The architecture of the DNA replication origin
recognition complex in Saccharomyces cerevisiae
Zhiqiang Chen*†‡, Christian Speck†§¶, Patricia Wendel§, Chunyan Tang*, Bruce Stillman§?, and Huilin Li*?**
*Biology Department, Brookhaven National Laboratory, Upton, NY 11973;§Cold Spring Harbor Laboratory, 1 Bungtown Road, Cold Spring
Harbor, NY 11742; and **Department of Biochemistry and Cell Biology, Stony Brook University, Stony Brook, NY 11794
Contributed by Bruce Stillman, April 21, 2008 (sent for review December 10, 2007)
The origin recognition complex (ORC) is conserved in all eu-
karyotes. The six proteins of the Saccharomyces cerevisiae ORC
that form a stable complex bind to origins of DNA replication and
recruit prereplicative complex (pre-RC) proteins, one of which is
Cdc6. To further understand the function of ORC we recently
determined by single-particle reconstruction of electron micro-
graphs a low-resolution, 3D structure of S. cerevisiae ORC and the
ORC–Cdc6 complex. In this article, the spatial arrangement of the
ORC subunits within the ORC structure is described. In one ap-
proach, a maltose binding protein (MBP) was systematically fused
to the N or the C termini of the five largest ORC subunits, one
subunit at a time, generating 10 MBP-fused ORCs, and the MBP
density was localized in the averaged, 2D EM images of the
MBP-fused ORC particles. Determining the Orc1–5 structure and
comparing it with the native ORC structure localized the Orc6
subunit near Orc2 and Orc3. Finally, subunit–subunit interactions
were determined by immunoprecipitation of ORC subunits syn-
thesized in vitro. Based on the derived ORC architecture and
existing structures of archaeal Orc1–DNA structures, we propose a
Cdc6. The studies provide a basis for understanding the overall
structure of the pre-RC.
electron microscopy ? structure ? ATPase
part of B1 define the binding sequence for the origin recognition
complex (ORC) (1–3). ORC binds to origin DNA in an ATP-
dependent manner and recruits other essential proteins, such as
the initiation factors Cdc6, Cdt1, and the presumptive DNA
helicase MCM, to the autonomously replicating sequence (ARS)
to form a prereplicative complex (pre-RC) before the initiation
of DNA replication that occurs in S phase (4–7). ORC consists
of six proteins named in the descending order of their relative
mass: Orc1 (120 kDa), Orc2 (71 kDa), Orc3 (62 kDa), Orc4 (56
kDa), Orc5 (53 kDa), and Orc6 (50 kDa). The calculated mass
of ORC is ?412 kDa. Only two of the ORC subunits (Orc1 and
Orc5) are known to bind ATP (8), although the largest five
subunits are predicted to contain an AAA? fold and a DNA-
binding winged helix domain (WHD) within their C-terminal
halves (9, 10).
The prokaryotic origin recognition proteins consist of a single
polypeptide that can form oligomeric structures (11–17), which
raises the question of why in eukaryotes ORC has six subunits
and why Cdc6 also contributes to origin recognition (18, 19).
Structural studies of the replication initiator proteins have begun
individual initiator proteins cooperate to promote initiation of
DNA replication is not clear. The eubacterial DnaA structure
In contrast, the recent structures of archaeal Orc1/Cdc6 in
complex with DNA implicates a different mechanism in which
the initiators wrap around DNA (14, 17). The low-resolution EM
structures of ORC from yeast and Drosophila are similar in size
n Saccharomyces cerevisiae origins of DNA replication contain
conserved A, B1, and B2 elements, where the A element and
and overall architecture, but with appreciable difference in
details (9, 10). So far, crystallographic studies of ORC have not
been successful, possibly because the multiple-subunit structure
is not very rigid in the absence of DNA and a DNA of ?50 base
pairs (bp) long might be required to fully stabilize the structure
(10, 20, 21). To gain further insight into the low-resolution EM
map, we studied the subunit arrangement of ORC by using a
structure-based strategy of systematic maltose binding protein
(MBP) fusion in combination with 2D image classification and
a biochemical strategy of in vitro transcription-translation and
coimmunoprecipitation to define subunit–subunit interactions.
A model for ORC is presented based on these studies.
EM and Single-Molecule Analysis of ORC Particles. PurifiedORCand
its derivatives either lacking the Orc6 subunit or containing MBP
attached to single subunits were visualized by transmission EM
after negative staining. ORC appears to be an elongated particle in
raw EM images (Fig. 1A). The raw particle images were subjected
to multivariate statistical analysis and classification. Images within
the same classes were averaged to produce the class averages. The
structural features in the 2D class averages of ORC match closely
with the reprojections of the 3D map that we have derived
previously (Fig. 1A Inset Left and Right, respectively). In the most
commonly observed view of ORC, there are seven high-density
regions that are labeled from top to bottom with ? through ? (Fig.
1B Upper Left). The 3D structure of ORC has the dimensions of
?160 ? 130 ? 100 Å (ref. 10 and Fig. 1C). Comparison of the class
tag associated with ORC. The subunit localization result by EM
Localization of Orc6 in ORC. In the absence of Orc6, Orc1-5 could
still form a stable subcomplex (21). This subcomplex provided us
an opportunity to localize the Orc6 in the ORC structure by
difference map in either 2D class averages or 3D reconstruc-
tions. The raw EM images of Orc1-5 were virtually indistinguish-
able from that of ORC. However, in selected 2D class averages,
the density in the region of Orc1-5 between ? and ?, as indicated
in the same region of ORC. We assigned this region as the
Author contributions: Z.C., C.S., B.S., and H.L. designed research; Z.C., C.S., P.W., C.T., and
H.L. wrote the paper.
The authors declare no conflict of interest.
Freely available online through the PNAS open access option.
†Z.C. and C.S. contributed equally to this work.
‡Present address: Institute of Advanced Materials and Renewable Energy, University of
Louisville, Louisville, KY 40292.
¶Present address: DNA Replication Group, Medical Research Council Clinical Sciences
Center, Imperial College Faculty of Medicine, Hammersmith Hospital Campus, Du Cane
Road, London W12 0NN, United Kingdom.
?To whom correspondence may be addressed. E-mail: email@example.com or firstname.lastname@example.org.
© 2008 by The National Academy of Sciences of the USA
July 29, 2008 ?
vol. 105 ?
location of Orc6 in ORC. We further determined the 3D
structure of this partial complex. The overall size and structure
of Orc1-5 subcomplex was similar to ORC (Fig. 1 B and C). The
3D difference map confirmed the location assignment of Orc6
(Fig. 1C Right). At the display level (3?), the main difference
peak encloses a volume corresponding to ?32 kDa of protein
mass. This mass accounts for 60% of the expected 50-kDa mass
of Orc6. We note that the peripheral location of Orc6 is
consistent with the knowledge that Orc6 is dispensable in the
DNA binding activity of ORC (21–23).
Preparation and Characterization of the MBP-Fused ORCs. Initially,
immunolabeling experiments were used to map the subunit
organization. ORC was incubated with the Fab fragments (?50
kDa) of four antibodies recognizing Orc1, Orc2, Orc3, and Orc4,
respectively, and the Fab–ORC complexes were further purified
by gel filtration column chromatography. Although stable Fab–
ORC complexes could be seen, no significant densities were
observed for the Fab in either the 2D class averages or the 3D
reconstructions of the four Fab–ORC complexes (data not
shown). The failure was probably caused by the low occupancy
To overcome the difficulty in immunolabeling, the Escherichia
coli MBP (?38 kDa) was fused to either the N or the C terminus
of each of the largest five subunits (Orc1, Orc2, Orc3, Orc4, and
Orc5), with a 9-aa linker. Of the 10 constructs prepared, we were
able to purify nine MBP-fused ORCs; the one exception was
ORC with MBP fused to the C terminus of Orc5 [Orc(1-4,6)–
Orc5–MBP]. SDS/PAGE showed that all nine purified MBP-
fused ORCs were properly assembled (data not shown). Fur-
thermore, all MBP-fused ORCs, together with the Orc1-5
subcomplex, were able to bind to the ARS1 origin and bind Cdc6
to form an ORC–Cdc6 complex on the origin DNA, as indicated
by the Cdc6-induced supershift (Fig. 2). The experiment dem-
onstrated that MBP fusion did not affect the ability of ORCs to
interact with DNA or Cdc6. In one case, the MBP–Orc1–ORC
complex, DNA binding was significantly enhanced compared
with native ORC, suggesting that the N terminus of Orc1 may
modulate DNA binding.
Subunit Localization in the MBP-Fused ORCs. The purified ORCs
that contained MBP were prepared for EM characterization. We
were able to detect MBP in seven of nine ORCs that contained
MBP; the MBP in the MBP–Orc1–Orc(2-6) and MBP–Orc2–
Orc(1, 3-6) complexes (i.e., MBP fused to the N termini) was not
visible, even though we could calculate well defined 2D class
averages of these complexes that were essentially the same as
reference-free 2D class averages representing the most common
view for each of the seven MBP-fused ORCs. Each of these
images was an average of a large number of raw particle images,
ranging from 45 to 450 particles. Therefore the structural
features in these averages are statistically significant. Well
defined density, away from the main body of ORC, was present,
and these densities in different images of the same fusion ORC
were generally clustered to a small area, as shown in illustrations
(Fig. 3 Far Right). We assigned the extra density to the ?38-kDa
MBP. Thus the location of the observed MBP density reflected
the approximate position of the N or C terminus of the corre-
sponding subunit onto which the MBP was fused.
When MBP was observed, its location varied over a small
region of the 2D image, most likely because of the ability of the
MBP on the N or C termini to occupy alternative, but discrete,
positions. The size of the region occupied by MBP most likely
A raw electron micrograph of the untagged ORC deeply stained by uranyl
acetate showing various views. (Inset) Seven reprojections of the derived ORC
3D map (Left) and the corresponding 2D class averages (Right). (B) A compar-
ison of three typical side views of ORC (Upper) with that of Orc1-5 (Lower)
view refers to the raw particle images used for calculating the reference-free
class average. The Greek letters ? through ? indicate the seven high-density
of 3?. The largest difference density represents the Orc6 position. b and c are
views rotated 90° from a around a vertical and a horizontal axis, respectively.
EM structures of the yeast ORC (Orc1-6) and Orc1-5 subcomplex. (A)
24610 12 14 16 18 20
28 30 32 34
or Orc1-5 subcomplex as indicated. The gel-shift assay was performed with 2
nM concentrations of the various ORCs, 0.2 nM of the concentrations labeled
ARS1 DNA, and 0.8–2.4 nM of the concentrations Cdc6.
The MBP-fused ORCs are active in binding to Cdc6 and DNA. Gel-shift
Chen et al.
July 29, 2008 ?
vol. 105 ?
no. 30 ?
reflected the degree of flexibility of the particular terminus of
the ORC subunit to which it was fused. The N and C termini of
each of the Orc3 and Orc4 subunits localized to similar sites;
however, this did not mean necessarily that they were in the same
positions (i.e., the subunit formed a hairpin-like structure),
because the images were 2D projections and therefore the N and
C termini could be localized in different planes and their
position within a 3D space be far apart from each other. MBPs
fused to Orc1, Orc4, and Orc5 localized to one side of the ORC,
and MBPs fused to Orc2 and Orc3 were located to the opposite
side of the ORC, indicating a polarity within the ORC.
ORC Subunit Interactions. To determine the subunit–subunit in-
teractions between the ORC subunits, genes encoding the
individual subunits were transcribed and the resulting mRNA
was translated in the presence of35S-methinoine. The labeled
proteins were immunoprecipitated either alone or in various
combinations (Fig. 4). Orc2 and Orc3 interacted with each other
in reciprocal immunoprecipitations (Ips) (Fig. 4 C and D, lanes
4). Orc4 bound to Orc5 (Fig. 4A, lane 16) but not to other
individual ORC subunits (Fig. 4A, lanes 13–15 and 17). The
Orc4/5 complex was able to interact with the Orc2/3 complex
(Fig. 4A, lane 21, long exposure) and separately with Orc1 (Fig.
4A, lane 22, long exposure), and with Orc2/3 and Orc1 to form
the Orc1-5 complex (Fig. 4A, lane 26). Consistent with these
complex in the presence of Orc1 (Fig. 4B, lanes 9–11). Orc1 did
not bind to the individual Orc4 or Orc5 subunits because the
bands present in Fig. 4B, lanes 4 and 5 were present in beads
alone (data not shown). For the same reason we could not test
whether Orc1 bound Orc2 or Orc3, but the Orc2/3 complex did
not bind to Orc1 (Fig. 4B, lane 6). Results for Orc6 binding in
these experiments were not consistent or were complicated by
Orc6 interacting with the beads in the absence of antibody; and
as a consequence, we did not interpret Orc6 interactions.
MBP Fusion Combined with 2D Image Classification Is an Effective
Approach for Mapping Needs in Molecular EM. Domain or subunit
mapping is essential in EM structural studies because of the
limited resolution. Antibody labeling or functionalized heavy
metal cluster labeling is often used (24, 25). The main difficulty
in using these noncovalent labels is low label occupancy, which
is further compounded by the flexible nature of the labels. To
overcome this difficulty, the use of proteins with encoded
molecular tags for EM contrasting was reported (26). The fusion
***** *** *
85 5897 11562105 567545 120
178274198 209184223230284 177355
340 102143165210132 170 132142112
271415 381 383414379349 383329451
MBP-fused ORC and the lower row shows the same images at higher contrast level (contrast ? 0.3). In the illustrations (Far Right), asterisks mark the observed
MBP locations relative to ORC. The number of particles used for calculating each average is indicated. 1C refers to the ORC with MBP fused at the C terminus
of Orc1 (ORC–Orc1–MBP). Likewise, the other complexes, 2C, 3N, 3C, 4N, 4C, and 5N also contain MBP at their N or C termini as indicated.
www.pnas.org?cgi?doi?10.1073?pnas.0803829105Chen et al.
protein approach eliminates the occupancy problem and was
used previously in the context of 3D difference mapping (27).
We found that the flexible fusion protein could be separated
efficiently into different 2D classes according to its positions by
reference-free 2D image classification. We demonstrated with
numerous MBP-fused ORCs that this approach is effective and
straightforward. We suggest that the approach of MBP fusion
combined with 2D image classification be added to the existing
toolkit for locating domains or subunits in large molecular
Flexible Domains in ORC. Our failure to detect MBP in MBP–
Orc1–Orc(2-6) and MBP–Orc2–Orc(1, 3-6) might be caused by
the flexibility of the N-terminal domains of Orc1 and Orc2.
Conceivably, a flexible MBP fused onto yet another flexible
domain would cause the MBP tag to move in too large a range
to be visible in averaged images. Similarly, the smaller than
expected difference density for Orc6 (Fig. 1C) is likely caused by
the partial flexibility of the subunit. Indeed, the N termini of
trypsin (data not shown), and these three subunits are the most
sensitive to proteases when ORC interacts with Cdc6 and DNA
(18). Secondary structure predictions also indicate that the N
termini of Orc1, Orc2, and Orc6 are enriched with flexible loop
regions (data not shown). The N terminus of Orc1 is involved in
interaction with Sir1 (28–30) and Cdc6 (31). These interactions
might require a certain degree of flexibility to achieve its
function in silencing. Based on these considerations, we suggest
that the proposed structure of ORC derived from the native
complex does not have all regions of the protein visible (Fig. 5
B and C).
ORC Architecture. Fig. 5 A and B summarizes our subunit inter-
these studies that Orc1, Orc4, and Orc5 occupy the upper half,
and Orc2, Orc3, and Orc6 occupy the lower half of the structure.
The two independent experimental approaches yielded consis-
tent results regarding the relative proximity of subunits to each
other. The derived ORC architecture is also in agreement with
earlier protein–DNA cross-linking experiments and recent yeast
++ + +
+++ + + + + + + + + + + + +
++ + + +
+ + +
Bead plus anti-Orc4 antibody
+ +++ ++
1 2 3 4 5 6 7 8 9 10 11
Bead plus anti-Orc1 antibody
1 2 3 4
Bead plus anti-
1 2 3 4
Bead plus anti-
1 2 3 4 5 6
7 8 9 1011
1214151617181920 21 132223 242526
ORC subunits, cloned into the transcription/translation vector (pCite-2a?),
were transcribed and the resulting mRNA was translated individually and in
various combinations, 5% of which was run on an SDS/PAGE gel as Input. The
remainder of the translation mix was then immunoprecipitated with 20 ?l
Gamma Bind G Sepharose and washed three times and 50% of the IP mix was
loaded onto SDS/PAGE gels. (A Lower and B–D) The IP of anti-Orc4 (A Lower),
anti-Orc1 (B), anti-Orc2 (C), and anti-Orc3 (D) antibodies are shown. (A Lower
Inset) A longer exposure of the gel, which was inserted over the top of the
lighter image to observed Orc1 protein. The Orc6 subunit was not observed
reliably in the IP lanes and hence we did not assign partners for this subunit.
In vitro subunit interaction. (A Upper) Genes encoding the individual
subunit was not included in the analysis. (B) A summary of subunit mapping
results. The approximate MBP position represents the location of the corre-
sponding N or C terminus of the subunit onto which the MBP was fused. For
an example, 1C refers to the average MBP position found in Orc(2-6)–Orc1–
0MBP. (C) A proposed model for ORC interaction with Cdc6 and origin DNA.
See Discussion for details.
The architecture of the yeast ORC. (A) A summary of ORC subunit
Chen et al.
July 29, 2008 ?
vol. 105 ?
no. 30 ?
two-hybrid experiments that suggest or demonstrate a direct
contact between Orc2p and Orc3p and between Orc4p and
Orc5p (21, 23, 32). It is significant to note that termini of each
of the five largest subunits of ORC are clustered to either the
upper right area or the lower left area of ORC within the ORC
structure (Fig. 5C). The C-terminal part of Orc1-5 contains the
proposed DNA-binding winged helix domain (WHD), and N
terminal to the WHD is a proposed domain belonging to the
AAA? complexes is the location of ATP binding sites at the
interface of two different subunits. The interface between
subunits frequently contains an arginine finger in one subunit
that interacts with the ATP bound by the adjacent subunit. It has
been shown that an arginine finger in Orc4 is required to
stimulate the ATPase within Orc1 (33). This result indicates that
Orc1 and Orc4 interact directly and are thus located adjacent to
each other, a result consistent with our proposed structure (Fig.
a complex with Orc5 (34–37), indicating that an Orc1–Orc4–
Orc5 complex interacts with Orc2–Orc3 via Orc5, and that the
Orc2-5 complex binds Orc1, suggesting an order in the complex
of Orc1,4,5,2/3. This arrangement is also consistent with a
low-resolution EM structure of ORC from Drosophila where
Orc5 was localized with the help of an antibody in the center of
the elongated complex (9). The relative order of Orc2/3 in our
proposed structure is not possible to assign based on existing
recent yeast two-hybrid data show an interaction between Orc2
and Orc5 (32), suggesting that hints that Orc3 might represent
the one end of the ORC if the AAA? subunits are arranged in
pseudolinear arrangement in the structure.
A Model for Interaction Between ORC, Cdc6, and Origin DNA. The
DNaseI footprint analysis suggests a protection of 48 bp of ARS1
origin DNA by the yeast ORC in the presence of ATP (19, 20).
The dimension of our EM structure of ORC is large enough to
accommodate five AAA? domains with WHD, and the 16-nm
length is appropriate for covering the 48-bp dsDNA. To estimate
the actual DNA contact sites, however, 3 to 5 bp from the ends
of the footprint should be subtracted, because DNase1 cannot
access the actual DNA binding site due to steric hindrance. Thus,
the protein–DNA contacts would likely cover ?38–44 bp of
a WHD and the other from the initiator-specific motif (ISM) in
the ?/? subdomain of the AAA? domain (14, 17). In the
protein–DNA structure, the WHD binding sites of Orc1–1 and
Orc1–3 are 9 bp apart, but at the secondary ISM DNA binding
sites these two Orc1 subunits are only 3 bp apart. It is uncertain
how a hypothetical third ORC subunit would bind to DNA if the
archaeal Orc1–DNA structure was extended by one subunit.
Because the WHD is the major DNA binding element, we
assume that the WHD defines the spacing between the neigh-
boring subunits. If this assumption were incorporated into a
model for the multisubunit eukaryotic ORC, it would result in
four adjacent WHD-containing origin-binding subunits in ORC
[3 ? 9 ? 15 ? 42 bp, where the 15 bp is the actual DNA-
contacting length of each subunit, as estimated from the Orc1–
DNA structure (14, 17)]. ORC is predicted to contain five
WHD-containing subunits, and we suggest that the fifth WHD
in ORC and the Cdc6 WHD might bind DNA when ORC forms
a complex with Cdc6, thereby greatly extending the interaction
with DNA to ?80 bp (10).
In the archaeal Orc1–DNA structure, Orc1-1 and Orc1-3 are
staggered ?60° around the DNA axis (14, 17). If this mode of
interaction is conserved in the eukaryotic ORC, we suggest a
model for ORC in which the origin DNA runs along its length
(Fig. 5C). In this model, the above-estimated four major DNA-
binding WHDs are ?60° apart along the DNA axis, resulting in
the first two C-terminal WHD (0° and 60°, respectively) on one
side and at one end, with the remaining two C-terminal WHD
end of the proposed structure. This arrangement agrees with the
observed clustering of C-terminally fused MBP to the lower left
and upper right sides of the ORC structure.
that Orc1 and Orc4 bind near the A element and Orc2 and Orc3
bind near the B1 element of the ARS1 origin (21). These data
suggest that the A element binds to the upper half of ORC
where Orc1 and Orc4 are located and B1 binds to the lower
half where Orc2 and Orc3 reside. Such a DNA orientation would
point Orc6, located at the bottom of ORC, toward B2 (Fig. 5C),
in agreement with the knowledge that Orc6 interacts with DNA
near B2. We previously showed that Cdc6 binds to the left side
of ORC and Cdc6 caused a profound lengthening of the DNA
contact (10). We speculate that B2 might bend back onto ORC
in the presence of Cdc6, which also contains a WHD near its C
terminus (Fig. 5C). It was shown that the putative replicative
helicase MCM loads at the B2 region and that Cdt1, the MCM
loading protein, interacts with Orc6 (22, 38, 39).
To summarize, our current understanding of ORC is that the
primary recognition is at the ARS1 A element, with a secondary
interaction with the B1 element. This bipartite recognition
determines, in part, the origin specificity of ORC and points the
non-AAA? domains in Orc6 and Orc2 toward the ARS1 B2
element. Cdc6 interaction with ORC extends the DNA interac-
tion toward the B2 element where Cdt1 cooperates to load the
MCM proteins. The overall ring-shaped structure of the ORC–
Cdc6 complex may facilitate interactions with the hexameric,
ring-shaped MCM proteins.
Materials and Methods
Preparation of the MBP-Fused ORCs. To generate ORCs with MBP at its C or N
terminus, the MBP coding region was amplified along with a 9-aa linker
(AAAAAIDTT) that was fused on either the C or N terminus of each ORC
subunit. This modified ORC subunit along with the remaining five nontagged
subunits were expressed from recombinant baculovirus vectors in insect cells
as described (21). The proteins were purified as described (10).
Gel-Shift Assay. The gel-shift assay was performed as described (10) using 0.2
nM ARS1 DNA, 2 nM ORC, ORC–MBP fusion complexes, or Orc1-5, and 0.8 or
2.4 nM Cdc6.
ORC Subunit Interactions. Genes encoding S. cerevisiae ORC subunits were
cloned individually by PCR primer extension and ligated into pCite-
2a(?)(Novagen/EMD) at the SacI/XhoI sites. Transcription/translation was car-
the standard reaction mix and 1 ?g of DNA per sample. mAbs [Orc1 (SB13),
Orc2 (SB67), Orc3 (SB3), Orc4 (SB6), Orc5 (SB5) and Orc6 (SB49)] were ?10
mg/ml by protein gel analysis and were used at 2.5 mg per IP reaction.
Translation products were diluted 1:5 in buffer [20 mM Hepes-KOH (pH 7.0),
75 mM NaCl, 0.02% Nonidet P-40, 5 mM MgAc, 5 mM ?-mercaptoethanol, 1
mM ATP (pH 7.0), 10% glycerol and Complete-EDTA Free Protease Inhibitor
tablets (Roche)], combined with clarified mAbs and immunoprecipitated by
and 50% IP) were visualized after electrophoresis through 10% acrylamide/
0.3%Bis SDS/PAGE gels, dried, and exposed to imaging film.
EM. The concentrations of ORC and its derivatives were adjusted to ?0.1
mg/ml in buffer containing 50 mM Hepes-KOH (pH 7.6), 100 mM potassium
glutamate, 5 mM MgCl2, 1 mM EGTA, and 1 mM ATP?S, and the diluted
applied to a glow-discharged 300-mesh copper grid covered with a thin layer
of carbon film, and after a brief incubation of 30–60 s, the excess sample
solution was blotted with a small piece of filter paper. The grid was then
stained in a deep stain procedure by three consecutive 5-?l drops of 2.0%
temperature before blotting. After blotting the last stain drop, the grid was
quickly dried by a stream of argon to prevent crystallization of the stain salt.
Micrographs of negatively stained specimens were recorded at a magnifica-
www.pnas.org?cgi?doi?10.1073?pnas.0803829105Chen et al.
tion of ?50,000 in a JEOL JEM-1200EX transmission EM, operated at an Download full-text
acceleration voltage of 120 kV. All images were recorded on Kodak SO-163
film by using the minimum-dose procedure at the defocus ranging from ?1.2
developer at 20°C. Selected micrographs were digitized with a Nikon Super-
cool Scan 8000ED using a step size of 12.7 ?m.
2D Image Classification and 3D Image Reconstruction. We used SPIDER (40),
raw particles in each data set was 22,238 (1N), 14,197 (1C), 16,780 (2N), 6,824
(2C), 10,762 (3N), 11,419 (3C), 25,578 (4N), 16,521 (4C), 12,178 (5N), and 8,935
(Orc1-5). The parameters of contrast transfer function for each micrograph
were calculated, and these parameters were used for phase flipping. Raw
particle images were band-pass-filtered to exclude the very low- and high-
frequency noise, normalized, centered, and then subjected to multivariate
statistical analysis (MSA) and classification. Several well centered class aver-
ages were selected as references to realign all of the raw particle images. The
recentered particle images were reclassified. Multireference alignment and
classification were carried out for several iterations by using well defined and
properly centered class averages from the previous iteration as references
until stable class averages were obtained. In a second approach, the MSA and
classification were simply carried out by ‘‘2drefine’’ in EMAN, (which incor-
a single batch script), with an initial class number of 100–150, depending on
the data set. The two approaches yielded comparable class averages. The 3D
map of the Orc1-5 subcomplex was obtained by refinement in EMAN by using
a low-pass-filtered ORC 3D map (30 Å) as the starting model (Speck 2005
NSMB). The 3D difference map was calculated by subtracting the Orc1-5 3D
map from the ORC 3D map, after proper scaling. The 3D maps were rendered
into surface views with Chimera (43).
ACKNOWLEDGMENTS. This work was supported by National Institutes of
Health Grants GM45436 (to B.S.) and GM74985 (to H.L.) and Brookhaven
National Laboratory Laboratory-Directed Research and Development Project
06-06 (to H.L.). C.S. was a fellow of the Leukemia and Lymphoma Society.
by multiple functional elements. Science 255:817–823.
binding site within yeast replicators. Proc Natl Acad Sci USA 92:2224–2228.
3. Rowley A, Cocker JH, Harwood J, Diffley JF (1995) Initiation complex assembly at
budding yeast replication origins begins with the recognition of a bipartite sequence
by limiting amounts of the initiator, ORC. EMBO J 14:2631–2641.
4. Bell SP, Dutta A (2002) DNA replication in eukaryotic cells. Annu Rev Biochem 71:333–
5. Diffley JF (2004) Regulation of early events in chromosome replication. Curr Biol
6. Mendez J, Stillman B (2003) Perpetuating the double helix: Molecular machines at
eukaryotic DNA replication origins. BioEssays 25:1158–1167.
8. Klemm RD, Austin RJ, Bell SP (1997) Coordinate binding of ATP and origin DNA
regulates the ATPase activity of the origin recognition complex. Cell 88:493–502.
9. Clarey MG, et al. (2006) Nucleotide-dependent conformational changes in the DnaA-
like core of the origin recognition complex. Nat Struct Mol Biol 13:684–690.
and Cdc6 to origin DNA. Nat Struct Mol Biol 12:965–971.
11. Erzberger JP, Mott ML, Berger JM (2006) Structural basis for ATP-dependent DnaA
assembly and replication-origin remodeling. Nat Struct Mol Biol 13:676–683.
12. Mott ML, Berger JM (2007) DNA replication initiation: Mechanisms and regulation in
bacteria. Nat Rev Microbiol 5:343–354.
13. Barry ER, Bell SD (2006) DNA replication in the archaea. Microbiol Mol Biol Rev
14. Dueber EL, Corn JE, Bell SD, Berger JM (2007) Replication origin recognition and
deformation by a heterodimeric archaeal Orc1 complex. Science 317:1210–1213.
15. Singleton MR, et al. (2004) Conformational changes induced by nucleotide binding in
Cdc6/ORC from Aeropyrum pernix. J Mol Biol 343:547–557.
16. De Felice M, et al. (2004) Modular organization of a Cdc6-like protein from the
crenarchaeon Sulfolobus solfataricus. Biochem J 381:645–653.
17. Gaudier M, Schuwirth BS, Westcott SL, Wigley DB (2007) Structural basis of DNA
replication origin recognition by an ORC protein. Science 317:1213–1216.
18. Mizushima T, Takahashi N, Stillman B (2000) Cdc6p modulates the structure and DNA
binding activity of the origin recognition complex in vitro. Genes Dev 14:1631–1641.
19. Speck C, Stillman B (2007) Cdc6 ATPase activity regulates ORC ? Cdc6 stability and the
selection of specific DNA sequences as origins of DNA replication. J Biol Chem
20. Bell SP, Stillman B (1992) ATP-dependent recognition of eukaryotic origins of DNA
replication by a multiprotein complex. Nature 357:128–134.
21. Lee DG, Bell SP (1997) Architecture of the yeast origin recognition complex bound to
origins of DNA replication. Mol Cell Biol 17:7159–7168.
22. Chen S, de Vries MA, Bell SP (2007) Orc6 is required for dynamic recruitment of Cdt1
during repeated Mcm2 7 loading. Genes Dev 21:2897–2907.
23. Chastain PD, 2nd, Bowers JL, Lee DG, Bell SP, Griffith JD (2004) Mapping subunit
location on the Saccharomyces cerevisiae origin recognition complex free and bound
to DNA using a novel nanoscale biopointer. J Biol Chem 279:36354–36362.
24. Boisset N, et al. (1988) Intramolecular localization of epitopes within an oligomeric
protein by immunoelectron microscopy and image processing. Proteins 3:161–183.
25. Hainfeld JF (1987) A small gold-conjugated antibody label: Improved resolution for
electron microscopy. Science 236:450–453.
26. Mercogliano CP, DeRosier DJ (2006) Gold nanocluster formation using metal-
lothionein: Mass spectrometry and electron microscopy. J Mol Biol 355:211–223.
27. Liu Z, et al. (2001) Three-dimensional reconstruction of the recombinant type 3
ryanodine receptor and localization of its amino terminus. Proc Natl Acad Sci USA
28. Hou Z, Bernstein DA, Fox CA, Keck JL (2005) Structural basis of the Sir1-origin recog-
nition complex interaction in transcriptional silencing. Proc Natl Acad Sci USA
29. Hsu HC, Stillman B, Xu RM (2005) Structural basis for origin recognition complex 1
protein-silence information regulator 1 protein interaction in epigenetic silencing.
Proc Natl Acad Sci USA 102:8519–8524.
30. Triolo T, Sternglanz R (1996) Role of interactions between the origin recognition
complex and SIR1 in transcriptional silencing. Nature 381:251–253.
31. Wang B, et al. (1999) The essential role of Saccharomyces cerevisiae CDC6 nucleotide-
Saccharomyces cerevisiae: Interaction between subunits and identification of binding
proteins. FEMS Yeast Res 7:1263–1269.
Mcm2–7 assembly at a defined origin of replication. Mol Cell 16:967–978.
34. Dhar SK, Delmolino L, Dutta A (2001) Architecture of the human origin recognition
complex. J Biol Chem 276:29067–29071.
35. Ranjan A, Gossen M (2006) A structural role for ATP in the formation and stability of
the human origin recognition complex. Proc Natl Acad Sci USA 103:4864–4869.
complex. J Biol Chem 282:32370–32383.
37. Vashee S, Simancek P, Challberg MD, Kelly TJ (2001) Assembly of the human origin
recognition complex. J Biol Chem 276:26666–26673.
Sci USA 99:101–106.
39. Zou L, Stillman B (2000) Assembly of a complex containing Cdc45p, replication protein
A, and Mcm2p at replication origins controlled by S-phase cyclin-dependent kinases
and Cdc7p-Dbf4p kinase. Mol Cell Biol 20:3086–3096.
40. Frank J, et al. (1996) SPIDER and WEB: processing and visualization of images in 3D
electron microscopy and related fields. J Struct Biol 116:190–199.
41. Ludtke SJ, Baldwin PR, Chiu W (1999) EMAN: Semiautomated software for high-
resolution single-particle reconstructions. J Struct Biol 128:82–97.
42. van Heel M, Harauz G, Orlova EV, Schmidt R, Schatz M (1996) A new generation of the
IMAGIC image processing system. J Struct Biol 116:17–24.
43. Pettersen EF, et al. (2004) UCSF Chimera: A visualization system for exploratory
research and analysis. J Comput Chem 25:1605–1612.
Chen et al.
July 29, 2008 ?
vol. 105 ?
no. 30 ?