Genome Biology 2007, 8:R48
2007Wang and Bradley Volume 8, Issue 4, Article R48
A recessive genetic screen for host factors required for retroviral
infection in a library of insertionally mutated Blm-deficient
embryonic stem cells
Wei Wang*† and Allan Bradley†
Addresses: *Department of Cell Biology and Genetics, College of Life Sciences, Peking University, Beijing 100871, PR China. †The Wellcome
Trust Sanger Institute, Hinxton, Cambridge CB10 1SA, UK.
Correspondence: Allan Bradley. Email: email@example.com
© 2007 Wang and Bradley; licensee BioMed Central Ltd.
This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Retroviral infection of embryonic stem cells<p>A recessive genetic screen of an insertionally mutated Blm-/- ES cell library identifies host factors required for retroviral infection, and confirms that mCat-1 is the ecotropic murine leukaemia virus receptor in ES cells.</p>
Background: Host factors required for retroviral infection are potential targets for the
modulation of diseases caused by retroviruses. During the retroviral life cycle, numerous cellular
factors interact with the virus and play an essential role in infection. Cultured embryonic stem (ES)
cells are susceptible to retroviral infection, therefore providing access to all of the genes required
for this process to take place. In order to identify the host factors involved in retroviral infection,
we designed and implemented a scheme for identifying ES cells that are resistant to retroviral
infection and subsequent cloning of the mutated gene.
Results: A library of mutant ES cells was established by genome-wide insertional mutagenesis in
Blm-deficient ES cells, and a screen was performed by superinfection of the library at high
multiplicity with a recombinant retrovirus carrying a positive and negative selection cassette.
Stringent negative selection was then used to exclude the infected ES cells. We successfully
recovered five independent clones of ES cells that are resistant to retroviral infection. Analysis of
the mutations in these clones revealed four different homozygous and one compound
heterozygous mutation in the mCat-1 locus, which confirms that mCat-1 is the ecotropic murine
leukemia virus receptor in ES cells.
Conclusion: We have demonstrated the feasibility and reliability of this recessive genetic
approach to identifying critical genes required for retroviral infection in ES cells; the approach
provides a unique opportunity to recover other cellular factors required for retroviral infection.
The resulting insertionally mutated Blm-deficient ES cell library might also provide access to
essential host cell components that are required for infection and replication for other types of
One characteristic of all viruses is dependency for replication
on components synthesized by host cells. All types of virus are
able to subvert the machinery in the host cell for replication
of the viral genome and expression of viral gene products .
Retroviral replication has a unique aspect, namely conversion
Published: 3 April 2007
Genome Biology 2007, 8:R48 (doi:10.1186/gb-2007-8-4-r48)
Received: 15 November 2006
Revised: 19 February 2007
Accepted: 3 April 2007
The electronic version of this article is the complete one and can be
found online at http://genomebiology.com/2007/8/4/R48
R48.2 Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley http://genomebiology.com/2007/8/4/R48
Genome Biology 2007, 8:R48
of genomic viral RNA into cellular DNA, which has been
exploited for the development of antiretroviral drugs. Inte-
gration of a retrovirus into the host genome concludes the
early stage of the life cycle, after which the virus can begin to
Study of the host factors that are involved in the retroviral life
cycle is important if we are to gain a detailed understanding
of the interaction between virus and host cell components.
Essential host cell components are potential targets for anti-
viral therapies that could be developed. Many viruses have
small genomes, and so the repertoire of components that can
be exploited as pharmaceutical targets is very limited. More-
over, because of their rapid replication, variants in the viral
genome that overcome the effect of inhibitors will be rapidly
selected, diminishing the effectiveness of antiviral agents.
The main drugs currently used to treat HIV infection are
inhibitors of two viral proteins, namely the reverse tran-
scriptase and the protease (encoded by the viral pol and gag
genes, respectively). Also, inhibitors of the HIV-1 entry and
fusion steps have been used as a third drug class in recent
years . Thus, therapeutic molecules targeting retroviral
host factors would be a potential new route to modulation of
diseases caused by retroviruses. Evidence of the importance
of host factors is provided by individuals who harbor
homozygous mutations in the gene encoding CC chemokine
receptor (CCR)5, who are extremely resistant to HIV infec-
tion. As a result of these observations, human antibodies to
CCR5 and small-molecule CCR5 antagonists are being inves-
tigated as potential HIV therapies [1,4].
Retroviral vectors are widely used as genetic vehicles or as
mutagens in embryonic stem (ES) cells. Comparatively few
studies have described the molecular components that are
essential for the interaction between retroviruses and ES
cells. In previous studies, several host genes required for viral
infection were identified by screening a gene trap library con-
structed in somatic cells . Here we describe a genetic
screen designed to identify host factors in ES cells that are
required for the early phase of the retroviral life cycle. This
recessive screen was conducted in a library of insertionally
mutated Blm-deficient ES cells. The random insertional
mutations in this library were generated using a recombinant
retroviral gene-trap vector, integration into genes of which
predominantly produces a loss of function mutation; the inte-
grated proviral DNA provides a sequence tag for identifying
the mutation . In principle the genome-wide gene-trap
mutations in this library should provide access to mutations
in the subset of genes expressed in ES cells . The Blm
(which encodes Bloom's syndrome protein)-deficient genetic
background of these ES cells is the second important feature
of this mutation library. Recessive genetic screens in a diploid
mammalian genome require an approach to generate cells
with homozygous mutations, which increases the complexity
of most genetic screens because of the low rate of loss of het-
erozygosity (LOH) of single allelic mutations in wild-type ES
cells. However, Blm-deficient ES cells have a 20-fold increase
in the rate of LOH , which offers a major advantage in
recessive screens. Indeed, two reports have described suc-
cessful use of Blm-deficient ES cells to identify recessive
mutations in genes required for DNA mismatch repair  and
the glycosylphosphatidylinasitol-anchor biosynthesis path-
For the screen described here, we have confirmed the utility
of this system in generating genome-wide homozygous muta-
tions to facilitate recessive genetic screens in vitro by identi-
fying mCat-1 as a critical gene in ES cells that is required for
retroviral infection. This screen was conducted by superinfec-
tion with a retroviral vector carrying the puro-Δtk (puromy-
cin-Δ-thymidine kinase) gene, a positive/negative selectable
marker . Clones surviving negative selection were shown
to be resistant to retroviral infection, and in every case the
mCat-1 gene was mutated. This success demonstrates the fea-
sibility of conducting genome-wide negative selection screens
for genes that confer resistance to infection.
Screening strategy for infection resistant mutants
The overall strategy for the screen is illustrated in Figure 1.
The principle behind the screen is selection against retrovi-
rally infected ES cells. The retrovirus used in the screen car-
ried the puro-Δtk positive/negative selection marker .
Infected cells expressing the puro-Δtk fusion gene are sensi-
tive to 1-(-2-deoxy-2-fluoro-1-β-D-arabino-furanosyl)-5-
iodouracil (FIAU) negative selection; thus, mutant ES cells
that cannot be infected by the virus will survive this negative
selection. This screen is therefore strongly dependent on the
ability to infect all cells in the culture with a retrovirus that
reliably expresses a negative selection cassette. A very high
infection efficiency must be achieved so that every single
infectable cell in the culture has at least one infection event.
In practice, the need to infect every cell in a culture of 109 cells
requires superinfection, in which every cell has between 10
and 20 independent viral insertions. Superinfection can be
achieved in a variety of ways, but in the screen described here
it was accomplished by co-cultivation of the gene-trap ES cell
library with the viral producer cells.
Superinfection with a puro-Δtk retroviral vector
To generate a viral producer cell line with a high titre, six dif-
ferent murine leukemia virus (MuLV) backbones were tested.
The puro-Δtk cassette was cloned into each backbone and the
vectors were tested for their efficiency in producing recom-
binant virus by transient transfection into phoenix packaging
cells . Viral titers were assessed using wild type ES cells,
and the WWF6 (Wang Wei female 6) vector had the highest
titer (Figure 2) because of several point mutations and a dele-
tion in its long terminal repeat that prevents transcriptional
suppression in ES cells . This recombinant vector was
used to generate the stable viral producer cell line B4-5 in
Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley R48.3
Genome Biology 2007, 8:R48
GPE-86 cells, another safe helper-free ecotropic packaging
cell line , which produced a titre of 2 × 104 colony-forming
Infection efficiency was investigated using both wild-type
(AB2.2) and Blm-deficient (NGG5.3) ES cells using irradiated
B4-5 producer cells. Following co-cultivation, 99.9% of the
both NGG5.3 and AB2.2 ES cells were resistant to puromycin
(Figure 3a). Furthermore, proviral copy numbers were inves-
tigated using clones isolated from these cultures without
selection. Southern blot analysis was performed using an
enzyme that cuts once in the provirus, and so each insertion
site will have a unique proviral-host genome junction frag-
ment (Figure 3b). This analysis revealed that, irrespective of
genotype, all clones had multiple insertion events, with an
average of five and a range from two to eight in this analysis.
These comparisons confirmed that co-cultivation resulted in
very efficient superinfection of Blm-deficient ES cells.
Screening for infection resistant mutants
The gene-trap library used in this study is sectored into eight
pools, each containing approximately 1,200 independent
gene-trap clones recovered by G418 selection (Figure 4a).
Each clone in the library has been expanded through a mini-
mum of 14 doublings . Given the total cell number, the
number of cell divisions and the rate of mitotic recombination
in Blm-deficient cells , each pool should contain a small
number of homozygous mutant cells derived from each inde-
pendent heterozygous insertion event.
The eight pools were separately co-cultivated with viral pro-
ducer cells and then selected in FIAU. After the first round of
FIAU selection, the surviving FIAU-resistant clones were
challenged a second time with the puro-Δtk retrovirus to
identify noninfectable mutant ES cells from those that were
not infected by chance. After three rounds of selection, 178
infection-resistant clones were recovered from five of the
eight pools. At the third round of selection, Blm-deficient ES
cells (NGG 5.3) were included as a positive control because
they are fully susceptible to viral infection. The phenotype of
the infection-resistant clones was clearly different from that
of the NGG5.3 cells (data not shown). Two representative
clones from each of the five pools were randomly selected to
test their degree of resistance to infection (Figure 4b). Among
the mutant clones from the same pool the resistance level was
quite similar, although it was variable between clones isolated
from different pools. Clones from pools 1 and 7 were almost
fully resistant to viral infection, whereas mutant clones from
pools 2, 3, and 4 were partially resistant, although still obvi-
ously different from wild-type cells.
To investigate the relationship between clones in same pools,
23 infection-resistant clones from pool 1 and 27 from pool 7
were analyzed by Southern blotting to detect the proviral-host
junction fragment of the gene-trap vector using a SAβgeo
probe. This analysis identified the same junction fragment in
clones from the same pool (data not shown), implying that
clones from the same pool were daughter clones. To deter-
mine whether this was the case for clones from all pools, a
representative sample of four clones from each pool was
tested (Figure 5a). In all cases, clones from the same pool
exhibited the same junction fragment, verifying the above
assumption. However, clones from the different pools had
different junction fragments, confirming that they had inde-
pendent mutations. A total of five independent clones were
recovered, corresponding to five pools of the library.
Cloning the viral insertion sites
The proviral-host junction from the five mutant clones was
recovered using a splinkerette polymerase chain reaction
(PCR)  and sequenced. The five sequences mapped to the
mCat-1 gene, a known receptor of MuLV in fibroblasts .
The proviral-host junction sequences mapped to five different
positions in introns 1 and 2 of the mCat-1 gene, demonstrat-
ing the independent origin of each of these five clones. The
Negative selection screen for infection-resistant cells
Negative selection screen for infection-resistant cells. A gene-trap library
in Blm-deficient embryonic stem (ES) cells was co-cultivated with the
irradiated viral producer cell line B4-5 and selected in 1-(-2-deoxy-2-
fluoro-1-β-D-arabino-furanosyl)-5-iodouracil (FIAU). Single cell clones
surviving the primary negative selection screen were tested twice by
exposure to viral supernatant and selection in puromycin to confirm
whether they were resistant to infection. The confirmed infection-
resistant clones were then molecularly characterized.
Negative selection (FIAU)
containing Blm-/-ES cells
(FIAUR & PuroS)
FIAU resistant colonies
Puro sensitive colonies
Puro resistant colonies
R48.4 Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley http://genomebiology.com/2007/8/4/R48
Genome Biology 2007, 8:R48
Puro-Δtk retroviral vector constructs and titres
Puro-Δtk retroviral vector constructs and titres. (a) All vectors were constructed from murine leukemia virus (MuLV) backbones and contained the Puro-
Δtk cassette. pWWF1, pWWF2, and pWWF3 are three revertible pBabe-based vectors containing a loxP site in the 3' long terminal repeat (LTR) .
pWWF4 and pWWF5 were constructed from vectors pQCXIX and pMSCV-Neo. pWWF6 was constructed in the pRetro-Super backbone. In some of
these vectors the 3' LTR has a self-inactivating mutation provided by an internal deletion to generate a self-inactivating provirus. (b) Titers of transiently
produced virus on ES cells after a single round of infection with 0.1 ml viral supernatant are illustrated by staining puromycin-resistant colonies in 24-well
Super-infection of ES cells by co-cultivation
Figure 3 (see following page)
Super-infection of ES cells by co-cultivation. (a) Efficiency of infection by co-cultivation with the viral producer cell line B4-5. AB2.2, wild-type (WT), and
NGG5.3 Blm-deficient embryonic stem cells cells were co-cultivated with irradiated B4-5 cells and plated in normal medium to determine plating efficiency
in the absence of selection and 1-(-2-deoxy-2-fluoro-1-β-D-arabino-furanosyl)-5-iodouracil (FIAU) to measure the frequency of noninfected cells. (b)
Assessment of proviral copy number in randomly picked nonselected wild-type and Blm-deficient embryonic stem (ES) cells after co-cultivation with the
viral producer cell line B4-5. Southern blot analysis was performed by using HindIII digested genomic DNA isolated from different single cell clones
cultured in nonselective medium. Each observed fragment represents a different proviral insertion. The probe is the PstI fragment generated from Puro-Δk
cassette. kb, kilobases; LTR, long terminal repeat.
pWWF1 from pCMV-Babe-oligo-revertible (pCBaOR)
SV40 OriSV40 Ori
pUC OripUC Ori
pUC Ori pUC Ori
pWWF2 from pBabe-EGFP-revertible (pBaER)
pWWF3 from pBabe-Oligo-revertible (pBaOR)
pWWF4 from pQCXIX
pWWF5 from pMSCV-Neo
pWWF6 from pRetro-super (pRS)
Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley R48.5
Genome Biology 2007, 8:R48
Figure 3 (see legend on previous page)
Number of colonies
H 5’ΔLTR H 3’ΔLTR
R48.6 Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley http://genomebiology.com/2007/8/4/R48
Genome Biology 2007, 8:R48
mutant clones from pools 1, 7, 2, 3 and 4 were termed V5, V4,
V3, V2 and V1, respectively (Figure 5b). In all cases the SAβ
geo cassette was in the appropriate transcriptional orienta-
tion. Because the ATG initiation codon of mCat-1 is in exon 3,
translation of the βgeo fusion gene uses its own start codon.
Furthermore, the integration sites of fully resistant clones V5
and V4 from pools 1 and 7 were downstream of exon 2 and
closer to the ATG start condon, whereas the partially resistant
clones were upstream of exon 2.
To determine whether the mutations were homozygous, these
clones were purified by single cell subcloning and Southern
analysis was performed using a probe from the mCat-1 gene.
This identified different proviral/mCat-1 junction fragments
in each clone, as expected (Figure 5c). Moreover, four of the
mutant clones (V1, V2, V4 and V5) lacked the wild-type allele
and were homozygous for the mutant allele, whereas the fifth
mutant clone (V3) appeared to be heterozygous because it
retained its wild-type mCat-1 allele.
To confirm the effect of the proviral insertions on expression
of the mCat-1 locus, reverse transcription (RT)-PCR was per-
formed using primers for exons 1 and 3, which spanned the
proviral insertion sites (Figure 5d). In all cases (including the
V3 clone), no product was detected. This confirmed that the
viral insertions affected the generation of a wild-type mCat-1
transcript upstream of the normal start codon of mCat-1. Fur-
thermore, the lack of a 'wild-type' exon 1 to 3 RT-PCR product
in the V3 clone suggested that this clone may carry a mutation
on the 'wild-type' allele that was not identified by Southern
analysis. RT-PCR was performed to detect possible tran-
scripts 3' of the proviral insertions. Using exon 4 to 7 and 8 to
12 primer pairs, an aberrant sized fragment was identified in
the exon 4 to 7 RT-PCR product from clone V3 (Figure 5d),
suggesting a potential splicing mutation on the 'wild-type'
allele from this clone. Sequence analysis of this product
revealed that the transcript of exon 4 to 7 from clone V3 was
shortened by skipping of exon 6 and part of exon 7 (Figure
5d), providing an explanation for the viral resistance of this
Confirmation of the causality of the mutations by Cre
Southern blotting analysis using a SAβgeo probe confirmed
that each of the five gene-trap clones had only a single gene-
trap viral insertion (Figure 5a), although in four out of five
cases this was bi-alleleic. To verify that the gene-trap inser-
tions in mCat-1 were directly responsible for the observed
phenotype, these were reverted with Cre, which deletes the
provirus, leaving a single long terminal repeat in the locus.
Reverted clones (1.1R, 7.1R, 2.1R, 3.1R, and 4.1R) were identi-
fied following transient expression of Cre by G418-sib selec-
tion and confirmed by genomic PCR, with primers for the
βgeo cassette (Figure 6a). Excision of the proviral insertion
restored the retroviral infection sensitivity of the revertants to
wild-type levels for each of the five clones tested (Figure 6b).
Successful infection by a retrovirus requires many host cell
factors . One of the most important of these is the receptor
present on the cell surface, which is necessary for retroviral
entry into the cell. Receptor-retroviral interactions can be
quite complex and may involve more than one molecule; for
instance, HIV requires a co-receptor in addition to CD4 .
mCat-1, which encodes a cationic amino acid transporter, was
identified as the receptor for murine ecotropic leukemia
viruses by an expression cloning strategy in fibroblasts .
In our study we used a loss-of-function assay, based on the
principle of negative selection of infected cells leading to the
Retroviral infection resistant mutants
Retroviral infection resistant mutants. (a) Retroviral gene-trap
mutagenesis. Illustration of the retroviral producer construct, RGTV-1,
with the SAβgeo gene trap cassette in the retroviral vector backbone, an
integrated provirus in the intron of a gene, and a Cre-reverted allele with a
single long terminal repeat (LTR) retained in the locus. βgeo, lacZ-neo
fusion gene; SA, splice acceptor. (b) Retroviral infection resistant
phenotype of five independent mutant clones. Daughter embryonic stem
(ES) cell clones from five different pools and controls were plated in
triplicate in 24-well plates. 'No drug' indicates that all clones grew without
selection; 'Puromycin selection' indicates that all clones were puromycin
sensitive; and 'Viral infection + Puromycin selection' indicates that all
clones were exposed to 1.2 × 105 colony-forming units of the B4-5 puro-
Δtk retroviral vector followed by selection. The NGG5.3 controls were
readily infected and clones from pools 1 and 7 were highly resistant to
infection; clones from pools 2, 3, and 4 were less resistant to infection.
NGG5.3 1.0 7.0 2.0 3.0 4.0
Full resistance Partial resistance
CMV- 5’TLR LoxP
Gene trap mutation
113 3 22
Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley R48.7
Genome Biology 2007, 8:R48
recovery of resistant cell clones. Because we are using a
retroviral vector to screen for mutants, the event that the
mCat-1 mutation blocks should occur at an early stage of the
retroviral life cycle, between receptor binding and integration
into the genome. Given previous data from other cell lines, we
believe that our identification of mCat-1 as the major MuLV
receptor in ES cells is reasonable.
In our screen, we recovered five independent mutations of
mCat-1 in a library of 10,000 independent gene-trap
Molecular analysis of mCat-1 mutations
Molecular analysis of mCat-1 mutations. (a) Junction fragment analysis of retroviral resistant clones. Southern blot of retroviral infection resistant clones
isolated from different subpools of the gene trap library. The clones isolated from different subpools have different HindIII proviral-host junction fragments
and are thus independent mutants, as expected. The clones isolated form the same pools share a common host-proviral junction fragment and thus appear
to be daughter clones in all cases. The probe is the ClaI fragment from the SA-βgeo cassette. LTR, long terminal repeat. (b) Insertion sites of gene-trap
virus in mCat-1. Part 1 shows the structure of the 5' end of mCat-1; the initiating ATG codon is in exon 3. The proviral-host junction at the end of 5' LTR
was sequenced by Splinkerette PCR. The location of the retroviral insertions is shown by arrows. Nucleotide positions are from National Center for
Biotechnology Information (NCBI) mouse build 35. The pools that correspond to V1 to V5 are shown in brackets. The structure of the fusion transcripts
for the insertions in introns 1 and 2 are also shown in parts 2 and 3, respectively. (c) Homozygosity analysis of mCat-1 in retroviral resistant clones.
Southern blot with an mCat-1 probe. Absence of the wild-type fragment in V1, V2, V4, and V5 indicates that the viral insertions are homozygous. The
presence of the wild-type fragment in the V3 clone reveals that in this clone the insertion is heterozygous. (d) Expression of wild-type and fusion
transcripts in gene-trap mutants. As shown in part 1, reverse transcription (RT)-polymerase chain reaction (PCR) with primers for mCat-1 exon 1 and lacZ
detected fusion transcripts of the expected size. Fusion transcripts of 238 base pairs (bp) were amplified from clones V1, V2 and V3, whereas transcripts
of 328 bp were detected from clones V4 and V5. The wild-type exon 1 to 3 transcript was only amplified from Blm-deficient ES cells. β-Actin (Actb) served
as a positive control for RT-PCR. As shown in part 2, RT-PCR with primers for mCat-1 exon 4 to 7 and exon 8 to 12 detected trace transcript levels of
mCat-1 from mutant clones except V1. The 403 bp product detected by RT-PCR for exon 4 to 7 primers in the V3 clone was sequenced and found to have
an aberrant splice, as illustrated in part 3.
Pool 1.0 2.0 3.0 4.0 7.0
1.1 1.2 1.3 1.4
2.1 2.2 2.3 2.4
3.1 3.2 3.3 3.4
4.1 4.2 4.3 4.4
7.1 7.2 7.3 7.4
V1, 2, 3
CG G GT
V1 V2 V3 V4 V5
Genomic location on Chr 5 (NCBI m36)
NGG5.3 V5(1.0) V4(7.0) V3(2.0) V2(3.0) V1(4.0)
3.1 3.24.1 4.2
Kpn I/Spe I
V1 V2 V3 V4 V5
(1.1) (7.1) (2.1) (3.1) (4.1)
4 5 6 7
4 5 7
mCat-1 Exon 4-7
V3 mutated transcript
NGG5.3 V5 V4 V3 V2 V1
R48.8 Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley http://genomebiology.com/2007/8/4/R48
Genome Biology 2007, 8:R48
mutations. The fact that we did not identify any other critical
gene in this screen illustrates that receptor-mediated entry is
perhaps the most nonredundant, essential step in the retrovi-
ral replication. Previously, Murray  and Sheng  and
their coworkers exploited gene-trap mutagenesis to generate
mutation libraries in several mammalian cell lines, and using
these libraries they identified genes required for viral replica-
tion of HIV-1, filoviruses, and reovirus These studies confirm
that insertional mutagenesis provides a rapid, genome-wide
approach to identifying host cellular factors that are required
for virus infection in different cell types.
The recovery of multiple independent mutations in the mCat-
1 gene in this study confirms its importance in MuLV replica-
tion. This library was used previously to screen for genes in
the DNA mismatch repair pathway . In the previous screen
seven independent homozygous mutations in Msh6 and two
independent mutations in Dnmt1 were recovered. The
number of independent mutations recovered in the Msh6 and
mCat-1 genes is greater than expected based on the complex-
ity of the library, which suggests that these genes might be
integration 'hot spots' for the gene-trap vector used as the
insertional mutagen for this library. A systematic analysis of
gene-trap 'hot spots' has been described by the German Gene
Trap Consortium (GGTC), which reported that 75% of gene-
trap mutations appeared only once in the gene-trap database
but 25% were represented by multiple clones and half of these
'hot spots' were vector specific . Thus, our previous fail-
ure to identify mutations in Msh2 and Mlh1 mismatch repair
MMR genes in our previous study, as well as other retroviral-
related host factor genes in this study, suggests that the use of
a single retroviral vector limited the coverage of the genome.
More than 10,000 genes are known to be expressed and
mutable by gene trapping in ES cells . Promoter gene-trap
mutagenesis is dependent on correct splicing between the
endogenous gene and the βgeo splice acceptor, and either the
generation of a fully functional fusion protein with the
trapped gene or insertion upstream of the normal initiation
codon. In this study only one potential reading frame of the
gene-trap virus was used, which limits the selectable inser-
tion events to those that occur in the appropriate reading
frame and produce a functional fusion protein. The gene-trap
fusion transcripts with mCat-1 and those isolated previously
(Msh6 and Dnmt1) were in the same reading frame. In prin-
ciple, we should be able to recover additional genes required
Reversibility of mCat-1 mutations with Cre
Reversibility of mCat-1 mutations with Cre. (a) Genomic polymerase chain reaction (PCR) with primers for LacZ and Neo detected the βgeo cassette in
the five mCat-1 mutations but was absent in five revertants. β-Actin (Actb) served as a positive control for genomic PCR. (b) Five revertant clones were
sensitive to G418, confirming excision of the virus and susceptibility to retroviral infection.
NGG5.3 1.1-R 7.1-R 2.1-R 3.1-R 4.1-R
Viral infection +
- + - + - + - + - +
Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley R48.9
Genome Biology 2007, 8:R48
for retroviral infection by constructing and screening libraries
with gene-trap vectors with alternative reading frames.
Another limitation of retroviral mutagenesis is the nonran-
dom nature of retroviral integration, which is known to favor
the 5' ends of genes. Vectors such as Piggybac  provide
alternative gene-trap insertional mutagenesis methods that
potentially are without such a bias. Finally, the chance of gen-
erating a homozygous mutation in Blm-deficient ES cells will
depend on its chromosomal position. Homozygous mutations
are more likely to be generated in genes located closer to the
telomere than those close to the centromere. Indeed, mCat-1
is located at 148.7 megabases (Mb), just 4 Mb from the tel-
omere of chromosome 5, whereas the gene recovered most
frequently in our previous screen , namely Msh6, is located
at 88.7 Mb, just 7 Mb from the telomere of chromosome 17. In
contrast, Dnmt1 is located at 20.7 Mb, suggesting that the
reduced frequency of recovering mutations in Dnmt-1 in pre-
vious studies may be related in part to its chromosomal
In order to avoid some of the biases associated with gene-trap
mutagenesis, highly efficient mutagenesis agents such as γ-
irradiation or chemical mutagens can be used. Indeed, ENU
(N-ethyl-N-nitrosourea) was successfully used to conduct a
screen with Blm-deficient ES cells previously . However,
with such approaches identification of the molecular change
can be extremely difficult, and in this respect retroviral vector
or transposon based gene-trap approaches offer a major
advantage. cDNA expression libraries provide an alternative
for functional screening for host genes that confer suscepti-
bility to viral infection. Such screens are generally configured
to rescue viral resistance of a cell line, which might operate at
any stage of the replicative cycle of the virus [23,24].
In a summary, in this screen we exploited an approach that
combined gene-trap insertional mutagenesis in Blm-deficient
ES cells with superinfection and negative selection, and
proved that mCat-1 is an essential host factor for retroviral
infection in ES cells. In principle, application of this screening
methodology with more complete libraries should identify
other cellular factors that are required in the early stages of
retroviral infection. Moreover, although the coverage of the
existing library is not complete, it should prove valuable for
recovery of essential host factors for other pathogenic agents.
Materials and methods
Construction of retroviral vectors
To generate a retroviral vector with the highest possible titre,
the puro-Δtk positive/negative selection cassette from
pYTC37  was cloned into several different retroviral vector
backbones. The backbones used were pCMV-Babe-Oligo-
Revertible (pCBaOR), pBabe-EGFP-Revertible (pBaER), and
pBabe-Oligo-Revertible (pBaOR); these three vectors were
modified on pBabe retroviral vectors [25,26], pQCXIX and
pMSCV-Neo (retroviral expression vectors; Clontech, Moun-
tain View, CA, USA) [27,28], and pRetro-Super (pRS; a gift
from Roderick Beijersbergen, The Netherlands Cancer Insti-
tute, Amsterdam, The Netherlands.)
Embryonic stem cell culture
ES cell culture was described in detail previously . Briefly,
ES cells were maintained on γ-irradiated feeder cell layers in
Dulbecco's modified Eagle's medium supplemented with 15%
fetal bovine serum, 2 mmol/l L-glutamine, 50 units/ml peni-
cillin, 40 μg/ml streptomycin, and 100 μmol/l β-mercap-
toethanol. Cells were cultured at 37°C with 5% carbon
Transient titer test
The titer of each retroviral vector was assessed by transient
transfection of 25 μg of each vector into phoenix helper-free
packaging cells  using calcium phosphate transfection
. Packaged virus was harvested 48 hours after
The viral titer was assessed using ES cells. Twenty four hours
before infection, ES cells were plated in 24-well plates at a
density of 3 × 105 cells per well. Viral supernatant from each
vector was filtered through a 0.45 μm filter, and polybrene
(hexadimethrine bromide) was added to a final concentration
of 10 μg/ml. One milliliter of each filtered supernatant was
applied to ES cells, and puromycin selection (3 μg/ml) was
initiated 24 hours after infection and continued for 8 days.
Drug-resistant ES colonies were fixed and stained with 2%
methylene blue in 70% ethanol and counted.
Construction of a stable viral producer cell line
pWWF6 (25 μg) was transfected into 1 × 107 GPE-86 cells 
by electroporation (290 V/cm and 960 μF). The cells were
plated and puromycin was added 48 hours after electropora-
tion. Puromycin-resistant colonies were picked into 24-well
plates and the titres of 72 independent colonies were assessed
as described previously. The 10 clones with the highest titers
were reassessed to identify one for use in future experiments.
The clone with highest titer was B4-5.
Superinfection of gene-trap library
The screen relies on superinfection of the subpools of the
gene-trap library. To maximize exposure of the ES cells to
viral particles, a co-cultivation strategy was used. B4-5 cells
were collected, suspended in media at a density of 1 × 107
cells/ml, and γ-irradiated with the dose of 6000 cGray. About
6 × 107 irradiated B4-5 cells were plated on to a 150 mm tissue
culture dish and 24 hours later 3.5 × 106 ES cells from each
pool of the gene-trap library were plated onto the irradiated
B4-5 cells. After six days of co-culture, the ES cells were con-
fluent. These cells were expanded and 1.8 × 107 cells were
selected in FIAU (0.2 μmol/l). Selection was maintained for
R48.10 Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley http://genomebiology.com/2007/8/4/R48
Genome Biology 2007, 8:R48
10 days. FIAU-resistant ES cell colonies were picked into 96-
well feeder plates for further assessment.
Second and third round infection assay
Second and third round infection was performed respectively
in 96-well and 24-well feeder plates using a sib-selection
strategy. For the second round of infection, each clone was
plated in duplicate 24 hours before infection and one dupli-
cate was exposed to 200 μl (approximately 4 × 103 colony-
forming units) of viral supernatant from B4-5 cells. This was
repeated 12 hours later. The following day one plate was
selected with puromycin and a duplicate copy was maintained
without selection. Clones that were infected were resistant to
puromycin and excluded. Puromycin-sensitive clones were
then tested a third time in 24-well plates using the same pro-
tocol to confirm their resistance to infection.
Isolation of the proviral junction
Proviral junction fragments were isolated using Splinkerette-
PCR, as described elsewhere . Briefly, genomic DNA was
digested with Sau3AI or FatI and ligated with the corre-
sponding Splikerette adaptors HMSp-Sau3AI or HMSp-FatI.
The Splikerette adaptors were generated by annealing Splik-
erette oligoes HMSpBb-Sau3AI or HMSpBb-FatI with
HMSpAa. The first found of PCR was carried out with viral
primer AB949new and Splinkerette primer HMSp1. The
nested PCR was carried out with the viral primer HM001 and
Splinkerette primer HMSp2. The specific PCR products were
gel purified and TA cloned  for sequencing.
Southern blotting and hybridization
Total genomic DNA was restricted, size fractionated on agar-
ose gels, blotted, and hybridized using standard procedures.
The following probes were used: LacZ, a 800 base pair ClaI-
digested fragment from pSAgeo , which is a plasmid con-
taining the SAβgeo cassette in pBS; and Puro-Δtk, a 1.2 kilo-
Sequence of nucleotides and primers.
HMSpAa5'-CGA AGA GTA ACC GTT GCT AGG AGA GAC CGT GGC TGA
ATG AGA CTG GTG TCG ACA CTA GTG G-3'
5'-gat cCC ACT AGT GTC GAC ACC AGT CTC TAA (T)10C(A)7-3'
5'-cat gCC ACT AGT GTC GAC ACC AGT CTC TAA (T)10C(A)7-3'
Splinkerette PCR primers
5'-GCT AGC TTG CCA AAC CTA CAG GTG G-3'
5'-GCC AAA CCT ACA GGT GGG GTC TTT-3'
5'-CGA AGA GTA ACC GTT GCT AGG AGA GAC C-3'
5'-GTG GCT GAA TGA GAC TGG TGT CGA C-3'
Primers for sequencing
M13 forward primer
M13 reverse primer
5'-CTG GCC GTC GTT TTA C-3'
5'-CAG GAA ACA GCT ATG AC-3'
5'-GGC CAC GCG TCG ACT AGT AC (T)17-3'
5'-GCG TGC GCC ATC CCC TCA GCT AGC A-3'
5'-CGA TGA TGT AGG AGA GAA TCA GGT-3'
5'-GTA CTT CAA GCG TGG CAA GAG-3'
5'-CTG GTG GAA AGT GCG CAG AGA-3'
5'-TCT CCT AGG CTC CAT GTT TCC C-3'
5'-CTA TCA GCA TCC ACA CTG CAA A-3'
5'-ATG TGC TGC AAG GCG ATT AAG-3'
Genomic PCR primers
5'-GTT TGA GAC CTT CAA CAC CCC-3'
5'-GTG GCC ATC TCC TGC TCG AAG TC-3'
PCR, polymerase chain reaction; RT, reverse transcription.
Genome Biology 2007, Volume 8, Issue 4, Article R48 Wang and Bradley R48.11
Genome Biology 2007, 8:R48
base PstI-digested fragment from pYTC37 , which is the
plasmid that contains the puro-Δtk cassette in pPGKbpA.
Reverse transcription polymerase chain reaction
Total RNA was extracted from the five mutant clones and
NGG5.3 cells as a negative control, and 5 μg total RNA was
used to generate the first strand cDNA. β-Actin (Actb) was
used as a positive control for RT-PCR. The PCR product of the
mutant mCat-1 transcript from exon 4 to 7 from clone V3 was
TA cloned for sequencing.
Five million ES cells were electroporated with 10 μg of the Cre
expression plasmid pCAG-Cre , plated at low density
(2,000 cells/90 mm plate), and grown without selection for 9
days. ES cell colonies were picked into 96-well plates. Clones
that had excised both copies of the gene-trap cassette were
identified as G418-sensitive clones by sib-selection and con-
firmed by genomic PCR, using primers that amplify a 578
base pair fragment in the βgeo cassette. Reinfection was per-
formed as described above to assess phenotypic reversal.
We would like to thank Ge Guo for discussion at an early phase in the
design of this screen; Zikai Xiong for great help with experimental instruc-
tion; Frances Law and Alistair Beasley for help with tissue culture; and
Antony Rodriguez, Shaun Cowley, and Haydn Prosser for their comments
on this manuscript. This work was supported by the Wellcome Trust &
Sanger Institute grant number 79643.
1. Wagner EK, Hewlett MJ: Basic Virology Malden, MA: Blackwell Science
2. Coffin MJ, Hughes SH, Varmus HE: Retrovirus New York: Cold Spring
Harbor Laboratory Press; 1997.
3.Rathbun RC, Lockhart SM, Stephens JR: Current HIV treatment
guidelines: an overview. Curr Pharm Des 2006, 12:1045-1063.
4.Deng H, Liu R, Ellmeier W, Choe S, Unutmaz D, Burkhart M, Di Mar-
zio P, Marmon S, Sutton RE, Hill CM, et al.: Identification of a
major co-receptor for primary isolates of HIV-1. Nature 1996,
5.Rubin DH, Ruley HE: Cellular genetics of host susceptibility and
resistance to virus infection. Crit Rev Eukaryot Gene Expr 2006,
6.Stanford WL, Cohn JB, Cordes SP: Gene-trap mutagenesis: past,
present and beyond. Nat Rev Genet 2001, 2:756-768.
7. Stryke D, Kawamoto M, Huang CC, Johns SJ, King LA, Harper CA,
Meng EC, Lee RE, Yee A, L'Italien L, et al.: BayGenomics: a
resource of insertional mutations in mouse embryonic stem
cells. Nucleic Acids Res 2003, 31:278-281.
8.Luo G, Santoro IM, McDaniel LD, Nishijima I, Mills M, Youssoufian H,
Vogel H, Schultz RA, Bradley A: Cancer predisposition caused by
elevated mitotic recombination in Bloom mice. Nat Genet
9. Guo G, Wang W, Bradley A: Mismatch repair genes identified
using genetic screens in Blm-deficient embryonic stem cells.
Nature 2004, 429:891-895.
10. Yusa K, Horie K, Kondoh G, Kouno M, Maeda Y, Kinoshita T, Takeda
J: Genome-wide phenotype analysis in ES cells by regulated
disruption of Bloom's syndrome gene. Nature 2004,
11. Chen YT, Bradley A: A new positive/negative selectable
marker, puDeltatk, for use in embryonic stem cells. Genesis
12.Kinsella TM, Nolan GP: Episomal vectors rapidly and stably pro-
duce high-titer recombinant retrovirus. Hum Gene Ther 1996,
Hawley RG, Lieu FH, Fong AZ, Hawley TS: Versatile retroviral
vectors for potential use in gene therapy. Gene Ther 1994,
Markowitz D, Goff S, Bank A: A safe packaging line for gene
transfer: separating viral genes on two different plasmids. J
Virol 1988, 62:1120-1124.
Mikkers H, Allen J, Knipscheer P, Romeijn L, Hart A, Vink E, Berns A:
High-throughput retroviral tagging to identify components
of specific signaling pathways in cancer. Nat Genet 2002,
Kim JW, Closs EI, Albritton LM, Cunningham JM: Transport of cat-
ionic amino acids by the mouse ecotropic retrovirus
receptor. Nature 1991, 352:725-728.
Albritton LM, Tseng L, Scadden D, Cunningham JM: A putative
murine ecotropic retrovirus receptor gene encodes a multi-
ple membrane-spanning protein and confers susceptibility to
virus infection. Cell 1989, 57:659-666.
Murray JL, Mavrakis M, McDonald NJ, Yilla M, Sheng J, Bellini WJ,
Zhao L, Le Doux JM, Shaw MW, Luo CC, et al.: Rab9 GTPase is
required for replication of human immunodeficiency virus
type 1, filoviruses, and measles virus. J Virol 2005,
Sheng J, Organ EL, Hao C, Wells KS, Ruley HE, Rubin DH: Mutations
in the IGF-II pathway that confer resistance to lytic reovirus
infection. BMC Cell Biol 2004, 5:32.
Hansen J, Floss T, Van Sloun P, Fuchtbauer EM, Vauti F, Arnold HH,
Schnutgen F, Wurst W, von Melchner H, Ruiz P: A large-scale,
gene-driven mutagenesis approach for the functional analy-
sis of the mouse genome. Proc Natl Acad Sci USA 2003,
Bradley A: Mining the mouse genome. Nature 2002,
Ding S, Wu X, Li G, Han M, Zhuang Y, Xu T: Efficient transposition
of the piggyBac (PB) transposon in mammalian cells and
mice. Cell 2005, 122:473-483.
Gao G, Goff SP: Isolation of suppressor genes that restore ret-
rovirus susceptibility to a virus-resistant cell line. Retrovirology
Naghavi MH, Valente S, Hatziioannou T, de Los Santos K, Wen Y,
Mott C, Gundersen GG, Goff SP: Moesin regulates stable micro-
tubule formation and limits retroviral infection in cultured
cells. EMBO J 2007, 26:41-52.
Morgenstern JP, Land H: Advanced mammalian gene transfer:
high titre retroviral vectors with multiple drug selection
markers and a complementary helper-free packaging cell
line. Nucleic Acids Res 1990, 18:3587-3596.
Guo G: Recessive genetic screen for mismatch repair compo-
nents in Blm deficient ES cells. PhD thesis [http://
www.sanger.ac.uk/Info/theses/]. University of Cambridge
Julius MA, Yan Q, Zheng Z, Kitajewski J: Q vectors, bicistronic ret-
roviral vectors for gene transfer. Biotechniques 2000,
Grez M, Akgun E, Hilberg F, Ostertag W: Embryonic stem cell
virus, a recombinant murine retrovirus with expression in
embryonic stem cells. Proc Natl Acad Sci USA 1990, 87:9202-9206.
Brummelkamp TR, Bernards R, Agami R: Stable suppression of
tumorigenicity by virus-mediated RNA interference. Cancer
Cell 2002, 2:243-247.
Ramirez-Solis R, Davis AC, Bradley A: Gene targeting in embry-
onic stem cells. Methods Enzymol 1993, 225:855-878.
Hillova J, Hill M, Goubin G, Dantchev D: Infectivity of Rous sar-
coma cell DNA: comparison of two techniques of transfec-
tion assay. Intervirology 1975, 5:367-374.
Jayaraman K, Puccini CJ: A PCR-mediated gene synthesis strat-
egy involving the assembly of oligonucleotides representing
only one of the strands. Biotechniques 1992, 12:392-398.
Soriano P, Friedrich G, Lawinger P: Promoter interactions in ret-
rovirus vectors introduced into fibroblasts and embryonic
stem cells. J Virol 1991, 65:2314-2319.
Araki K, Araki M, Miyazaki J, Vassalli P: Site-specific recombina-
tion of a transgene in fertilized eggs by transient expression
of Cre recombinase. Proc Natl Acad Sci USA 1995, 92:160-164.