Mutations in the M112/M113-coding region facilitate murine cytomegalovirus replication in human cells.

JOURNAL OF VIROLOGY, Aug. 2010, p. 7994–8006 Vol. 84, No. 16
0022-538X/10/$12.00 doi:10.1128/JVI.02624-09
Copyright © 2010, American Society for Microbiology. All Rights Reserved.
Mutations in the M112/M113-Coding Region Facilitate Murine
Cytomegalovirus Replication in Human Cells�
Uwe Schumacher, Wiebke Handke, Igor Jurak,§ and Wolfram Brune*
Division of Viral Infections, Robert Koch Institute, Nordufer 20, 13353 Berlin, Germany
Received 15 December 2009/Accepted 21 May 2010
Cytomegaloviruses, representatives of the Betaherpesvirinae, cause opportunistic infections in immunocom-
promised hosts. They infect various cells and tissues in their natural host but are highly species specific. For
instance, human cytomegalovirus (HCMV) does not replicate in mouse cells, and human cells are not
permissive for murine cytomegalovirus (MCMV) infection. However, the underlying molecular mechanisms are
so far poorly understood. In the present study we isolated and characterized a spontaneously occurring MCMV
mutant that has gained the capacity to replicate rapidly and to high titers in human cells. Compared to the
parental wild-type (wt) virus, this mutant formed larger nuclear replication compartments and replicated viral
DNA more efficiently. It also disrupted promyelocytic leukemia (PML) protein nuclear domains with greater
efficiency but caused less apoptosis than did wt MCMV. Sequence analysis of the mutant virus genome revealed
mutations in the M112/M113-coding region. This region is homologous to the HCMV UL112-113 region and
encodes the viral early 1 (E1) proteins, which are known to play an important role in viral DNA replication.
By introducing the M112/M113 mutations into wt MCMV, we demonstrated that they are sufficient to facilitate
MCMV replication in human cells and are, at least in part, responsible for the efficient replication capability
of the spontaneously adapted virus. However, additional mutations probably contribute as well. These results
reveal a previously unrecognized role of the viral E1 proteins in regulating viral replication in different cells
and provide new insights into the mechanisms of the species specificity of cytomegaloviruses.
Cytomegaloviruses (CMVs) are prototypes of the � subfam-
ily of the Herpesviridae. Representatives of this subfamily have
been identified in various animal species, and these viruses
cause similar symptoms in their respective hosts (36). HCMV
is an opportunistic pathogen that causes generally mild infec-
tions in people with a fully functional immune system. How-
ever, this virus is also responsible for serious medical problems,
particularly in newborns and immunocompromised patients
(39).
Since their first isolation in cell culture, CMVs have been
recognized as highly species specific (57). They replicate only
in cells of their own or a closely related species. For instance,
simian CMV can replicate in human fibroblasts (32), and
HCMV can replicate in chimpanzee skin fibroblasts (41). Sim-
ilarly, murine cytomegalovirus (MCMV) productively infects
rat cells (7, 46), but a rat cytomegalovirus did not replicate in
murine fibroblasts (7). However, cells of other more distantly
related species are usually nonpermissive to infection. Several
studies have shown that CMVs can enter cells of other species
and express a subset of viral genes (19, 20, 29, 32). This finding
has led to the conclusion that the restriction to CMV replica-
tion in nonpermissive cells is associated with a postpenetration
block to viral gene expression and DNA replication but not
due to a failure to enter the cell (36).
Recently, we picked up on this topic and tried to gain new
insights into the molecular mechanisms underlying the species
specificity of CMVs. We showed that CMVs of mice and rats
induce apoptosis when they infect human fibroblasts or retinal
epithelial cells (26). The induction of apoptosis prevented a
sustained replication of these viruses in human cells and re-
duced progeny production to insignificant levels. When apop-
tosis was inhibited by the overexpression of Bcl-2 or a func-
tionally similar protein, MCMV was able to replicate to
substantial titers in human cells. These results indicated that
the induction of apoptosis is an important limitation to cyto-
megalovirus cross-species infections (26). However, the fact
that MCMV replication in human cells in the presence of
apoptosis inhibition was somewhat delayed and less efficient
than that in murine cells indicated that other limiting factors
likely exist. Another study suggested that MCMV can replicate
to low levels in human cells with the help of HCMV immedi-
ate-early 1 (IE1) and HCMV tegument proteins (50).
In the present study, we describe the isolation and char-
acterization of a mutant MCMV that has spontaneously
acquired the ability to replicate rapidly and to high titers in
human retinal pigment epithelial (RPE-1) cells. We show
that this virus induces less apoptosis and replicates its DNA
faster than the parental wild-type (wt) MCMV. Moreover,
the mutant virus disrupts intranuclear sites of intrinsic an-
tiviral defense more efficiently than the wt virus. Sequence
analysis of the human cell-adapted MCMV strain revealed
several alterations, including mutations in the M112/M113-
coding region. By targeted mutagenesis we showed that mu-
tations in M112/M113 are sufficient to facilitate MCMV
replication in human cells. However, additional mutations
most likely contribute to the remarkably efficient replication
of the adapted strain.
* Corresponding author. Present address: Heinrich Pette Institute for
Experimental Virology and Immunology at the University of Hamburg,
Martinstr. 52, 20251 Hamburg, Germany. Phone: 49 40 48051-351. Fax: 49
40 48051-103. E-mail: wolfram.brune@hpi.uni-hamburg.de.
§ Present address: Department of Biological Chemistry and Molec-
ular Pharmacology, Harvard Medical School, Boston, MA 02115.
� Published ahead of print on 2 June 2010.
7994

MATERIALS AND METHODS
Cells and viruses. hTERT RPE-1 (ATCC CRL-4000) cells are telomerase-
immortalized retinal pigment epithelial cells (6). MRC-5 cells (ATCC CCL-171)
are primary human embryonic lung fibroblasts. 10.1 cells are spontaneously
immortalized murine embryonic fibroblasts (22). Primary human foreskin fibro-
blasts (HFFs) were a gift from Jens von Einem (University of Ulm, Ulm, Ger-
many), and primary human umbilical vein endothelial cells (HUVECs) were
provided by Regine Heller (University of Jena, Jena Germany). HUVECs were
cultured in complete endothelial cell growth medium (Promocell). All other cells
were cultured in complete Dulbecco’s modified Eagle’s medium (DMEM) sup-
plemented with 10% fetal calf serum at 37°C and 5% CO2 in a humidified
atmosphere. wt and mutant MCMVs were propagated in 10.1 fibroblasts as
described previously (26). MCMV/h was propagated in human RPE-1 cells.
Bacterial artificial chromosome (BAC)-derived HCMV strains AD169 and
TB40/E (23, 45) were propagated on MRC-5 fibroblasts. MCMV titers were
determined on mouse 10.1 cells and HCMV titers were determined on MRC-5
cells by using the median tissue culture infective dose (TCID50) method (33).
Plasmids and transfection. The M112/M113-coding sequence plus �50-nucle-
otide (nt) extensions on either side (for homologous recombination) were PCR
amplified from wt MCMV and MCMV/h by using primers M112_Not_fwd
(GTT-CCT-GCG-GCC-GCG-GTA-GAT-TAC-GTG-CCC-ACT-TTTC) and
M113_Pst_Mfe_rev (CAA-CTG-CAG-TCA-GTT-AGA-GTT-TAC-AGA-GCA-
TCA-TTT-CTT-TAT-CCA-TCT-TTCAATTGA-GAT-CAA-TTA-AGA-TCA-
TCG-AACACA). The amplification products were cleaved with NotI and PstI
(underlined), cloned into pBluescript KS(�), and verified by sequencing. An Flp
recombination target (FRT)-flanked kanamycin resistance gene (excised with
EcoRI from pSLFRTkn [5]) was cloned into the MfeI site (underlined) located
behind the M112/M113-coding region, yielding plasmids pBS-M112-kn and pBS-
M112mut-kn. For protein expression by transient transfection, the M112/M113-
coding sequence (without the kan cassette) was excised with NotI and PstI from
the pBluescript backbone and inserted into pFlagCMV5a (Sigma). Plasmid
pp89UC, encoding MCMV IE1, and pcDNA-IE1, encoding HCMV IE1, were
kindly provided by Martin Messerle (Hannover Medical School, Hannover, Ger-
many) and Michael Nevels (University of Regensburg, Regensburg, Germany),
respectively. The green fluorescent protein (GFP) expression plasmid
pEGFP-C1 was obtained from Clontech. Plasmid transfections were done by
using Polyfect transfection reagent (Qiagen).
Generation of gene knockdown cells. Plasmids carrying the short hairpin RNA
(shRNA) expressing retroviral vectors pHM2237/empty, pHM2238/shC,
pHM2240/shDaxx1, and pHM2243/shPML2 (52) were kindly provided by
Thomas Stamminger and Nina Tavalai (University of Erlangen, Erlangen, Ger-
many). Retroviruses were produced by transfecting the vector plasmids into the
Phoenix Ampho packaging cell line by calcium phosphate transfection, as de-
scribed previously (48), and were used for the transduction of RPE-1 cells.
Nonclonal cell populations stably expressing shRNA were obtained by selection
with 5 �g/ml puromycin for 7 days.
Antibodies and immunodetection. Monoclonal antibodies against the MCMV
IE1 (CROMA101) and early 1 (E1) (CROMA103) proteins were provided by
Stipan Jonjic (University of Rijeka, Rijeka, Croatia), monoclonal antibodies
against M44 (3B9.22A) and MCMV glycoprotein B (gB) (2E8.21A) were pro-
vided by Lambert Loh (University of Saskatchewan, Saskatoon, Saskatchewan,
Canada), and monoclonal antibodies against HCMV IE1 (1B12) were provided
by Thomas Shenk (Princeton University, Princeton, NJ). A polyclonal rabbit
antiserum against MCMV IE3 was provided by Eva Borst (Hannover Medical
School, Hannover, Germany). Antibodies recognizing the promyelocytic leuke-
mia (PML) protein (H-238; Santa Cruz Biotechnology), hDaxx (E94; Epitomics),
and �-actin (AC-74; Sigma) were purchased from the indicated suppliers.
For immunofluorescence analyses cells were seeded onto coverslips on the day
before infection/transfection. After 24 h, cells were washed twice with phosphate-
buffered saline (PBS), fixed for 30 min at 4°C with 4% paraformaldehyde,
neutralized with 50 mM ammonium chloride, permeabilized with 0.3% Triton
X-100, and blocked with 0.2% cold-water fish gelatin (Sigma). Proteins of inter-
est were detected by indirect immunofluorescence using secondary antibodies
coupled to Alexa Fluor 568 or Alexa Fluor 488 (Invitrogen). Nuclei were coun-
terstained with 4�,6�-diamidino-2-phenylindole (DAPI). Confocal laser scanning
microscopy was performed by using a Zeiss LSM510 Meta microscope. To
analyze PML disruption or the formation of replication compartments, at least
three different experiments were done, and a minimum of 150 infected cells from
each experiment were evaluated.
For Western blot analyses, cells were infected at a multiplicity of infection
(MOI) of 5 TCID50/cell, harvested at the indicated time points, and lysed with
lysis buffer containing 20 mM Tris-HCl (pH 7.5), 300 mM NaCl, 1% Na-deoxy-
cholate, 1% Triton X-100, and 0.1% SDS. Proteins samples were separated by
SDS-PAGE and transferred onto positively charged nitrocellulose membranes.
Proteins of interest were detected by using protein-specific primary antibodies,
horseradish peroxidase-coupled secondary antibodies (Dako Cytomation), and
enhanced chemiluminescence (ECL) reagents (Amersham).
BAC mutagenesis. All recombinant viruses were constructed on the basis of
MCMV-GFP (9) by using BAC technology (11). To construct the �M112/M113
mutant, a zeocin resistance gene was PCR amplified by using primers
M112_zeo_fwd (5�-ACG-TGC-CCA-CTT-TTC-TCG-TCG-CGA-CCG-GTG-
AAA-AGA-CCT-TCG-TTC-GGA-CCT-gtt-gac-aat-taa-tca-tcg-gcat-3�) and
M113_zeo_rev (5�-AGT-CAG-TTA-GAG-TTT-ACA-GAG-CAT-CAT-TTC-
TTT-ATC-CATCTTT-CAT-GAG-At-cag-tcc-tgc-tcc-tcg-gcca-3�) to introduce
50-nt homology arms (shown in uppercase type) upstream and downstream of
the M112/M113-coding region. This linear PCR fragment was used for homol-
ogous recombination in Escherichia coli strain DY380 containing the MCMV-
GFP BAC (10). Mutant BACs were analyzed by restriction digestion and agarose
gel electrophoresis. In a second step, linear fragments containing the wt or
mutant M112/M113-coding sequence and the kanamycin cassette were excised
with NotI and PstI from plasmids pBS-M112-kn and pBS-M112mut-kn, respec-
tively, and used for homologous recombination. Clones were selected with kana-
mycin, and recombinant BACs were analyzed by restriction digestion. The kana-
mycin cassette was removed by using Flp recombinase as previously described
(10). An independent M112/M113 mutant virus was constructed by using the
galK system (56). Briefly, an �M112/M113 mutant was constructed as described
above by using galK instead of zeo for positive selection. In a second step, the
mutant M112/M113 sequence was reinserted by using galK for negative selection.
Mutant and control viruses were reconstituted by transfecting purified BAC
DNA into mouse fibroblasts using Polyfect transfection reagent (Qiagen).
Genome sequencing. MCMV/h was derived from a single plaque of infected
RPE-1 cells and was passaged continuously on RPE-1 cells. For the preparation
of MCMV/h virion DNA, viral particles in the supernatant of infected RPE-1
cultures were pelleted by centrifugation for 3 h at 25,000 � g. Pellets were
resuspended in a solution containing 10 mM Tris-HCl (pH 7.8), 5 mM EDTA,
and 0.5% SDS; digested overnight with 50 �g/ml proteinase K at 56°C; and finally
subjected to phenol-chloroform extraction and DNA precipitation. Purified
DNA was analyzed by restriction digestion. High-quality virion DNA was sent to
Macrogen for shotgun sequencing. The resulting contigs were aligned, and the
remaining gaps were filled by primer walking. In this way, almost the entire
MCMV/h genome was sequenced save for a few kilobases at the genome termini.
Growth kinetics. For growth kinetics, cells were seeded into six-well dishes and
infected with the viruses of interest. At 4 h postinfection (hpi), the medium was
removed, cells were washed twice with PBS, and fresh medium was added. At the
indicated time points, titers in the supernatant were determined by using the
TCID50 method. All growth kinetic experiments were done in triplicate.
Southern blot analysis and real-time PCR. Cells were infected at an MOI of
5 TCID50/cell for 4 h. At the indicated time points, adherent and detached cells
were collected, and complete genomic DNA was extracted by proteinase K
digestion and phenol-chloroform extraction. For slot blot analysis, 1 �g of DNA
of each sample was denatured for 10 min at 95°C in 6� SSC (0.9 M NaCl, 90 mM
Na-citrate [pH 7.0]), cooled rapidly on ice, transferred onto a nylon membrane
by vacuum suction using a slot blot apparatus (Roth), and fixed by using a UV
cross-linker (Stratagene). For Southern blot analysis, equal amounts of DNA
were digested with EcoRI, separated by gel electrophoresis, and blotted onto a
nylon membrane. Hybridization with a digoxigenin (DIG)-labeled MCMV M45
probe and detection by ECL were done by using a DIG-High Prime DNA
labeling and detection kit (Roche).
For real-time PCR analysis cells were infected at an MOI of 0.5 TCID50/cell,
and DNA was extracted at the indicated time points by using the DNeasy blood
and tissue kit (Qiagen). A SYBR green PCR master mix (ABI Applied Biosys-
tems) was used for the quantitative detection of the cellular myc gene and the
viral IE1 gene. The appropriate primers were described previously (8, 55). For
the determination of viral and cellular genome copies, defined dilutions of
plasmids containing the cloned PCR products were used. Genome copies were
determined in triplicates from three separate infection assays.
Cell viability and cell death assays. Cell viability was measured by using a
3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-
tetrazolium (MTS) assay (CellTiter 96 AQueous; Promega). Mean values and
standard deviations of data from at least six parallel experiments are shown.
Significance levels were calculated by using a Student’s t test. To analyze DNA
fragmentation as a sign of apoptosis, cells were grown on coverslips and infected
at an MOI of 5 TCID50/cell. Nuclear DNA fragmentation was detected by using
a terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling
VOL. 84, 2010 E1 MUTATIONS MEDIATE MCMV REPLICATION IN HUMAN CELLS 7995

(TUNEL) assay (TMR red in situ cell death detection kit; Roche) and evaluated
by using a Zeiss Axiovert 200 M epifluorescence microscope.
RESULTS
Isolation of a mutant MCMV spontaneously adapted to
human cells. In a previous study we showed that MCMV does
not replicate to significant levels in human cells and that in-
fected human cells undergo apoptosis. The inability of MCMV
to replicate and spread in human cells could be overcome by
overexpressing a Bcl-2-like antiapoptotic protein (26). During
one of our experiments, we fortuitously observed the sponta-
neous outgrowth of a rapidly spreading virus in a cell culture
dish of human retinal pigment epithelial (RPE-1) cells that had
been infected 2 weeks earlier with MCMV-GFP, a GFP-ex-
pressing MCMV variant. We named this spontaneously arisen
mutant MCMV/h to indicate its adaptation to human cells.
MCMV/h expressed GFP, caused a typical cytopathic effect
(CPE), formed plaques, and could be serially passaged on
RPE-1 cells. Moreover, it grew much faster and to higher titers
than MCMV-bclXL (26), a mutant MCMV that overexpresses
Bcl-xL, in RPE-1 cells, human embryonic lung fibroblasts
(MRC-5), human foreskin fibroblasts (HFFs), and human um-
bilical vein endothelial cells (HUVECs) (Fig. 1A to F). In
MRC-5 fibroblasts, MCMV/h did not replicate as efficiently as
in RPE-1 cells upon low-MOI infection (Fig. 1D). By micro-
scopic inspection we observed that MCMV/h was capable of
spreading from the initially infected fibroblasts to neighboring
cells, forming foci and even small plaques of infected (GFP-
positive) cells. After a few days, these foci stopped expanding
and gradually regressed (data not shown). The same replica-
tion phenotype was observed for HFFs, and the virus repli-
cated to almost identical titers in these cells (data not shown).
The reason for this limited replication in human fibroblasts
after low-MOI infection is so far unknown. In murine fibro-
blasts, however, the virus replicated to titers similar to those of
parental MCMV-GFP even after low-MOI infection (Fig. 1G).
MCMV/h replicated only transiently in HUVECs after high-
MOI infection (Fig. 1F) and to titers at or below the detection
limit after low-MOI infection (data not shown).
Virion DNA was extracted from an MCMV/h stock grown
on RPE-1 cells and subjected to EcoRI and NheI restriction.
The restriction patterns of MCMV/h DNA were very similar to
the patterns of MCMV-GFP. However, a few differences were
evident (Fig. 1H), suggesting that genetic alterations had oc-
curred. The transfection of purified MCMV/h DNA into
RPE-1 cells resulted in a reconstitution of infectious
MCMV/h. This largely ruled out the possibility that the repli-
cation of MCMV/h in human cells was caused by the presence
of an undetected helper RNA virus, as RNA viruses would not
withstand the DNA purification procedure. To exclude a pos-
sible contamination with HCMV, BAC DNA of HCMV labo-
ratory strain AD169 was included. The digestion pattern of
MCMV/h is completely different from that of AD169. Hence,
the adapted virus is not an HCMV.
Phenotypic characterization of MCMV/h. We have previ-
ously described that MCMV induces apoptosis in infected hu-
man cells (26). Compared to wt MCMV, MCMV/h induced
less apoptosis, as seen morphologically by microscopic inspec-
tion as well as by analyzing nuclear DNA fragmentation with a
TUNEL assay (Fig. 2A and B). MCMV/h decreased the via-
bility of infected RPE-1 cells, as measured by an MTS assay,
but the loss of viability was less pronounced than that after
infection with the parental virus, MCMV-GFP (Fig. 2C). The
viability of murine fibroblasts remained largely unaltered be-
tween 24 and 72 hpi, no matter whether the cells had been
infected with wt or mutant MCMVs (Fig. 2D).
Next, we analyzed viral gene expression in human and mu-
rine cells. The expressions of the immediate-early 1 (IE1),
early 1 (E1), M44, and glycoprotein B (gB) proteins as repre-
sentatives of the immediate-early, early, early-late, and late
kinetic classes, respectively, were measured by Western blot-
ting at different times postinfection. In murine fibroblasts, viral
protein expression levels in cells infected with MCMV/h were
very similar to those in cells infected with the parental virus
with the exception of gB, whose expression level rose slightly
earlier in MCMV/h-infected cells (Fig. 3A). In RPE-1 cells, the
MCMV/h protein expression pattern was quite similar to the
one observed for 10.1 fibroblasts. However, M44 expression
was delayed in RPE-1 cells, and gB expression was lower at 24
hpi. In contrast, viral protein expression vanished in wt
MCMV-infected cells between 48 and 72 hpi, because infected
cells had died (Fig. 3A), consistent with previous observations
(26). Viral DNA replication was observed only transiently in wt
MCMV-infected RPE-1 cells but occurred efficiently in
MCMV/h-infected cells, as determined by slot blot analysis
and real-time PCR (Fig. 3B and C).
Sequence analysis of MCMV/h reveals genetic alterations.
We wondered whether a genetic alteration or an epigenetic
modification of the MCMV genome was responsible for the
ability of MCMV/h to replicate efficiently in human cells. As
MCMV/h retained its ability to replicate in RPE-1 cells even
after repeated passage in murine fibroblasts (data not shown),
we considered it unlikely that an epigenetic modification,
which might have occurred in human cells, was responsible.
Moreover, the changes in the restriction pattern of MCMV/h
compared to the expression pattern of the parental virus (Fig.
1H) suggested that the phenotype could have been caused by
one or more mutations in the viral genome. As the sequencing
of a few candidate genes (such as genes with known antiapop-
totic functions) did not reveal any differences from wt MCMV,
we decided to have the whole MCMV/h genome sequenced.
The MCMV/h sequence was then compared to the previously
reported MCMV Smith sequence (42). The results of this
comparison are listed in Table 1. The most striking alteration
was a 10.3-kb deletion (nt 9561 to 19848, open reading frames
[ORFs] m10 to m18) combined with a 12.5-kb inverted dupli-
cation of ORFs m58 to m73 (nt 91689 to 104152). In addition,
several point mutations and small deletions were detected.
Some of these were also found in the parental MCMV BAC,
suggesting that they are probably not responsible for the phe-
notype of MCMV/h. As MCMV/h originated from a single
plaque of infected RPE-1 cells and was passaged repeatedly on
the same cell line, we assumed that MCMV/h represents a
single mutant clone. However, we did not subject this virus to
numerous rounds of plaque purification, which would be nec-
essary to formally exclude the possibility that MCMV/h con-
sists of more than one mutant clone.
To analyze which of the detected mutations contributed to
the extended-host-range phenotype of MCMV/h, we decided
7996 SCHUMACHER ET AL. J. VIROL.

to engineer selected mutations into the wt MCMV-GFP ge-
nome. The deletion of the region of m10 to m18 did not
facilitate MCMV replication in RPE-1 cells (data not shown),
indicating that this region does not contain a gene that actively
restricts MCMV’s host range. The reconstruction of the dupli-
cation spanning ORFs m58 to m73 is technically very difficult
and was therefore deferred.
Mutations in M112/M113 facilitate MCMV replication in
RPE-1 cells. Next, we focused on the mutations in M112/M113.
This region encodes the E1 proteins, which exist in at least four
differentially spliced isoforms (13, 16) (Fig. 4A). It is highly
homologous in its structure to the UL112-113 region of human
cytomegalovirus (16, 47, 60). Although the mechanisms of ac-
tion of the E1 proteins have not been fully elucidated, a num-
ber of studies have shown that they play an important role in
the formation of nuclear replication compartments (4, 38, 40),
in viral DNA replication (37, 61), and in the regulation of gene
expression (24, 28, 49). Hence, we thought that the mutations
in M112/M113 might play a role in the ability of MCMV/h to
replicate in human cells.
FIG. 1. Replication of MCMV/h in human and murine cells. (A to F) Human RPE-1 cells, HFFs, MRC-5 cells, and HUVECs were infected
at an MOI of 0.2 or 5 TCID50/cell with MCMV/h, MCMV-bclXL, or the parental wt virus. HCMV strains AD169 and TB40/E are shown for
comparison. (G) Murine 10.1 fibroblasts were infected with the same viruses described above (A) at an MOI of 0.1 TCID50/cell. Virus replication
was determined by titration. DL, detection limit. (H) Virion DNA was extracted from MCMV/h and the parental wt MCMV, digested with
restriction enzymes, and separated by gel electrophoresis. MCMV and HCMV (strain AD169) BAC DNAs are shown for comparison. Differences
in the restriction fragment patterns of BAC and virion DNAs are expected, as virion DNA is linear and BAC DNA is circular, and BAC DNA
contains the BAC replicon in addition to the viral genome. The terminal fragments of the virion DNA and the corresponding joined fragment of
the BAC DNA are indicated by arrowheads.
VOL. 84, 2010 E1 MUTATIONS MEDIATE MCMV REPLICATION IN HUMAN CELLS 7997

MCMV/h contains a 9-nucleotide deletion in M112/M113
that should result in a 3-amino-acid deletion affecting all four
E1 isoforms (Table 1 and Fig. 4A). In addition, two point
mutations were present, which were predicted to cause amino
acid exchanges only in the large 87-kDa isoform. A third point
mutation at position 163813 is likely silent. To test whether
these mutations influence MCMV replication in RPE-1 cells,
we first deleted the entire E1-coding region from the MCMV-
GFP BAC as shown in Fig. 4B. We then reinserted either the
wt or the mutant M112/M113 sequence. The resulting BACs
were named RevM112 and M112mut, respectively. The recom-
binant MCMV BACs were verified by restriction digestion
(Fig. 4C) and transfected into fibroblasts to obtain recombi-
nant viruses. The M112mut and RevM112 viruses were easily
reconstituted, but numerous attempts to reconstitute the
�M112/M113 deletion mutant were unsuccessful. This indi-
cated either that the E1 proteins are essential for MCMV
replication or that viral replication is so severely compromised
in their absence that a deletion mutant can be regenerated only
in complementing cells. RevM112 and M112mut replication in
murine fibroblasts was indistinguishable from that of the pa-
rental virus (data not shown). RevM112 also behaved like the
parental wt virus in that it did not replicate in RPE-1 cells. In
contrast, M112mut replicated to high titers in RPE-1 cells (Fig.
5A). However, it did not replicate as rapidly as MCMV/h,
suggesting that additional mutations present in MCMV/h must
contribute to its highly efficient replication in human cells. To
exclude the possibility that the FRT site, which remained be-
hind the M112/M113-coding region (Fig. 4B), had a negative
influence on the replication of M112mut, we constructed an
independent M112/M113 mutant MCMV using the galK sys-
tem (56). The resulting virus replicated with the same kinetics
in RPE-1 cells as M112mut (data not shown), indicating that
the FRT site did not have a detrimental effect. We also tested
to which extent the constructed viruses induced cell death in
RPE-1 cells. Again, RevM112 behaved like the parental wt
FIG. 2. MCMV/h induces less cell death than the parental wt MCMV. (A) RPE-1 cells were infected at an MOI of 5 TCID50/cell. Fluorescent
images were taken 24 and 48 hpi. Infected cells show green fluorescence, as both viruses express GFP. (B) RPE-1 cells were infected as described
above, and nuclear DNA fragmentation as a sign of apoptosis was determined at 48 hpi by using a TUNEL assay. Fragmented DNA is indicated
by red fluorescence. Nuclei were stained with DAPI. (C and D) Human RPE-1 cells (C) and murine 10.1 fibroblasts (D) were infected at an MOI
of 5 TCID50/cell with MCMV/h, the parental wt virus, or MCMV-bclXL. Cell viability was measured by using an MTS assay and is shown relative
to the viability at 0 hpi. The differences in viability between the wt and MCMV/h at 48 and 72 hpi were significant (P � 0.01 by a Student’s t test).
7998 SCHUMACHER ET AL. J. VIROL.

virus, and M112mut displayed an intermediate phenotype; it
induced apoptosis less efficiently than wt MCMV but more
efficiently than MCMV/h (Fig. 5B).
Next, we asked whether the mutations in M112/M113 had an
impact on the abundance of individual E1 protein isoforms in
infected RPE-1 cells. As shown in Fig. 6A, the abundances of
the 36- and 38-kDa E1 proteins were similar for all viruses up
to 48 hpi. The 33-kDa isoform was not detected by the
CROMA103 antibody, presumably because CROMA103 rec-
ognizes an epitope encoded by exon 2. The expression of the
87-kDa E1 protein was maintained up to 96 hpi in cells in-
fected with MCMV/h or the M112mut virus but was reduced at
48 hpi and almost absent at later times in cells infected with wt
MCMV or the revertant, RevM112. As the large E1 isoform
was previously reported to control the repressive effect of IE3
on the major immediate-early (MIE) promoter (49), we tested
whether the loss of the 87-kDa E1 protein correlated with a
reduced level of expression of the MIE proteins IE1 and IE3.
No obvious differences in the abundances of IE1 and IE3 were
detected up to 24 hpi (Fig. 6A), but at later times (particularly
at 72 and 96 hpi), the IE1 and IE3 levels were reduced. How-
ever, it is difficult to determine to which extent this effect was
caused by the 87-kDa E1 protein, because the onset of cell
death, which starts at around 48 hpi and is more pronounced in
wt- and RevM112-infected cells, impairs viral protein and
DNA levels in general.
The E1 proteins form intranuclear structures that also con-
tain IE3 and viral DNA polymerase and accumulate viral DNA
(4, 49). Therefore, they are considered to represent replication
compartments. They start out as small punctate foci that coa-
lesce to a large lobulated compartment at late times after
infection (13, 40). To test whether wt and mutant MCMVs
differ in their abilities to form replication compartments, we
analyzed the distributions of the E1, IE3, and M44 proteins in
infected RPE-1 cells by immunofluorescence. In all cases, IE3
colocalized in nuclear replication compartments with the DNA
polymerase processivity factor M44 and the E1 proteins, but
there were clear differences in the sizes of the replication
compartments (Fig. 6B and C). A statistical evaluation re-
vealed that the majority of cells infected with MCMV/h con-
tained large lobulated replication compartments, whereas only
small punctate structures were found in the vast majority
( 80%) of cells infected with wt MCMV or the revertant,
RevM112 (Fig. 6D). As in previous experiments, the con-
structed M112mut mutant showed an intermediate phenotype,
with large lobulated replication compartments in approxi-
mately 40% of infected RPE-1 cells and small punctate repli-
cation foci in approximately 60% of infected RPE-1 cells.
HCMV and MCMV DNA replication compartments initiate
from the periphery of promyelocytic leukemia protein-associ-
ated nuclear bodies (PML bodies, also known as ND10) (4,
49). The PML bodies contain proteins such as PML, Daxx, and
Sp100, which can function as transcriptional repressors (18, 34,
53). At early times after CMV infection, PML bodies are
dispersed, and the PML protein is found diffusely distributed
throughout the nucleus (3, 21, 31, 44, 58). ND10 disruption has
been correlated with efficient viral gene transcription and
DNA replication (2). Hence, we wondered whether MCMV
was capable of disrupting PML domains in human RPE-1 cells.
As shown in Fig. 7A and B, ND10 disruption was observed for
only a minority of wt MCMV-infected cells but for the majority
of MCMV/h-infected cells. The M112mut mutant also had an
increased ability to disrupt PML bodies but was less efficient
than MCMV/h (Fig. 7B). This raised the question of whether
FIG. 3. Viral gene expression and DNA replication. (A) Human RPE-1 and murine 10.1 cells were infected with wt MCMV or MCMV/h. The
levels of expression of the viral IE1, E1, M44, and gB proteins were determined by Western blotting. (B and C) Viral DNA replication was
determined by slot blot hybridization (B) or by real-time PCR (C). Viral genome amplification is shown relative to the value determined at 8 hpi
(representing input viral genomes before replication).
VOL. 84, 2010 E1 MUTATIONS MEDIATE MCMV REPLICATION IN HUMAN CELLS 7999

the E1 proteins encoded by the M112/M113 locus were respon-
sible for the ND10 disruption or whether these proteins pro-
mote ND10 disruption more indirectly. To resolve this ques-
tion, we tested if E1 proteins expressed from a plasmid vector
were able to disrupt PML bodies. Neither the E1 proteins
expressed from a wt M112/M113 gene sequence nor those
expressed from the MCMV/h-derived mutant M112/M113
gene were able to disperse PML nuclear domains. In contrast,
the transfection of an MCMV IE1 expression plasmid resulted
in the efficient disruption of PML domains, similar to the effect
of HCMV IE1 (Fig. 7C). These results suggested that the viral
E1 proteins did not disrupt ND10 by themselves but rather
promoted ND10 disruption by an indirect mechanism.
Recent studies have corroborated the repressive effect of
PML and hDaxx on HCMV replication by showing that the
elimination of these proteins by small interfering RNA
(siRNA)-mediated gene knockdown enhances viral replication
(51, 52, 59). Based on these observations we reasoned that a
knockdown of PML or hDaxx might also facilitate MCMV
replication in human RPE-1 cells. Therefore, we used previ-
ously described shRNA-expressing retroviruses (52) to gener-
ate stable PML and hDaxx knockdown cells (Fig. 7D) and
analyzed MCMV replication in these cells. As shown in Fig.
7E, PML knockdown did not facilitate the replication of wt
MCMV in RPE-1 cells but enhanced the replication of the
M112mut mutant. In contrast, the knockdown of hDaxx de-
creased the replication of M112mut in RPE-1 cells (Fig. 7E).
The reason for the decreased replication in Daxx knockdown
cells is unclear. However, it seems likely that a previously
reported increased sensitivity of Daxx knockdown cells to pro-
apoptotic stimuli is responsible (14, 35).
Taken together, this study shows that MCMV can sponta-
neously gain the ability to replicate in human cells by acquiring
a few small mutations in the M112/M113-coding region. These
mutations decrease the proapoptotic effect of MCMV infec-
tion on human cells and increase the formation of replication
TABLE 1. Sequence alterations detected in MCMV/h compared to the published sequence of the MCMV Smith strainf
ORF(s) Position(s) Sequence alteration Predicted consequence
Presence in
parental virus
m10–m18 9561–19848 �10288 nt �m10–m18 No
104152–91689 Inverse duplication Duplication of m58–M71 No
m20 20958 �G ORF extended Yesa
M26–M27 31973 �G Intergenic region Yes
m29 36197 �G Frameshift, ORF truncated Yesb
m29.1 36197 �G Frameshift, ORF extended Yesb
m30 37263 �C Frameshift, ORF extended Yesb
M31 38803 �G Frameshift, ORF extended Yesa
M32 40000 C3T Silent mutation Yes
M45 61917 �C Fusion of ORFs M45 and m45.1 Yesc
m58 92088 �A Frameshift Yes
92178 C3G ? Yes
92179 �CG Frameshift Yes
92206 CG3GC ? Yes
92226 G3A ? Yes
92230 G3A ? Yes
92358 �G Frameshift Yes
M71–M72 102942 �A ? (intergenic region) Yes
M78–M79 112624 �T ? (intergenic region) Yes
M86 123941 A3C N1104H Yes
123951 T3G L1107R Yes
M93 135803 A3C Silent mutation Yes
M102 146853 G3A D420N Yes
m108 162352–162389 �37 nt microRNAd deleted (positions 162364–162385) No
162374 A3G Silent mutation Yesd
162380 A3G Silent mutation Yesd
M112 163118–163126 �9 nt �GSP8-10 No
163813 C3T Silent mutation No
164243 A3G T352A No
164442 G3A G418E No
M116 168256 G3A G206V No
168370 G3A A256V No
168376 G3A S258F No
168526 C3A T296I No
m119.1 172403 C3A G197W No
M122 Ex5 179136 G3A S227F Yes
m129 187786 �T Frameshift, ORF truncated Yes
m143 201402 �G Frameshift, ORF extended Yese
a See reference 27.
b See reference 1.
c See reference 9.
d See references 12 and 17.
e See references 15 and 54.
f Reported under GenBank accession number NC_004065 (42).
8000 SCHUMACHER ET AL. J. VIROL.

compartments and the efficiency of viral DNA replication. The
mutations also enhance the ability of MCMV to disrupt PML
nuclear domains. However, PML-mediated repression of viral
transcription and/or replication does not seem to be the pri-
mary barrier to cross-species infection but rather limits its
efficiency. The introduction of the M112/M113 mutations into
wt MCMV did not fully reproduce the remarkably rapid rep-
lication of MCMV/h in human RPE-1 cells, suggesting that
additional mutations must contribute to its phenotype.
DISCUSSION
In this study we describe and characterize an MCMV mutant
that has spontaneously obtained the capacity to replicate in
human cells. At first, the identification of this mutant virus was
surprising to us. However, a literature search revealed that a
spontaneous adaptation of MCMV to human cells has been
seen before. In 1969, Raynaud and colleagues described the
adaptation of a field mouse (Apodemus sylvaticus) cytomega-
lovirus to monkey kidney cells and human diploid cells by first
passaging it through embryonic hamster cells and a baby ham-
ster kidney (BHK21) cell line (43). In monkey kidney cells,
their virus replicated to a titer of 107 TCID50/ml, a titer similar
to the titer that our MCMV/h isolate reached in human RPE-1
cells. A later study by Kim et al. compared the Raynaud strain
(prior to adaptation) to the MCMV Smith strain and found
that the two strains were indistinguishable in terms of replica-
tion kinetics, CPE morphology, virus particle density, and an-
tibody cross-reactivity, suggesting that the two viruses were
identical (30). Kim and colleagues concluded that the virus
must have undergone a remarkable change during tissue cul-
ture passage or that the virus that replicated in human cells was
a contaminant and not the original mouse virus. As the studies
by Raynaud et al. and Kim et al. were done before the advent
of molecular biology, it is impossible to tell whether the two
MCMV strains were genetically identical or only similar in
their phenotypes. Be that as it may, the present study clearly
shows that even the MCMV Smith strain can spontaneously
adapt to human cells and reveals a genetic basis for this ex-
tended-host-range phenotype.
In a previous study we showed that MCMV induces apop-
tosis in human cells and that an inhibition of apoptosis allows
FIG. 4. Construction of mutant MCMV genomes. (A) Structure of the M112/M113-coding region according to data described previously by
Ciocco-Schmitt et al. (16). Mutations detected in MCMV/h are indicated by arrows, and the predicted changes in the amino acid sequence are
shown below. (B) An MCMV M112/M113 deletion mutant was constructed by the insertion of a zeocin resistance gene (zeo) replacing the
M112/M113-coding region. The wt or mutant M112/M113 sequence was then reinserted by using a kan gene flanked by FRT sites as a selectable
marker. The kan gene was subsequently removed by Flp recombination. NheI restriction sites are indicated with N, and the expected NheI
fragment sizes are shown. (C) wt and mutant BACs were digested with NheI and separated by gel electrophoresis.
VOL. 84, 2010 E1 MUTATIONS MEDIATE MCMV REPLICATION IN HUMAN CELLS 8001

the virus to replicate in human cells (26). Surprisingly, the
human cell-adapted MCMV/h still induced apoptosis in a large
proportion of infected RPE-1 cells, although it was less than
that induced by the parental wt virus (Fig. 2C). A sequence
analysis of MCMV/h also did not reveal any mutations in
known antiapoptotic genes of MCMV or in MCMV homologs
of HCMV cell death suppressors. Instead, we demonstrated
that mutations in the M112/M113-coding sequence facilitate
MCMV replication in human cells and are at least in part
responsible for the remarkably efficient growth of MCMV/h in
RPE-1 cells. The mutations in MCMV/h enabled the virus to
express its proteins and replicate its DNA faster or more effi-
ciently than the parental wt virus (Fig. 3 and 6A). On the other
hand, the mutant virus also induced less apoptosis than wt
MCMV (Fig. 5B), and this most probably contributed to the
enhanced protein expression and DNA replication. However,
MCMV/h still has a more pronounced apoptosis-inducing ef-
fect on human cells than on murine cells. MCMV/h titers drop
quite rapidly after high-MOI infection of human cells, partic-
ularly if sensitive cells such as HUVECs are used (Fig. 1F).
Hence, MCMV/h is not yet perfectly adapted to human cells,
and apoptosis of infected cells must still be considered an
important limitation for MCMV cross-species infections.
The MCMV M112/M113 and the HCMV UL112-113 re-
gions are highly similar in their locations within the viral ge-
nome and share an almost identical splicing pattern (16). Dif-
ferentially spliced mRNAs encode at least four protein
products of 33, 36, 38, and 87 kDa in the case of MCMV (16)
and 34, 43, 50, and 84 kDa in the case of HCMV (60). Sec-
ondary modifications such as phosphorylation are responsible
for differences between the predicted and the apparent molec-
ular weights of some of the protein variants (13, 60). Although
the E1 proteins were identified many years ago, their mecha-
nisms of action remain poorly understood. E1 proteins become
detectable shortly after the major immediate-early proteins (4,
13), suggesting that very little IE protein is necessary to acti-
vate the 112/113 promoter. Together with the MIE proteins,
the E1 proteins can enhance the expression of other viral
genes, particularly those involved in DNA replication (24, 28).
In addition, the E1 proteins may also have a more direct role
in viral DNA replication, as they can bind single- and double-
stranded DNA and accumulate at nuclear replication sites
(25). Moreover, UL112-113 proteins are necessary for the
transient complementation of HCMV oriLyt-dependent DNA
replication (37), and antisense RNA to UL112-113 transcripts
inhibits viral DNA replication (61).
It seems likely that the different E1 protein isoforms have
nonidentical functions, but how the different isoforms function
and how they interact with each other and with other proteins
have only begun to be understood. Of the three relevant mu-
tations in MCMV/h, one mutation (the 3-amino-acid deletion)
affects all E1 protein isoforms, whereas the two point muta-
tions lead to amino acid exchanges only in the large 87-kDa
isoform (Fig. 4A). However, we cannot exclude that the point
mutations have an impact on splicing efficiency and shift the
balance between the various E1 isoforms. The higher abun-
dance of the 87-kDa E1 protein at late times postinfection
(Fig. 6A) might indeed reflect altered splicing but might also
be based on differences in protein stability between the wt and
the mutant 87-kDa proteins. Interestingly, it was recently
shown that the 87-kDa protein interacts with IE3 and modu-
lates the repressive effect of IE3 on the MIE promoter (49).
However, marked differences in IE1 or IE3 levels between wt
and mutant viruses were detected only from 48 hpi onward
(Fig. 6A), the same time when massive cell death starts, par-
ticularly in cells infected with wt MCMV. Hence, it is difficult
to determine the contributions of the 87-kDa E1 protein and
cell death to the reduced IE1 and IE3 levels, respectively. In
fact, previous work has shown that extended cell survival due
to the expression of an antiapoptotic protein correlated with
higher levels of viral protein expression at 72 and 96 hpi (26).
PML nuclear domains (ND10) consist predominantly of
proteins functioning as transcriptional repressors (18, 34, 53).
Viral genomes are frequently detected in close association with
PML domains, suggesting that the PML domains exert a cell-
intrinsic antiviral defense. This notion is supported by the fact
that some DNA viruses are capable of dispersing PML nuclear
bodies (18, 34, 53). Moreover, the inactivation of PML, Daxx,
or Sp100 by gene knockdown significantly enhances viral rep-
lication in permissive cells (52, 53, 59). However, can the dis-
ruption of PML structures also facilitate virus replication in
cells that are usually nonpermissive for a particular virus? We
previously constructed MCMV-hIE1 (an MCMV mutant ex-
pressing HCMV IE1), and this virus did not replicate to sig-
nificant titers in human RPE-1 cells or MRC-5 fibroblasts (26),
even though it disrupted ND10 in more than 60% of infected
FIG. 5. Viral replication and cell death induction in RPE-1 cells.
(A) RPE-1 cells were infected at an MOI of 0.2 TCID50/cell with wt
and mutant MCMVs as indicated. Virus replication was determined by
titration. DL, detection limit. (B) RPE-1 cells were infected at an MOI
of 5 TCID50/cell. Cell viability was measured by using an MTS assay
and is shown relative to the viability at 0 hpi. The differences in viability
between the wt and M112mut at 48 and 72 hpi were significant (P �
0.01 by a Student’s t test), but there was no significant difference
between the wt and RevM112.
8002 SCHUMACHER ET AL. J. VIROL.

RPE-1 cells (data not shown). Moreover, a knockdown of PML
or Daxx in the same cells did not facilitate the replication of wt
MCMV (Fig. 7E), indicating that ND10 disruption alone is not
sufficient for MCMV to replicate and spread efficiently. How-
ever, PML knockdown did enhance the replication of the
M112mut virus. Interestingly, using the same MCMV-hIE1
mutant, Tang and Maul previously showed a low level of rep-
lication in human fibroblasts (50). Those authors concluded
that HCMV IE1 can help MCMV to cross the species barrier.
The reason for this minor discrepancy between the two studies
FIG. 6. M112/M113 protein expression and formation of replication compartments. (A) Cells were infected with wt and mutant MCMVs. The
levels of expression of the IE1, IE3, E1, M44, and gB proteins were determined by Western blotting. The 36-, 38-, and 87-kDa E1 proteins were
detected on the same blot, but a longer exposure is shown for the large 87-kDa isoform, as its level of expression is much lower than the levels
of expression of the 36- and 38-kDa isoforms. Viral DNA (vDNA) levels were determined by Southern blotting. (B and C) Viral replication
compartments were stained at 24 hpi with antibodies against IE3, M44, and E1. (D) The relative abundances of the two different forms of
replication compartments at 24 hpi were determined by evaluating more than 150 nuclei per sample in four independent experiments. The
differences between wt and mutant MCMVs (MCMV/h and M112mut, respectively) were significant (P � 0.01 by a Student’s t test).
VOL. 84, 2010 E1 MUTATIONS MEDIATE MCMV REPLICATION IN HUMAN CELLS 8003

is unclear, and further investigation is needed to consolidate
these findings. However, we believe that it is fair to draw the
conclusion that PML disruption has little (or no) effect on
MCMV replication in human cells, unless the virus carries
additional mutations facilitating its replication in these cells.
Mutations in M112/M113 were sufficient to afford MCMV
replication in human RPE-1 cells. However, the M112mut
virus clearly did not replicate and spread as rapidly as
MCMV/h (Fig. 5A). Hence, additional mutations in MCMV/h
must contribute to its remarkably efficient replication. Many of
FIG. 7. ND10 disruption by MCMV in human cells. (A) RPE-1 cells were infected with wt MCMV, MCMV/h, or HCMV strain AD169.
Infected cell nuclei were stained at 24 hpi with antibodies against the viral E1 (for MCMV) or IE1 (for HCMV) protein. ND10 was stained with
a PML-specific antibody. (B) The relative abundance of punctate versus dispersed ND10 was determined by evaluating more than 150 nuclei per
sample in three independent experiments. Differences between wt and mutant MCMVs were significant (P� 0.01 by a Student’s t test). (C) RPE-1
cells were transfected with plasmids encoding enhanced GFP (EGFP), HCMV or MCMV IE1, and wt or mutant M112/M113 (E1) proteins.
Transfected cells were stained with antibodies specific for the transiently expressed proteins, and ND10 was stained with an anti-PML antibody.
(D) RPE-1 cells were transduced with retroviral vectors expressing PML- or Daxx-specific shRNAs, a control shRNA, or an empty vector as
described previously (52). Successful gene knockdown was verified by Western blotting. An unspecific background band recognized by the
anti-PML antibody is marked with an asterisk. (E) The shRNA-expressing cells were infected at an MOI of 0.2 TCID50/cell, and the replication
kinetics of wt MCMV (solid symbols) and the M112mut virus (open symbols) were determined by titration.
8004 SCHUMACHER ET AL. J. VIROL.

the mutations listed in Table 1 are in largely uncharacterized
ORFs or noncoding regions of the MCMV genome. There-
fore, it is difficult to predict which of them contributes to the
phenotype. The identification and characterization of addi-
tional contributing mutations should provide new insights into
the functions of as-yet-uncharacterized genes or regulatory
regions and into the mechanisms of the species specificity of
cytomegaloviruses in general. Once we have a better under-
standing of how MCMV can overcome the obstacles to its
replication in human cells, we might be able to construct
HCMV mutants that can replicate in murine cells or even in
the mouse, thereby providing the long-sought-after small-ani-
mal model for the study of HCMV infection in vivo.
ACKNOWLEDGMENTS
We thank E. Borst, R. Heller, S. Jonjic, L. Loh, M. Messerle, M.
Nevels, T. Shenk, T. Stamminger, and J. von Einem for providing cells,
plasmids, and antibodies; K. Berger for technical assistance; and M.
Budt, M. Nevels, and S. Voigt for critically reading the manuscript.
This work was supported by grant BR 1730/3-1 from the Deutsche
Forschungsgemeinschaft and an intramural research grant from the
Robert Koch Institute.
REFERENCES
1. Ahasan, M. M., and C. Sweet. 2007. Murine cytomegalovirus open reading
frame m29.1 augments virus replication both in vitro and in vivo. J. Gen.
Virol. 88:2941–2951.
2. Ahn, J. H., and G. S. Hayward. 2000. Disruption of PML-associated
nuclear bodies by IE1 correlates with efficient early stages of viral gene
expression and DNA replication in human cytomegalovirus infection.
Virology 274:39–55.
3. Ahn, J. H., and G. S. Hayward. 1997. The major immediate-early proteins
IE1 and IE2 of human cytomegalovirus colocalize with and disrupt PML-
associated nuclear bodies at very early times in infected permissive cells.
J. Virol. 71:4599–4613.
4. Ahn, J. H., W. J. Jang, and G. S. Hayward. 1999. The human cytomegalovirus
IE2 and UL112-113 proteins accumulate in viral DNA replication compart-
ments that initiate from the periphery of promyelocytic leukemia protein-
associated nuclear bodies (PODs or ND10). J. Virol. 73:10458–10471.
5. Atalay, R., A. Zimmermann, M. Wagner, E. Borst, C. Benz, M. Messerle, and
H. Hengel. 2002. Identification and expression of human cytomegalovirus
transcription units coding for two distinct Fcgamma receptor homologs.
J. Virol. 76:8596–8608.
6. Bodnar, A. G., M. Ouellette, M. Frolkis, S. E. Holt, C. P. Chiu, G. B. Morin,
C. B. Harley, J. W. Shay, S. Lichtsteiner, and W. E. Wright. 1998. Extension
of life-span by introduction of telomerase into normal human cells. Science
279:349–352.
7. Bruggeman, C. A., H. Meijer, P. H. Dormans, W. M. Debie, G. E. Grauls, and
C. P. van Boven. 1982. Isolation of a cytomegalovirus-like agent from wild
rats. Arch. Virol. 73:231–241.
8. Brune, W., M. Hasan, M. Krych, I. Bubic, S. Jonjic, and U. H. Koszinowski.
2001. Secreted virus-encoded proteins reflect murine cytomegalovirus pro-
ductivity in organs. J. Infect. Dis. 184:1320–1324.
9. Brune, W., C. Me´nard, J. Heesemann, and U. H. Koszinowski. 2001. A
ribonucleotide reductase homolog of cytomegalovirus and endothelial cell
tropism. Science 291:303–305.
10. Brune, W., M. Nevels, and T. Shenk. 2003. Murine cytomegalovirus m41
open reading frame encodes a Golgi-localized antiapoptotic protein. J. Virol.
77:11633–11643.
11. Brune, W., M. Wagner, and M. Messerle. 2006. Manipulating cytomegalo-
virus genomes by BAC mutagenesis: strategies and applications, p. 61–89. In
M. J. Reddehase (ed.), Cytomegaloviruses: molecular biology and immunol-
ogy. Caister Academic Press, Norfolk, United Kingdom.
12. Buck, A. H., J. Santoyo-Lopez, K. A. Robertson, D. S. Kumar, M. Reczko,
and P. Ghazal. 2007. Discrete clusters of virus-encoded microRNAs are
associated with complementary strands of the genome and the 7.2-kilobase
stable intron in murine cytomegalovirus. J. Virol. 81:13761–13770.
13. Bu¨hler, B., G. M. Keil, F. Weiland, and U. H. Koszinowski. 1990. Charac-
terization of the murine cytomegalovirus early transcription unit e1 that is
induced by immediate-early proteins. J. Virol. 64:1907–1919.
14. Chen, L. Y., and J. D. Chen. 2003. Daxx silencing sensitizes cells to multiple
apoptotic pathways. Mol. Cell. Biol. 23:7108–7121.
15. Child, S. J., L. K. Hanson, C. E. Brown, D. M. Janzen, and A. P. Geballe.
2006. Double-stranded RNA binding by a heterodimeric complex of murine
cytomegalovirus m142 and m143 proteins. J. Virol. 80:10173–10180.
16. Ciocco-Schmitt, G. M., Z. Karabekian, E. W. Godfrey, R. M. Stenberg, A. E.
Campbell, and J. A. Kerry. 2002. Identification and characterization of novel
murine cytomegalovirus M112-113 (e1) gene products. Virology 294:199–
208.
17. Do¨lken, L., J. Perot, V. Cognat, A. Alioua, M. John, J. Soutschek, Z.
Ruzsics, U. Koszinowski, O. Voinnet, and S. Pfeffer. 2007. Mouse cyto-
megalovirus microRNAs dominate the cellular small RNA profile during
lytic infection and show features of posttranscriptional regulation. J. Vi-
rol. 81:13771–13782.
18. Everett, R. D. 2006. Interactions between DNA viruses, ND10 and the DNA
damage response. Cell. Microbiol. 8:365–374.
19. Fioretti, A., T. Furukawa, D. Santoli, and S. A. Plotkin. 1973. Nonproductive
infection of guinea pig cells with human cytomegalovirus. J. Virol. 11:998–
1003.
20. Garcia-Ramirez, J. J., F. Ruchti, H. Huang, K. Simmen, A. Angulo, and P.
Ghazal. 2001. Dominance of virus over host factors in cross-species activa-
tion of human cytomegalovirus early gene expression. J. Virol. 75:26–35.
21. Ghazal, P., A. E. Visser, M. Gustems, R. Garcia, E. M. Borst, K. Sullivan, M.
Messerle, and A. Angulo. 2005. Elimination of ie1 significantly attenuates
murine cytomegalovirus virulence but does not alter replicative capacity in
cell culture. J. Virol. 79:7182–7194.
22. Harvey, D. M., and A. J. Levine. 1991. p53 alteration is a common event in
the spontaneous immortalization of primary BALB/c murine embryo fibro-
blasts. Genes Dev. 5:2375–2385.
23. Hobom, U., W. Brune, M. Messerle, G. Hahn, and U. H. Koszinowski. 2000.
Fast screening procedures for random transposon libraries of cloned her-
pesvirus genomes: mutational analysis of human cytomegalovirus envelope
glycoprotein genes. J. Virol. 74:7720–7729.
24. Iskenderian, A. C., L. Huang, A. Reilly, R. M. Stenberg, and D. G. Anders.
1996. Four of eleven loci required for transient complementation of human
cytomegalovirus DNA replication cooperate to activate expression of repli-
cation genes. J. Virol. 70:383–392.
25. Iwayama, S., T. Yamamoto, T. Furuya, R. Kobayashi, K. Ikuta, and K. Hirai.
1994. Intracellular localization and DNA-binding activity of a class of viral
early phosphoproteins in human fibroblasts infected with human cytomega-
lovirus (Towne strain). J. Gen. Virol. 75(Pt. 12):3309–3318.
26. Jurak, I., and W. Brune. 2006. Induction of apoptosis limits cytomegalovirus
cross-species infection. EMBO J. 25:2634–2642.
27. Kattenhorn, L. M., R. Mills, M. Wagner, A. Lomsadze, V. Makeev, M.
Borodovsky, H. L. Ploegh, and B. M. Kessler. 2004. Identification of proteins
associated with murine cytomegalovirus virions. J. Virol. 78:11187–11197.
28. Kerry, J. A., M. A. Priddy, T. Y. Jervey, C. P. Kohler, T. L. Staley, C. D.
Vanson, T. R. Jones, A. C. Iskenderian, D. G. Anders, and R. M. Stenberg.
1996. Multiple regulatory events influence human cytomegalovirus DNA
polymerase (UL54) expression during viral infection. J. Virol. 70:373–382.
29. Kim, K. S., and R. I. Carp. 1972. Abortive infection of human diploid cells
by murine cytomegalovirus. Infect. Immun. 6:793–797.
30. Kim, K. S., V. Sapienza, and R. I. Carp. 1974. Comparative studies of the
Smith and Raynaud strains of murine cytomegalovirus. Infect. Immun. 10:
672–674.
31. Korioth, F., G. G. Maul, B. Plachter, T. Stamminger, and J. Frey. 1996. The
nuclear domain 10 (ND10) is disrupted by the human cytomegalovirus gene
product IE1. Exp. Cell Res. 229:155–158.
32. Lafemina, R. L., and G. S. Hayward. 1988. Differences in cell-type-specific
blocks to immediate early gene expression and DNA replication of human,
simian and murine cytomegalovirus. J. Gen. Virol. 69:355–374.
33. Mahy, B. W. J., and H. O. Kangro. 1996. Virology methods manual. Aca-
demic Press, San Diego, CA.
34. Maul, G. G. 2008. Initiation of cytomegalovirus infection at ND10. Curr.
Top. Microbiol. Immunol. 325:117–132.
35. Michaelson, J. S., and P. Leder. 2003. RNAi reveals anti-apoptotic and
transcriptionally repressive activities of DAXX. J. Cell Sci. 116:345–352.
36. Mocarski, E. S., and C. T. Courcelle. 2001. Cytomegaloviruses and their
replication, p. 2629–2673. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A.
Lamb, M. A. Martin, B. Roizman, and S. E. Straus (ed.), Fields virology, 4th
ed. Lippincott Williams & Wilkins, Philadelphia, PA.
37. Pari, G. S., and D. G. Anders. 1993. Eleven loci encoding trans-acting factors
are required for transient complementation of human cytomegalovirus ori-
Lyt-dependent DNA replication. J. Virol. 67:6979–6988.
38. Park, M. Y., Y. E. Kim, M. R. Seo, J. R. Lee, C. H. Lee, and J. H. Ahn. 2006.
Interactions among four proteins encoded by the human cytomegalovirus
UL112-113 region regulate their intranuclear targeting and the recruitment
of UL44 to prereplication foci. J. Virol. 80:2718–2727.
39. Pass, R. F. 2001. Cytomegalovirus, p. 2675–2705. In D. M. Knipe, P. M.
Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E.
Straus (ed.), Fields virology, 4th ed. Lippincott Williams & Wilkins, Phila-
delphia, PA.
40. Penfold, M. E., and E. S. Mocarski. 1997. Formation of cytomegalovirus
DNA replication compartments defined by localization of viral proteins and
DNA synthesis. Virology 239:46–61.
41. Perot, K., C. M. Walker, and R. R. Spaete. 1992. Primary chimpanzee skin
VOL. 84, 2010 E1 MUTATIONS MEDIATE MCMV REPLICATION IN HUMAN CELLS 8005

fibroblast cells are fully permissive for human cytomegalovirus replication.
J. Gen. Virol. 73:3281–3284.
42. Rawlinson, W. D., H. E. Farrell, and B. G. Barrell. 1996. Analysis of the
complete DNA sequence of murine cytomegalovirus. J. Virol. 70:8833–8849.
43. Raynaud, J., P. Atanasiu, C. Barreau, and M. Jahkola. 1969. Adaptation
d’un virus cytome´galique provenant do Mulot (Apodemus sylvaticus) sur
diffe´rantes cellules he´te´rologiques, y compris les cellules humaines. C. R.
Hebd. Seances Acad. Sci. Hebd. Seances Acad. Sci. D 269:104–106.
44. Sandford, G. R., U. Schumacher, J. Ettinger, W. Brune, G. S. Hayward,
W. H. Burns, and S. Voigt. 2009. Deletion of the rat cytomegalovirus imme-
diate early 1 gene results in a virus capable of establishing latency but with
lower levels of acute virus replication and latency that compromise reacti-
vation efficiency. J. Gen. Virol. 91:616–621.
45. Sinzger, C., G. Hahn, M. Digel, R. Katona, K. L. Sampaio, M. Messerle, H.
Hengel, U. Koszinowski, W. Brune, and B. Adler. 2008. Cloning and se-
quencing of a highly productive, endotheliotropic virus strain derived from
human cytomegalovirus TB40/E. J. Gen. Virol. 89:359–368.
46. Smith, C. B., L. S. Wei, and M. Griffiths. 1986. Mouse cytomegalovirus is
infectious for rats and alters lymphocyte subsets and spleen cell proliferation.
Arch. Virol. 90:313–323.
47. Staprans, S. I., and D. H. Spector. 1986. 2.2-kilobase class of early transcripts
encoded by cell-related sequences in human cytomegalovirus strain AD169.
J. Virol. 57:591–602.
48. Swift, S., J. Lorens, P. Achacoso, and G. P. Nolan. 2001. Rapid production
of retroviruses for efficient gene delivery to mammalian cells using 293T
cell-based systems. Curr. Protoc. Immunol. 10:10.17C.
49. Tang, Q., L. Li, and G. G. Maul. 2005. Mouse cytomegalovirus early M112/
113 proteins control the repressive effect of IE3 on the major immediate-
early promoter. J. Virol. 79:257–263.
50. Tang, Q., and G. G. Maul. 2006. Mouse cytomegalovirus crosses the species
barrier with help from a few human cytomegalovirus proteins. J. Virol.
80:7510–7521.
51. Tavalai, N., P. Papior, S. Rechter, M. Leis, and T. Stamminger. 2006.
Evidence for a role of the cellular ND10 protein PML in mediating intrinsic
immunity against human cytomegalovirus infections. J. Virol. 80:8006–8018.
52. Tavalai, N., P. Papior, S. Rechter, and T. Stamminger. 2008. Nuclear domain
10 components promyelocytic leukemia protein and hDaxx independently
contribute to an intrinsic antiviral defense against human cytomegalovirus
infection. J. Virol. 82:126–137.
53. Tavalai, N., and T. Stamminger. 2008. New insights into the role of the
subnuclear structure ND10 for viral infection. Biochim. Biophys. Acta 1783:
2207–2221.
54. Valchanova, R. S., M. Picard-Maureau, M. Budt, and W. Brune. 2006.
Murine cytomegalovirus m142 and m143 are both required to block protein
kinase R-mediated shutdown of protein synthesis. J. Virol. 80:10181–10190.
55. Voigt, S., A. Mesci, J. Ettinger, J. H. Fine, P. Chen, W. Chou, and J. R.
Carlyle. 2007. Cytomegalovirus evasion of innate immunity by subversion of
the NKR-P1B:Clr-b missing-self axis. Immunity 26:617–627.
56. Warming, S., N. Costantino, D. L. Court, N. A. Jenkins, and N. G. Copeland.
2005. Simple and highly efficient BAC recombineering using galK selection.
Nucleic Acids Res. 33:e36.
57. Weller, T. H. 1970. Cytomegaloviruses: the difficult years. J. Infect. Dis.
122:532–539.
58. Wilkinson, G. W., C. Kelly, J. H. Sinclair, and C. Rickards. 1998. Disruption
of PML-associated nuclear bodies mediated by the human cytomegalovirus
major immediate early gene product. J. Gen. Virol. 79:1233–1245.
59. Woodhall, D. L., I. J. Groves, M. B. Reeves, G. Wilkinson, and J. H. Sinclair.
2006. Human Daxx-mediated repression of human cytomegalovirus gene
expression correlates with a repressive chromatin structure around the major
immediate early promoter. J. Biol. Chem. 281:37652–37660.
60. Wright, D. A., and D. H. Spector. 1989. Posttranscriptional regulation of a
class of human cytomegalovirus phosphoproteins encoded by an early tran-
scription unit. J. Virol. 63:3117–3127.
61. Yamamoto, T., S. Suzuki, K. Radsak, and K. Hirai. 1998. The UL112/113
gene products of human cytomegalovirus which colocalize with viral DNA in
infected cell nuclei are related to efficient viral DNA replication. Virus Res.
56:107–114.
8006 SCHUMACHER ET AL. J. VIROL.