notable movement of
the ssDNA in these two intermediate states (Fig. 5a,c–f). For simplicity,
we refer to these states as IS1 and IS2. When IS1 is rotated approximately
120° in an anticlockwise direction, both the ssDNA and the helicase
domain align closely with those in IS2 (Fig. 5f). However, there remains
a question, which of these intermediate states appear early during the
process of DNA unwinding?
Fortunately, we also capture an ATP-γS-ADP-ssDNA form, in which
four ATP-γS and two ADP molecules were bound (Fig. 2e). Although
the density of the N-terminal primase domain in this form is weaker
compared to the ATP-ADP-apo-ssDNA states, it is still visible when the
map is viewed at a low threshold (Extended Data Fig. 9a). The DNA trans-
location process appears to be slowed down by the slower hydrolysis
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
of ATP-γS, as the ADP in D5F has not been released. By aligning the
N-terminal primase domain in ATP-γS-ADP-ssDNA states with that in IS1
(Extended Data Fig. 9b) and IS2 (Extended Data Fig. 9c), we found that
IS1 is fitted well with ATP-γS-ADP-ssDNA form. This suggests that IS1
occurs earlier than IS2 during DNA unwinding. For DNA translocation,
β-hairpins C to F are raised to take the position of β-hairpins A to D, and
β-hairpins A and B fall to bind new nucleotides after releasing the now
being binding nucleotides (Fig. 5b,d). The movements of β-hairpins
result in the 3′–5′ direction of right-handed spiral DNA unwinding,
which is consistent with a previous study36. Two loose protomers rotat-
ing counterclockwise for 120° consume two ATP molecules and move
two nucleotides forward in the 5′ direction.
We further conducted a comparison of each protomer and
observed that the collar domains remained relatively unchanged when
these structures were superimposed (Extended Data Fig. 10a,b). We
noticed that the inclined angles of the AAA+ helicase domain varied
among the ATP, ADP and apo nucleotide-binding states (Extended
Data Fig. 10a,b). In the ADP-bound structure of D5, two ADP mole-
cules are presented in the ADP loose form whereas the remaining four
ADP-bound protomers acquire a smaller inclined angle, which might
represent another working state of D5, so we called this the ‘ADP tight
form’ (Fig. 4). The apo form of D5 acquires the largest inclined angle,
once the ATP/ADP or ssDNA binding occurs, so it could trigger the
smaller inclined angle (Extended Data Fig. 10a). Based on these obser-
vations, we propose that the tilt angle changes of the AAA+ helicase
protomers are correlated with distinct nucleotide-binding states and
are supported by a collar pivot, which in turn affects the positions of
the β-hairpins in this protomer36.
The putative working mechanism of D5 helicase
The helicases, such as E1 of the papillomavirus, the large tumor
antigen of simian virus 40 and Rep40 of the adeno-associated virus
belonging to the SF-III family AAA+ helicase, all processively trans-
locate in the 3′–5′ direction of the DNA substrate36,46–48. A proposed
model, known as the concerted escort translocation model, explains
how the hexameric helicase threads the DNA substrate through the
spiral staircase formed by the central channel β-hairpins. This model
accounts for the events of ssDNA translocation, ATP binding, hydroly-
sis and ADP release, which are coupled to conformational changes
within the helicase36.
When aligned with other members of the SF-III family, the heli-
case of D5 shows a closer resemblance to papillomavirus E1, both of
which share common features such as the collar domain and AAA+
helicase domain46 (Extended Data Fig. 10d). However, the detailed
working mechanism may differ remarkably. Three ATP hydrolysis
models have been proposed for hydrolytic cycling, including the con-
certed hydrolysis model (all six subunits bind, hydrolyze and release
ATP in synchrony), stochastic hydrolysis model (ATP hydrolysis at
the random site) and rotary hydrolysis model (ATP hydrolysis in a
sequential order around the collar domain)36. Similar to E1, D5 dis-
played a series of ATPase states that progress from ATP- to ADP- and
then apo-like configurations in sequential order, supporting the rotary
mechanism. However, the 4ATP-1ADP-1apo binding mode observed
in D5 is distinct from the 3ATP-2ADP-1apo binding pattern seen in
E1 (ref. 35).
More importantly, we captured two ATP-ADP-apo-ssDNA inter-
mediate states, which represent two static snapshots of processive
movements along ssDNA. When bound to any substrates (ATP, ADP,
ssDNA or nucleotide-ssDNA), a striking shared structural feature of
ring-shaped D5 is that four protomers are tightly connected with each
other, regarded as ‘tight form’. Among ring-shaped hexamer of D5,
there are always two protomers that weakly interacting with adjacent
protomers, which are called ‘loose form’ here. The collar domain acts
as a fulcrum allowing the AAA+ helicase domain to produce tilt angles
depending on the rotation of the tight or loose form (Fig. 6).
Article
Based on the findings and observations, a working model for the
D5 helicase can be proposed (Fig. 6). In the apo state, where D5 is not
associated with a DNA substrate, there are six loose forms of protom-
ers. This state allows ATP molecules to enter the ATP binding pocket
of each protomer. The ATP hydrolysis controls the movement of the
β-hairpins. When ssDNA binds with the β-hairpins in the central chan-
nel of the ring, two ATP molecules in four tight protomers undergo
hydrolysis, while the remaining two loose protomers release ADP and
reacquire ATP molecules. This step allows D5 to catch the ssDNA in one
working step. In this model, it is proposed that the two loose protomers
rotate counterclockwise with each step during ssDNA translocation,
coupled with the hydrolysis of two ATP molecules. This movement
enables the helicase to advance the ssDNA by two nucleotides in
the 5′ direction.
Discussion
The genome of MPXV consists of a single, contiguous dsDNA chain
that forms a linear duplex with covalently closed hairpin termini49,50.
During the infection process, the expression of the genes in poxviruses
is tightly controlled and can be classified into the early and late gene
expression stages51,52. D5 protein is expressed at an early stage, with a
peak at 2 to 3 hours after infection, and declines to be undetectable after
5 hours, coinciding with the onset of viral DNA synthesis28,53. Whether
the natural activity of D5 protein is timely regulated remains unclear.
In VACV, low-resolution structures of the D5 protein have revealed
its hexameric assembly state. However, the specific binding modes
of DNA and ATP, as well as the mechanism of ssDNA translocation,
remain unknown26,31.
Our study has successfully revealed the inhibitory effect of the
primase domain on the helicase activity of D5, supported by both
biochemical and structural evidence. Blocking the ssDNA exit from the
collar domain might play an important regulatory role in the persistent
helicase activity of D5. Through the analysis of D5full-length (apyrase)
and its re-acquisition of helicase activity, we have speculated that the
competition between ADP and ATP binding may influence the blockage
of the C-terminal helicase channel by the primase domains. To further
investigate this hypothesis, we altered the order of reactant addition
in our helicase activity assays. We observed improved unwinding effi-
ciency when fork dsDNA was added before ADP and ATP (Extended Data
Fig. 7c–e). We postulate that this improvement is due to the entry of
fork dsDNA into the cavity, which occupies the space and prevents the
binding of the primase domain to the collar domain through a confor-
mational change induced by ADP binding. However, it is important to
note that despite the partial regulatory role of ADP or ATP on the heli-
case activity of D5full-length, the truncation form (D5238–785) still exhibited
higher helicase activity under these assay conditions. This suggests
that factors other than ADP or ATP binding may also contribute to
the overall helicase activity of full-length D5. Furthermore, the triple
mutation (E79K/D88K/R128E) that disrupts the interaction between the
primase and helicase domains notably enhanced the helicase activity
of D5, underscoring the importance of the primase–helicase domain
interaction in regulating D5′s helicase activity.
Notably, the primase domain of D5 may be subject to regulation by
other components of the DNA replication machinery within the host
cell environment. For example, in VACV, the processivity factor A20 has
been implicated in the regulation of DNA replication and may control
the inhibitory effect of the primase domain on the helicase activity of
D5. Additionally, there may be other unidentified factors involved in
modulating the helicase activity of D5 by interacting with the primase
domain. Further investigation is needed to identify and characterize
these potential regulatory factors.
Hutin et al. resolved a low-resolution (3.9 Å) structure of D5 in
VACV and built a model using Alphafold2 prediction, without the clear
density of the N-terminal primase domain31. Moreover, no information
about ATP/ADP-binding mode could be obtained from their model. By
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
contrast, we captured seven distinctive transition-state intermediates
of the hexamer in apo and ATP/ADP substrate binding states or during
processive translocation along ssDNA. And our structures are good
enough to assign the ssDNA rather than dsDNA to the central channel.
These structures reveal how D5 forms a homo-hexameric ring-shaped
AAA+ helicase, using a spiral staircase encircling an ssDNA substrate
and coupling ATP hydrolysis energy to highly coordinated motions.
We provided solid evidence that two ATP molecules were consumed to
drive DNA to move forward by two nucleotides, which could simply be
proposed by the energetic consumption associated with spontaneous
base pair unwinding, limiting helicase movement to a maximum step
size of one or two bases in each NTP turnover36.
Both the D5 protein of poxvirus and E1 of papillomavirus make use
of the pore loops to lever up and down depending on the nucleotide
status of their associated protomers for ssDNA translocation36,45,54.
However, the ATP- and/or ADP-binding mode and the ssDNA direc-
tions in E1 (5′ toward the collar domain) and D5 (3′ toward the collar
domain) structures are different, suggesting a possible different work-
ing mechanism of the D5 protein. Translocation by E1 might depend
on a coordinated escort mechanism in which a loop dives toward the
5′ DNA end, engages a base and then moves upward, remaining associ-
ated with that base until it is released at the end of the ATPase cycle for
that subunit. However, in the ATP-ADP-apo-ssDNA intermediate states,
every turnover might consume two ATP molecules, which drives the
DNA forward by two nucleotides. Subsequently, the timely coupling of
primase and helicase in D5 might accelerate the synthesis of new viral
DNA. In conclusion, our findings represent a remarkable advancement
in our understanding of the structure and function of the poxvirus
helicase-primase.
Online content
Any methods, additional references, Nature Portfolio reporting sum-
maries, source data, extended data, supplementary information,
acknowledgements, peer review information; details of author contri-
butions and competing interests; and statements of data and code avail-
ability are available at https://doi.org/10.1038/s41594-023-01142-0.
References
1. Lum, F. M. et al. Monkeypox: disease epidemiology, host
immunity and clinical interventions. Nat. Rev. Immunol. 22,
597–613 (2022).
2. Kumar, N. et al. The 2022 outbreak and the pathobiology of the
monkeypox virus. J. Autoimmun. 131, 102855 (2022).
3. Liang, C., Qian, J. & Liu, L. Biological characteristics, biosafety
prevention and control strategies for the 2022 multi-country
outbreak of monkeypox. Biosaf. Health 4, 376–385 (2022).
4. Lewis-Jones, S. Zoonotic poxvirus infections in humans.
Curr. Opin. Infect. Dis. 17, 81–89 (2004).
5. Farasani, A. Monkeypox virus: future role in Human population.
J. Infect. Public Health 15, 1270–1275 (2022).
6. Henderson, D. A. et al. Smallpox as a biological weapon -
medical and public health management. J. Am. Med Assoc. 281,
2127–2137 (1999).
7. Tiecco, G. et al. Monkeypox, a literature review: what is new
and where does this concerning virus come from? Viruses 14,
1894 (2022).
8. Cho, C. T. & Wenner, H. A. Monkeypox virus. Bacteriol. Rev. 37,
1–18 (1973).
9. McFadden, G. Poxvirus tropism. Nat. Rev. Microbiol. 3,
201–213 (2005).
10. Pennington, T. H. & Follett, E. A. C. Vaccinia virus replication in
enucleate Bsc-1 cells: particle production and synthesis of viral
DNA and proteins. J. Virol. 13, 488–493 (1974).
11. Yang, Z. & Moss, B. Decoding poxvirus genome. Oncotarget 6,
28513–28514 (2015).
Article
12. Moss, B. Poxvirus DNA replication. Cold Spring Harb. Perspect.
Biol. 5, a010199 (2013).
13. Greseth, M. D. & Traktman, P. The life cycle of the vaccinia virus
genome. Annu Rev. Virol. 9, 239–259 (2022).
14. Stanitsa, E. S., Arps, L. & Traktman, P. Vaccinia virus uracil
DNA glycosylase interacts with the A20 protein to form a
heterodimeric processivity factor for the viral DNA polymerase.
J. Biol. Chem. 281, 3439–3451 (2006).
15. Davis, R. E. & Mathews, C. K. Acidic C terminus of vaccinia virus
DNA-binding protein interacts with ribonucleotide reductase.
Proc. Natl Acad. Sci. USA 90, 745–749 (1993).
16. McDonald, W. F., Klemperer, N. & Traktman, P. Characterization of
a processive form of the vaccinia virus DNA polymerase. Virology
234, 168–175 (1997).
17. Rochester, S. C. & Traktman, P. Characterization of the single-
stranded DNA binding protein encoded by the vaccinia virus I3
gene. J. Virol. 72, 2917–2926 (1998).
18. Roseman, N. A. & Hruby, D. E. Nucleotide sequence and transcript
organization of a region of the vaccinia virus genome which
encodes a constitutively expressed gene required for DNA
replication. J. Virol. 61, 1398–1406 (1987).
19. Baker, R. O., Bray, M. & Huggins, J. W. Potential antiviral
therapeutics for smallpox, monkeypox and other orthopoxvirus
infections. Antivir. Res 57, 13–23 (2003).
20. De Clercq, E. Cidofovir in the treatment of poxvirus infections.
Antivir. Res 55, 1–13 (2002).
21. Massung, R. F. et al. Analysis of the complete genome of
smallpox variola major virus-strain Bangladesh-1975. Virology 201,
215–240 (1994).
22. Isidro, J. et al. Phylogenomic characterization and signs of
microevolution in the 2022 multi-country outbreak of monkeypox
virus. Nat. Med. 28, 1569–1572 (2022).
23. Kannan, S. R. et al. Mutations in the monkeypox virus replication
complex: potential contributing factors to the 2022 outbreak.
J. Autoimmun. 133, 102928 (2022).
24. Boyle, K. A., Arps, L. & Traktman, P. Biochemical and genetic
analysis of the vaccinia virus D5 protein: multimerization-
dependent ATPase activity is required to support viral DNA
replication. J. Virol. 81, 844–859 (2007).
25. Kilcher, S. et al. siRNA screen of early poxvirus genes identifies
the AAA+ ATPase D5 as the virus genome-uncoating factor.
Cell Host Microbe 15, 103–112 (2014).
26. Hutin, S. et al. Domain organization of vaccinia virus helicase-
primase D5. J. Virol. 90, 4604–4613 (2016).
27. Evans, E. et al. The vaccinia virus D5 protein, which is
required for DNA replication, is a nucleic acid-independent
nucleoside triphosphatase. J. Virol. 69, 5353–5361 (1995).
28. Evans, E. & Traktman, P. Molecular genetic-analysis of a vaccinia
virus gene with an essential role in DNA-replication. J. Virol. 61,
3152–3162 (1987).
29. Iyer, L. M. et al. Origin and evolution of the archaeo-eukaryotic
primase superfamily and related palm-domain proteins:
structural insights and new members. Nucleic Acids Res. 33,
3875–3896 (2005).
30. De Silva, F. S. et al. Poxvirus DNA primase. Proc. Natl Acad. Sci.
USA 104, 18724–18729 (2007).
31. Hutin, S. et al. The vaccinia virus DNA helicase structure from
combined single-particle cryo-electron microscopy and
AlphaFold2 prediction. Viruses 14, 2206 (2022).
32. De Silva, F. S., Paran, N. & Moss, B. Products and substrate/
template usage of vaccinia virus DNA primase. Virology 383,
136–141 (2009).
33. Singleton, M. R., Dillingham, M. S. & Wigley, D. B. Structure
and mechanism of helicases and nucleic acid translocases.
Annu. Rev. Biochem. 76, 23–50 (2007).
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
34. Goetz, G. S., Dean, F. B., Hurwitz, J. & Matson, S. W. The unwinding
of duplex regions in DNA by the simian virus-40 large tumor
antigen-associated DNA helicase activity. J. Biol. Chem. 263,
383–392 (1988).
35. Medagli, B. & Onesti, S. Structure and mechanism of hexameric
helicases. DNA Helicases DNA Mot. Proteins 973, 75–95 (2013).
36. Fernandez, A. J. & Berger, J. M. Mechanisms of hexameric
helicases. Crit. Rev. Biochem. Mol. 56, 621–639 (2021).
37. McCraith, S. et al. Genome-wide analysis of vaccinia virus
protein-protein interactions. Proc. Natl Acad. Sci. USA 97,
4879–4884 (2000).
38. Contesto-Richefeu, C. et al. Crystal structure of the vaccinia
virus DNA polymerase holoenzyme subunit D4 in complex
with the A20 N-terminal domain. PLoS Pathog. 10, e1003978
(2014).
39. Tarbouriech, N. et al. The vaccinia virus DNA polymerase structure
provides insights into the mode of processivity factor binding.
Nat. Commun. 8, 1455 (2017).
40. Burmeister, W. P. et al. Crystal structure of the vaccinia virus
uracil-DNA glycosylase in complex with DNA. J. Biol. Chem. 290,
17923–17934 (2015).
41. Bersch, B. et al. Solution structure of the C-terminal domain of
A20, the missing brick for the characterization of the interface
between vaccinia virus DNA polymerase and its processivity
factor. J. Mol. Biol. 433, 167009 (2021).
42. Agah, A. et al. A multi-enzyme model for pyrosequencing. Nucleic
Acids Res. 32, e166 (2004).
43. Holm, L. DALI and the persistence of protein shape. Protein Sci.
29, 128–140 (2020).
44. Baranovskiy, A. G. et al. Crystal structure of the human primase.
J. Biol. Chem. 290, 5635–5646 (2015).
45. Enemark, E. J. & Joshua-Tor, L. Mechanism of DNA translocation
in a replicative hexameric helicase. Nature 442, 270–275
(2006).
46. Joo, S. et al. Ring-shaped replicative helicase encircles double-
stranded DNA during unwinding. Nucleic Acids Res. 47,
11344–11354 (2019).
47. Gai, D. et al. Mechanisms of conformational change for a
replicative hexameric helicase of SV40 large tumor antigen. Cell
119, 47–60 (2004).
48. James, J. A. et al. Structure of adeno-associated virus type 2
Rep40-ADP complex: insight into nucleotide recognition and
catalysis by superfamily 3 helicases. Proc. Natl Acad. Sci. USA 101,
12455–12460 (2004).
49. Baroudy, B. M. & Moss, B. Sequence homologies of diverse
length tandem repetitions near ends of vaccinia virus
genome suggest unequal crossing over. Nucleic Acids Res. 10,
5673–5679 (1982).
50. Baroudy, B. M., Venkatesan, S. & Moss, B. Incompletely
base-paired flip-flop terminal loops link the 2 DNA strands of the
vaccina-virus genome into one uninterrupted polynucleotide
chain. Cell 28, 315–324 (1982).
51. Buller, R. M. L. & Palumbo, G. J. Poxvirus pathogenesis. Microbiol
Rev. 55, 80–122 (1991).
52. Moss, B. Regulation of vaccinia virus transcription. Annu. Rev.
Biochem. 59, 661–688 (1990).
53. McDonald, W. F., Crozel-Goudot, V. & Traktman, P. Transient
expression of the vaccinia virus DNA polymerase is an
intrinsic feature of the early phase of infection and is unlinked
to DNA replication and late gene expression. J. Virol. 66,
534–547 (1992).
54. Khan, Y. A., White, K. I. & Brunger, A. T. The AAA+ superfamily:
a review of the structural and mechanistic principles of
these molecular machines. Crit. Rev. Biochem. Mol. Biol. 57,
156–187 (2022).
Article
Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner)
holds exclusive rights to this article under a publishing agreement
with the author(s) or other rightsholder(s); author self-archiving
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
of the accepted manuscript version of this article is solely
governed by the terms of such publishing agreement and
applicable law.
© The Author(s), under exclusive licence to Springer Nature America,
Inc. 2024
Article
Methods
Protein expression and purification
The full-length open reading frame of OPG117 of MPXV (MPXV 2022
West African strain, GenBank ON563414.3) and variants including
K509A/T511A/R514A/F630A/D656A mutant (M5t), R619/620A mutant,
R585A mutant, F588A mutant, R585A/F588A mutant and E79K/D88K/
R128E mutant, were cloned into the pCAG vector (Invitrogen) with an
N-terminal FLAG tag, respectively. Point mutations were introduced
using a standard two-step PCR and verified by DNA sequencing. The
primase domain truncations of D5, residues 238–785 and residues
323–785, were constructed into the pCAG vector with a N-terminal FLAG
tag. All the plasmids used to transfect cells were prepared by GoldHi
EndoFree Plasmid Maxi Kit (CWBIO).
The recombinant protein was overexpressed using the HEK293F
mammalian cells (Invitrogen) at 37 °C under 5% CO2 in a Multitron-Pro
shaker (Infors, 130 r.p.m.). When the cell density reached 2.0 × 106 cells
per ml, the plasmid was transiently transfected into the cells. To trans-
fect 1 l of cell culture, about 1.5 mg of the plasmid was premixed with
3 mg of polyethylenimines (Polysciences) in 50 ml of fresh medium
for 15 min before adding to the cell culture. Cells were collected by
centrifugation at 4,000g for 10 min after 60 h of transfection and resus-
pended in a buffer containing 25 mM HEPES (pH 7.5), 150 mM NaCl and
mixture of three protease inhibitors, aprotinin (1.3 μg ml−1, AMRESCO),
pepstatin (0.7 μg ml−1, AMRESCO) and leupeptin (5 μg ml−1, AMRESCO).
For purification of the D5 protein, the cells were subjected to
lysis by sonication and cell debris was removed by centrifugation at
16,000g for 50 min at 4 °C. The supernatant was loaded to the M2 resin,
rinsed with the wash (W) buffer containing 25 mM HEPES (pH 7.5) and
150 mM NaCl and eluted by W buffer supplemented with 0.2 mg ml−1
FLAG peptide. The eluent was then concentrated and applied to SEC
(Superose 6 Increase 10/300 GL, GE Healthcare) in the W buffer. The
peak fractions were collected and concentrated for EM analysis and
in vitro biochemical assays. For obtaining D5 (apyrase) protein, the
eluent was digested by apyrase (500 U ml−1, New England Biolabs)
at 4 °C for 16 h. The digested eluent was then applied to SEC (Super-
ose 6 Increase 10/300 GL, GE Healthcare) in the W buffer. The peak
fractions were also collected and concentrated for EM analysis and
functional assay.
Cryo-EM sample preparation and data acquisition
The short DNA duplex used for cryo-EM analysis was the 12/16
primer template from the Pol δ-DNA complex (Protein Data Bank
(PDB) ID 3IAY), containing the 12-nt oligonucleotide primer
(5′-ATCCTCCCCTAC-3′) annealed to the 16-nt template (5′-TAAGGT
AGGGGAGGAT-3′)55. The fork DNA duplex used for cryo-EM analysis
was the 49/47 primer template, the 49-nt oligonucleotide (5′-AGCT-
ACCATGCCTGCACGAATTAAGCAATTCGTAATCATGGTCATAGCT-3′)
annealed to the 47-nt (5′-AGCTATGACCATGATTACGAATTGCTTG-
GAATCCTGACGAACTGTAG-3′).
For the D5-ATP-ADP and D5-ADP forms, the peak fractions of fresh
D5 complex at 3 mg ml−1 plus 1 mM MgCl2 were used for cryo-EM sample
preparation (different batch proteins). For the D5-ATP-ADP-apo-ssDNA
form, the peak fractions of fresh D5 complex at 3 mg ml−1 were incu-
bated with 20 μM ‘short DNA duplex’ and 1 mM MgCl2 for about 1 h.
For the D5-apo form, about 12 mg of D5 was incubated with 10 U of
apyrase (500 U ml−1, New England Biolabs) at 4 °C for 16 h, and then
the mixture was subjected to SEC in W buffer. For the D5-apo-ssDNA
form, the peak fractions of apo protein were diluted to 3 mg ml−1 and
incubated with 20 μM ‘short DNA duplex’ and 1 mM MgCl2. For the
D5-ATP-γS-ADP-ssDNA form, the apo protein was supplemented 1 mM
ATP-γS, 1 mM MgCl2 without or with 20 μM ‘short DNA duplex’. For the
D5-ADP-ssDNA form, the apo protein was incubated with 1 mM ATP,
20 μM fork DNA duplex and 1 mM MgCl2 for 5 min.
Aliquots (3.3 μl) of the protein complex were placed on
glow-discharged holey carbon grids (Quantifoil Au R1.2/1.3). The grids
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
were blotted for 3.0 or 3.5 s and flash-frozen in liquid ethane cooled
by liquid nitrogen with Vitrobot (Mark IV, Thermo Fisher Scientific).
The prepared grids were transferred to a Titan Krios operating at
300 kV equipped with Gatan K3 detector and GIF Quantum energy
filter. Video stacks were automatically collected using EPU software
(Thermo Fisher Scientific), with a slit width of 20 eV on the energy filter
and a defocus range from −1.4 to −1.8 µm at a nominal magnification
of ×81,000. The total dose rate was approximately 50 e−/Å2 for each
stack. The stacks were motion corrected with MotionCor2 (ref. 56)
and binned twofold, resulting in pixel sizes of 1.095, 1.072 or 1.077 Å
per pixel. Meanwhile, dose weighting was performed57. The defocus
values were estimated with Gctf58.
Data processing
The cryo-EM structures of the complex were solved in Relion v.3.0.6
and cryoSPARC v.3.3.1. Particles were automatically picked using
Relion v.3.0.6 (refs. 59–62) from manually selected micrographs.
After 2D classification with Relion, good particles were selected and
subject to several cycles of 2D classification, ab initio reconstruc-
tion and multiple cycle of heterogeneous refinement without sym-
metry using cryoSPARC63. The good particles were selected and
subjected to local contrast transfer function refinement with C1 sym-
metry, nonuniform refinement, resulting in three-dimensional (3D)
reconstruction for whole structures. The particles in the data sets
ATP-ADP-apo-ssDNA form, ADP-ssDNA form and ATP-γS -ADP-ssDNA
form were subject to 3D classification and focused refinement
with Relion.
The resolution was estimated with the gold-standard Fourier shell
correlation 0.143 criterion64 with high-resolution noise substitution65.
Refer to Extended Data Figs. 2–5 and Table 1 for details of data collec-
tion and processing.
Model building and structure refinement
Predicted models of D5 was first obtained using Alphafold2 (ref. 66),
which was further manually adjusted based on the cryo-EM map
with COOT67. Each residue was manually checked with the chemical
properties taken into consideration during model building. Several
segments, whose corresponding densities were invisible, were not
modeled. Structural refinement was performed in Phenix 68 with
secondary structure and geometry restraints to prevent overfitting.
To monitor the potential overfitting, the model was refined against
one of the two independent half maps from the gold-standard 3D
refinement approach. Then, the refined model was tested against
the other map. Statistics associated with data collection, 3D recon-
struction and model building are summarized in Table 1 and Sup-
plementary Table 1.
ATPase assay
ATPase activity was determined in 25 mM HEPES pH 7.5, 150 mM NaCl
by Na+-K+-ATP activity assay kits (Beijing Solarbio Biotechnology Co.,
Ltd). The released phosphate from ATP hydrolysis was determined over
time using the kit. The kit specifies that the amount of 1 μM inorganic
phosphorus produced by each milligram of enzyme catalyzing ATP
hydrolysis in 1 h is a unit of enzyme activity (U mg−1 protein). The protein
concentration of WT D5 was determined by BioRad assay with BSA as a
standard. Then an aliquot (10 μl) of WT D5 and each variant was mixed
with 5× sample buffer (250 mM Tris-HCl (pH 6.8), 10% SDS, 50% glyc-
erol, 0.2% bromphenol blue and 500 mM dithiothreitol), and loaded
to SDS–PAGE gels. After electrophoresis and staining with Coomassie,
the gels were exposed by ChampGel7000. The band intensities of WT
D5 and variants were measured by densitometry using the software
ImageJ v.2.9.0. The protein concentration of each variant was normal-
ized based on the protein concentration and band intensities of the WT
D5. The resulting data collection was repeated at least three times and
data were analyzed by GraphPad Prism v.9.0.
Article
Electrophoretic mobility shift assay
For ssDNA binding assays, the substrate was a 49 nt and 3′ labeled
Cyanine5 (Cy5) oligonucleotide (5′-AGCTACCATGCCTGCACGAAT
TAAGCAATTCGTAATCATGGTCATAGCT). DNA-binding assays with Cy5-
labeled substrate (0.2 μM) and each protein (0.2 μM), were performed
in 25 mM HEPES pH 7.5, 150 mM NaCl. Reactions were incubated for 1 h
at 4 °C and terminated by adding 10% glycerol and 0.13% w/v bromo-
phenol blue. Products were separated on a 4 to 20% gradient gel, and
gels were exposed to MiniChemi610 Plus for imaging. The free DNA,
unbound by protein, were measured with ImageJ v.2.9.0 for quantita-
tive evaluation of DNA-binding ability. Three replications were taken
independently for data analysis and the data have been presented as
mean ± standard deviation.
Helicase activity assay
The oligonucleotides 5′-AGCTACCATGCCTGCACGAATTAAGCAATTCG
TAATCATGGTCATAGCT with 3′ labeled Cy5 and 5′-AGCTATGACCAT
GATTACGAATTGCTTGGAATCCTGACGAACTGTAG were annealed to
generate the Fork dsDNA substrate. Helicase assays with Cy5-labeled
substrates (0.4 μM) and each protein (0.1, 0.2, 0.4 and 0.8 μM) were
performed in 25 mM HEPES pH 7.5, 150 mM NaCl, 4 mM MgCl2, 4 mM
ATP, which was named the ATP-only group. For ADP only group,
Cy5-labeled substrates and each protein (apyrase digested) were
performed in 25 mM HEPES pH 7.5, 150 mM NaCl, 4 mM MgCl2, 4 mM
ATP. For the ADP, then fork, then ATP group, each protein (apyrase
digested) was incubated with 1 mM ADP at 4 °C for 30 min and mixed
with Cy5-labeled dsDNA substrates in 25 mM HEPES pH 7.5, 150 mM
NaCl, 4 mM MgCl2 and 4 mM ATP for reaction. For the Fork, then ADP,
then ATP group, each protein (apyrase digested) was incubated with
0.4 μM Cy5-labeled dsDNA substrates at 4 °C for 1 h, and incubated
with 1 mM ADP at 4 °C for 30 min. Finally, the preformed protein was
placed in 25 mM HEPES pH 7.5, 150 mM NaCl, 4 mM MgCl2, 4 mM ATP
for reaction. All reactions under various conditions were delivered for
1 h at 4 °C and added 10% glycerol, 20 mM EDTA, 0.5% SDS and 0.2%
bromphenol blue to terminate the reaction. Products were separated
on a 20% polyacrylamide–TBE gel, and gels were exposed to Min-
iChemi610 Plus for imaging. The free DNA unwound by helicase was
quantified by software ImageJ v.2.9.0. All experiments were performed
in triplicate, and the data were presented as mean ± standard deviation
by GraphPad Prism v.9.0.
Reporting summary
Further information on research design is available in the Nature Port-
folio Reporting Summary linked to this article.
Data availability
Cryo-EM maps and molecular models have been deposited in the
EM Data Bank and PDB, respectively. Accession codes are listed
here and in Table 1. Atomic coordinates and cryo-EM density maps
of D5 protein in ATP-ADP-apo-ssDNA form IS1 (PDB 8HWA whole
map EMD-35051), ATP-ADP-apo-ssDNA form IS2 (PDB 8HWB
whole map EMD-35052), apo form (PDB 8HWC whole map EMD-
35053), ADP form (PDB 8HWD whole map: EMD-35054), ATP-ADP
form (PDB 8HWE whole map EMD-35055), ADP-ssDNA form (PDB
8HWF whole map EMD-35056), ATP-γS-ADP-ssDNA form (PDB
8HWG whole map EMD-350567) and apo-ssDNA form (PDB 8HWH
whole map EMD-350568) have been deposited to the PDB (http://
www.rcsb.org) and the Electron Microscopy Data Bank (https://
www.ebi.ac.uk/pdbe/emdb/), respectively. All other data will be made
available upon request. Source data are provided with this paper.
References
55. Swan, M. K. et al. Structural basis of high-fidelity DNA synthesis
by yeast DNA polymerase delta. Nat. Struct. Mol. Biol. 16,
979–986 (2009).
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
56. Zheng, S. Q. et al. MotionCor2: anisotropic correction of
beam-induced motion for improved cryo-electron microscopy.
Nat. Methods 14, 331–332 (2017).
57. Grant, T. & Grigorieff, N. Measuring the optimal exposure for
single particle cryo-EM using a 2.6 A reconstruction of rotavirus
VP6. eLife 4, e06980 (2015).
58. Zhang, K. Gctf: real-time CTF determination and correction.
J. Struct. Biol. 193, 1–12 (2016).
59. Zivanov, J. et al. New tools for automated high-resolution
cryo-EM structure determination in RELION-3. eLife 7,
e42166 (2018).
60. Kimanius, D. et al. Accelerated cryo-EM structure determination
with parallelisation using GPUs in RELION-2. eLife 5, e18722 (2016).
61. Scheres, S. H. RELION: implementation of a Bayesian
approach to cryo-EM structure determination. J. Struct. Biol. 180,
519–530 (2012).
62. Scheres, S. H. A Bayesian view on cryo-EM structure
determination. J. Mol. Biol. 415, 406–418 (2012).
63. Punjani, A. et al. cryoSPARC: algorithms for rapid unsupervised
cryo-EM structure determination. Nat. Methods 14, 290–296
(2017).
64. Rosenthal, P. B. & Henderson, R. Optimal determination of
particle orientation, absolute hand, and contrast loss in
single-particle electron cryomicroscopy. J. Mol. Biol. 333,
721–745 (2003).
65. Chen, S. et al. High-resolution noise substitution to measure
overfitting and validate resolution in 3D structure determination
by single particle electron cryomicroscopy. Ultramicroscopy 135,
24–35 (2013).
66. Jumper, J. et al. Highly accurate protein structure prediction with
AlphaFold. Nature 596, 583–589 (2021).
67. Emsley, P. et al. Features and development of Coot. Acta
Crystallogr. Sect. D., Biol. Crystallogr. 66, 486–501 (2010).
68. Adams, P. D. et al. PHENIX: a comprehensive Python-based
system for macromolecular structure solution. Acta Crystallogr.
Sect. D., Biol. Crystallogr. 66, 213–221 (2010).
Acknowledgements
We thank the Cryo-EM Facility of Southern University of Science and
Technology (SUSTech) for providing the facility support. We thank
X. Ma, L. Zhang and P. Li at the Cryo-EM Center of SUSTech for technical
support in EM data acquisition. We thank Z. Liu for technical support
on computing environment. This work was funded by the Science,
Technology and Innovation Commission of Shenzhen Municipality
(grant no. JSGG20220226085550001 to R.Y.) and the National Natural
Science Foundation of China (grant no. 82202517 to R.Y.).
Author contributions
R.Y. conceived the project. R.Y., Y.G., Y.L. and J.Z. designed the
experiments. Y.L. did the molecular cloning, protein purification,
cryo-EM data collection and processing, and model building. Y.G.
and J.Z. did the protein purification, cryo-EM data collection and
ATPase assay, DNA-binding assay and the helicase assay. All authors
contributed to data analysis. R.Y. and Y.G. wrote the paper.
Competing interests
The authors declare no competing interests.
Additional information
Extended data is available for this paper at
https://doi.org/10.1038/s41594-023-01142-0.
Supplementary information The online version
contains supplementary material available at
https://doi.org/10.1038/s41594-023-01142-0.
Article
Correspondence and requests for materials should be addressed to
Yingying Guo or Renhong Yan.
Peer review information Nature Structural & Molecular Biology
thanks Hauke Hillen and the other, anonymous, reviewer(s) for their
contribution to the peer review of this work. Peer reviewer reports
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
are available. Primary Handling Editor: Katarzyna Ciazynska, in
collaboration with the Nature Structural & Molecular Biology team.
Peer reviewer reports are available.
Reprints and permissions information is available at
www.nature.com/reprints.
Article
Extended Data Fig. 1 | Sequence alignments of the D5 protein from different
poxviruses. The hallmark motifs (Walker a and b) and the Arginine finger
residues are shown by yellow, green and black dashed box, respectively. Various
colors are used to distinguish the primase domain (blue color), Zn-binding motif
(orange color), collar-domain (yellow color), AAA+ helicase domain (green color)
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
and C-terminal domain (purple color). Gene and protein ID are as follows: MPXV
2022 West African strain (Genebank: ON563414.3), MPXV_Zaire_96_I_16 strain
(Genebank: NP_536529.1), MPXV_UK_P3 strain (Genebank: YP_010377099.1),
VACV (Genebank: AGB75830.1), VARV (Genebank: ABG44074.1). Sequences were
aligned by MultAlin and ESPript 3.0.
Article https://doi.org/10.1038/s41594-023-01142-0

Extended Data Fig. 2 | See next page for caption.

Nature Structural & Molecular Biology

Article
Extended Data Fig. 2 | Cryo-EM analysis of D5 in ATP-ADP-Apo-ssDNA form.
a, Represent micrographs and 2D class averages of D5. Scale bar of 2D, 10 nm.
A total of 2237 micrographs were collected. b and c, Flowchart for cryo-EM data
processing. Please refer to the ‘Data Processing’ in Methods section for details.
(c) Local resolution map for the 3D reconstruction of the overall structure.
d, FSC curve. The resolution was estimated with the gold-standard Fourier shell
correlation 0.143 criterion with high-resolution noise substitution. e-h, Euler
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
angle distribution in the final 3D reconstruction of class 1–4. i and j, FSC curve of
the refined model versus the overall structure of class 1/2 that it is refined against
(black); of the model refined against the first half map versus the same map (red);
and of the model refined against the first half map versus the second half map
(green). The small difference between the red and green curves indicates that the
refinement of the atomic coordinates did not suffer from overfitting.
Article https://doi.org/10.1038/s41594-023-01142-0

Extended Data Fig. 3 | See next page for caption.

Nature Structural & Molecular Biology

Article
Extended Data Fig. 3 | Cryo-EM analysis of D5 structures without ssDNA.
a-c, Represent micrographs and 2D class averages of D5 Apo, ADP, and ATP-ADP
form. Scale bar of 2D, 10 nm. A total of 3504, 1514 and 1004 micrographs were
collected, respectively. d and e, Flowchart for cryo-EM data processing. Please
refer to the ‘Data Processing’ in Methods section for details. e, Local resolution
map for the 3D reconstruction of the overall structure. f, FSC curve. The
resolution was estimated with the gold-standard Fourier shell correlation 0.143
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
criterion with high-resolution noise substitution. g, Euler angle distribution in
the final 3D reconstruction. h, FSC curve of the refined model versus the overall
structure that it is refined against (black); of the model refined against the first
half map versus the same map (red); and of the model refined against the first half
map versus the second half map (green). The small difference between the red
and green curves indicates that the refinement of the atomic coordinates did not
suffer from overfitting.
Article https://doi.org/10.1038/s41594-023-01142-0

Extended Data Fig. 4 | See next page for caption.

Nature Structural & Molecular Biology

Article
Extended Data Fig. 4 | Cryo-EM analysis of D5 structures with ssDNA.
a-c, Represent micrographs and 2D class averages of D5-ADP-ssDNA,
ATP-γS -ADP-ssDNA, and Apo-ssDNA form. Scale bar of 2D, 10 nm. A total of
4594, 1431 and 201 micrographs were collected, respectively. d and e, Flowchart
for cryo-EM data processing. Please refer to the ‘Data Processing’ in Methods
section for details. e, Local resolution map for the 3D reconstruction of
the overall structure. f, FSC curve. The resolution was estimated with the
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
gold-standard Fourier shell correlation 0.143 criterion with high-resolution noise
substitution. g, Euler angle distribution in the final 3D reconstruction. h, FSC
curve of the refined model versus the overall structure that it is refined against
(black); of the model refined against the first half map versus the same map (red);
and of the model refined against the first half map versus the second half map
(green). The small difference between the red and green curves indicates that the
refinement of the atomic coordinates did not suffer from overfitting.
Article https://doi.org/10.1038/s41594-023-01142-0

Extended Data Fig. 5 | Representative cryo-EM density maps of D5. a, Cryo-EM density maps of protein are shown at threshold of 7 σ. b, Cryo-EM density map of
ssDNA is shown at threshold of 5 σ. c-e, Cryo-EM density maps of ligand and the binding pocket are shown at threshold of 7 σ. f, Density maps show the C-terminal motif
could interact with the helicase domain in neighboring protomer.

Nature Structural & Molecular Biology

Article https://doi.org/10.1038/s41594-023-01142-0

Extended Data Fig. 6 | See next page for caption.

Nature Structural & Molecular Biology

Article
Extended Data Fig. 6 | Structural characteristics of D5 of MPXV. a, The
organization of four layers of D5. Left: Structure of one protomer of full-length
D5 is showed by cartoon. The color strategy is consistent with Fig.1a. Right:
Overall structures of D5 is showed by surface and the color strategy is consistent
with Fig.1d. b, Structural analysis of N-terminal primase domain of D5. Shown
here is cartoon presentation of domain-colored cryo-EM structures of N-terminal
primase domain of D5 and structural alignment between RRM in protomer A and
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
human PrimPol (PDB ID: 5L2X). RMSD between 44 pruned atom pairs is
1.210 Å and across all 207 pairs is 9.768 Å. Insets: In top panel, structural similarity
between RRM in protomer A (red) and human PrimPol (cyan) indicates the similar
functions and potential ATP binding pocket in N-terminal primase domain of
MPVX. Bottom panel shows the density at potential ATP binding pocket allow us
to build an ATP molecular. Cryo-EM density map is shown at threshold of 7 σ.
Article https://doi.org/10.1038/s41594-023-01142-0

Extended Data Fig. 7 | See next page for caption.

Nature Structural & Molecular Biology

Article
Extended Data Fig. 7 | Helicase activity of D5 and the variants. a-d, Helicase
activity of D5 (Apyrase) and truncation D5238-785 (Apyrase). When ATP acts as an
energy substrate, D5full length (Apyrase) has weak helicase activity, while truncation
D5238-785 (Apyrase) shows apparent increasing in helicase activity. Whereas, ADP
can’t catalyze the reaction. When apyrase treated proteins were pre-incubated
with ADP, the helicase activity was reduced. However, for D5full length (Apyrase),
pre-incubation with Fork DNA before ADP restored its activity to ATP only group,
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
which indicates that ADP may have a regulatory effect on helicase activity of full
length D5. e, Free DNA releasing by proteins was delivered quantitative analysis
with ImageJ 2.9.0 to evaluate the helicase ability of various proteins. Three
replications were taken independently for data analysis and the data has been
presented as mean ± standard deviation. The samples derive from the same
experiment and that gels were processed in parallel.
Article https://doi.org/10.1038/s41594-023-01142-0

Extended Data Fig. 8 | See next page for caption.

Nature Structural & Molecular Biology

Article
Extended Data Fig. 8 | Functional activity of D5 and the variants. a, ATP
hydrolysis activity of D5 and the mutants. WT D5 shows obvious ATPase activity.
Comparing with WT D5, R619/620A mutant and M5t (K509A/ T511A/ R514A/
F630A/ R656A) mutant show apparent decreasing in ATPase activity. Also,
ATP-γS has a great impact on the ATPase activity of the proteins. b, DNA binding
ability of D5 and the mutants. WT D5 has DNA binding activity, and R585A,
F588A and R585A/F588A mutants could impair the DNA binding ability of D5.
All experiments were performed in triplicates and the free DNA was delivered
quantitative analysis to evaluate the DNA binding ability of various proteins. In
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
(a) and (b), **** indicates extremely significant difference (p < 0.0001),
** indicates significant difference (p < 0.01), ns indicates no significance.
Three replications were taken independently for data analysis and the data has
been presented as mean ± standard deviation. For multiple comparisons, P values
were derived from ordinary one-way ANOVA with Šídák′s multiple comparisons
test. The default parament settings were applied to multiple comparisons.
P values< 0.05 (two side) were considered significant. Graphs were prepared in
Graphpad prism (Version 9.0).
Article https://doi.org/10.1038/s41594-023-01142-0

Extended Data Fig. 9 | See next page for caption.

Nature Structural & Molecular Biology

Article
Extended Data Fig. 9 | Structural comparison among N-terminal domain of
D5-ATP-γS-ADP ssDNA form and D5 ATP-ADP-Apo ssDNA form IS1/2. a, Here
shows the cryo-EM map and model of D5 ATP-γS-ADP-ssDNA form. Density of D5
is colored light blue and cartoon presentation of D5 and DNA are colored gray
and black, respectively. Alignments among N-terminal primase domain of D5
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
ATP-γS-ADP ssDNA form and D5 ATP-ADP-Apo ssDNA form IS1/2 shows that DNA
in ATP-γS-ADP ssDNA form is consistent with it in IS1 (b) but not IS2 (c). RMSD
between 3139 Cα pairs of D5 IS1 and D5 ATP-γS-ADP-ssDNA form, D5 IS2 and D5
ATP-γS-ADP-ssDNA form are 1.118 Å and 4.840 Å, respectively.
Article https://doi.org/10.1038/s41594-023-01142-0

Extended Data Fig. 10 | See next page for caption.

Nature Structural & Molecular Biology

Article
Extended Data Fig. 10 | Comparison of D5 bound to different ligands and E1
protein. a, The structural comparison of D5 bound to different ligands shows
there is no shift between collar domains and dramatic shifts in AAA+ helicase
domain. Right panel is the diagram to show the shifts angles in AAA+ helicase
domain. AAA+ helicase domain of D5-ATP form (yellow), D5-ADP tight form
(green), D5-Apo bound DNA form (purple), D5-ADP loose form (blue), and Apo
(magenta) are looser in turn. b, The interaction network of ADP in D5-ADP loose
form. To make the shift more clearly, superpositions of ATP form and ADP tight
form, ADP tight form and Apo bound DNA form, Apo bound DNA form and ADP
loose form, and ADP loose form and Apo are showed from left to right. RMSD
Nature Structural & Molecular Biology
https://doi.org/10.1038/s41594-023-01142-0
between 378 cα pairs (323 to 700 amino acid) of the chain in ATP form and ADP
tight form, ADP tight form and Apo bound DNA form, Apo bound DNA form and
ADP loose form, and ADP loose form and Apo are 1.791, 2.931, 4.220, and 1.990.
c, The interaction network of ADP in D5-ADP loose form. d, Structural
comparison of D5 of MPXV and E1 of papillomavirus. Middle: Structural
alignment of D5 and E1. D5 are colored blue and E1 are colored green. D5 and E1
are showed at left and right panel. Collar domain are colored light blue (D5) and
light green (E1), and AAA+ helicase domains are colored dark blue (D5) and dark
green (E1).
 nature portfolio | reporting summary
Corresponding author(s): Renhong Yan
Last updated by author(s): Aug 29, 2023

Reporting Summary
Nature Portfolio wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Portfolio policies, see our Editorial Policies and the Editorial Policy Checklist.

Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.

A description of all covariates tested

A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient)
AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)

For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.

For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Our web collection on statistics for biologists contains articles on many of the points above.

Software and code

Policy information about availability of computer code
Data collection EPU software

Data analysis Relion 3.0.6, cryoSPARC v3.3.1, ChimeraX1.4 v3.3.1, MotionCor2 1.1.0, GCTF 1.06, Phenix 1.11.1, Coot 0.8.2, Pymol 1.8.6.0, GraphPad Prism
9.0, ImageJ 2.9.0
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and
reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Portfolio guidelines for submitting code & software for further information.

Data
Policy information about availability of data
March 2021

All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A description of any restrictions on data availability
- For clinical datasets or third party data, please ensure that the statement adheres to our policy

Atomic coordinates and cryo-EM density maps of D5 protein in ATP-ADP-Apo-ssDNA form IS1 (PDB: 8HWA, whole map: EMD-35051), ATP-ADP-Apo-ssDNA form IS2
(PDB 8HWB, whole map: EMD-35052), Apo form (PDB: 8HWC, whole map: EMD-35053), ADP form (PDB 8HWD, whole map: EMD-35054), ATP-ADP form (PDB:
8HWE, whole map: EMD-35055), ADP-ssDNA form (PDB: 8HWF, whole map: EMD-35056), ATP-γS-ADP-ssDNA form (PDB: 8HWG, whole map: EMD-350567), and

1
Apo-ssDNA form (PDB:8HWH, whole map:EMD-350568) have been deposited to the Protein Data Bank (http://www.rcsb.org) and the Electron Microscopy Data
Bank (http://www.ebi.ac.uk/pdbe/emdb/), respectively.

nature portfolio | reporting summary

Human research participants
Policy information about studies involving human research participants and Sex and Gender in Research.

Reporting on sex and gender Human research participants were not involved in this study.

Population characteristics Human research participants were not involved in this study.

Recruitment Human research participants were not involved in this study.

Ethics oversight Human research participants were not involved in this study.

Note that full information on the approval of the study protocol must also be provided in the manuscript.

Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.

Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf

Life sciences study design

All studies must disclose on these points even when the disclosure is negative.
Sample size Sample size calculation was performed based on the previous knowledge. Cryo-EM datasets were collected and generated interpretable map
with sufficient resolution reported in this study. Due to the resolution range is 3.0-3.9 angstroms, we judged the sample size is enough.

Data exclusions No data was excluded.

Replication For the biochemical assay, each experiment was replicated at least 3 times on separate occasions. Experimental findings were reliably
reproduced. For D5 protein expression and purification were examined multiple times independently. Multiple cryo-EM grids were prepared
and frozen, datas were collected for two or three selected grids of the ATP-ADP-Apo-ssDNA, Apo, ADP, ATP-ADP, ADP-ssDNA, ATP-γS-ADP-
ssDNA and Apo-ssDNA structures. Representative ones of protein purification and micrographs were shown in the supplementary
information.

Randomization Randomization is not relevant for this study, as there no groups allocated in any of the experiments. This study did not involve animals or
humans research participants.

Blinding Blinding is not applicable to this research as this is deemed not practically feasible.

Reporting for specific materials, systems and methods

We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material,
system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.

Materials & experimental systems Methods

n/a Involved in the study n/a Involved in the study
Antibodies ChIP-seq
Eukaryotic cell lines Flow cytometry
March 2021

Palaeontology and archaeology MRI-based neuroimaging

Animals and other organisms
Clinical data
Dual use research of concern

2
Eukaryotic cell lines

nature portfolio | reporting summary

Policy information about cell lines and Sex and Gender in Research
Cell line source(s) HEK293F (invitrogen)

Authentication No further authentication was performed for commercially available cell lines.

Mycoplasma contamination Not tested for mycoplasma contamination.

Commonly misidentified lines No commonly misidentified cell lines were used.

(See ICLAC register)

March 2021