Current Gene Therapy, 2011, 11, 000-000 1
1566-5232/11 $58.00+.00 © 2011 Bentham Science Publishers
Therapeutic Applications of the PhiC31 Integrase System
Christopher L. Chavez and Michele P. Calos*
Department of Genetics, Stanford University School of Medicine, Stanford, CA 94305-5120, USA
Abstract: The potential use of the ?C31 integrase system in gene therapy opens up the possibilities of new treatments for
old diseases. ?C31 integrase mediates the integration of plasmid DNA into the chromsomes of mammalian cells in a se-
quence-specific manner, resulting in robust, long-term transgene expression. In this article, we review how ?C31 integrase
mediates transgene integration into the genomes of target cells and summarize the recent preclinical applications of the
system to gene therapy. These applications encompass in vivo studies in liver and lung, as well as increasing ex vivo uses
of the system, including in neural and muscle stem cells, in cord-lining epithelial cells, and for the production of induced
pluripotent stem cells. The safety of the ?C31 integrase system for gene therapy is evaluated, and its ability to provide
treatments for hemophilia is discussed. We conclude that gene therapy strategies utilizing ?C31 integrase offer great
promise for the development of treatments in the future.
Keywords: att site, genetic disease, genomic integration, hemophilia, induced pluripotent stem cells, liver, non-viral, phage,
serine recombinase, site-specific, stem cells.
To ensure that a given genetic modification has the de-
sired therapeutic effect, chromosomal integration of the
transgene is often required. This is because correction of
most genetic disorders requires the long-term expression of
the missing gene product. Therefore, the stability conferred
by integration of a transgene into the chromosomes of self-
renewing or long-lived cells could potentiate a life-long cor-
rection of the genetic disorder. Transgene integration by
some commonly used vectors, such as retroviruses and
transposons, is nearly random. For these types of vectors,
there are millions of potential integration sites present in a
given genome. Therefore, when used in gene therapy, each
transgene integration is unique and unpredictable. Since
chromosomal context has a strong influence on gene expres-
sion, a random integration pattern will result in a variable
transgene expression profile. Indeed, some of the integrated
transgenes will fail to express at all. Integration of transgenes
into open chromatin regions leads to a more uniform and
higher level of transgene expression, which is a goal in gene
In addition to non-uniform transgene expression, another
challenge with the use of randomly integrating vectors is the
very real prospect of insertional mutagenesis. The greatest
concern is that the vector will integrate near an oncogene and
that the regulatory elements associated with the vector could
promote oncogene expression, leading to the induction of
cancer. This was indeed the case in both a gene marking
study in mice and a clinical trial for X-linked severe combined
immune deficiency [1,2]. Another concern with the use of
these types of vectors is possible disruption of a tumor sup-
pressor gene, which could also lead to the formation of cancer.
*Address correspondence to this author at the Department of Genetics,
Stanford University School of Medicine, Stanford, CA 94305-5120, USA;
Tel: 650-723-5558; Fax: 650-725-1534; E-mail: firstname.lastname@example.org
gration mechanism that is sequence-specific and therefore
has fewer and more predictable integration sites. Retroviral
integrases and transposases lack these features, being derived
from multicellular organisms, where random integration is
more easily tolerated due to large regions of intergenic and
intronic DNA. Zinc finger nucleases, while designed to be
site-specific, sometimes cleave off-target sites, causing tox-
icity. Zinc finger nucleases are also complex to engineer and
can be quite costly . By contrast, integrases isolated from
bacteriophages offer a simpler solution. These enzymes are
often highly sequence-specific and require the pairing of
attachment (att) sites for recombination to occur. Bacterial
genomes are small and compact, requiring integrating phages
to be highly specific as to where they integrate. Random
integration into a bacterial genome could easily result in a
lethal mutation killing the cell and the phage along with it.
Demonstrating the specificity achieved by a phage integrase,
a study done in three different human cell lines of hundreds
of ?C31 integrase integration sites revealed the presence of
distinct hotspots. Of these hotspots, 19 integration sites ac-
counted for ~56% of the integration events . These pre-
ferred sites were scattered across intergenic regions, introns,
and exons of the genome in approximately the same propor-
tions that these elements existed in the genome, with a slight
preference for transcribed regions. Since open chromatin has
been proposed to be more accessible for integration, this
feature could explain the robust expression typically seen for
genes integrated with ?C31 integrase [4,5]. We and others
have taken advantage of the natural properties of the ?C31
integrase to integrate plasmid DNA in a sequence-specific,
unidirectional manner into mammalian genomes. The use of
phage integrases for gene therapy has been reviewed previ-
ously [6-8]. Here, we focus on developments that have oc-
curred since 2006.
One way to overcome these difficulties is to use an inte-
2 Current Gene Therapy, 2011, Vol. 11, No. 5 Chavez and Calos
THE MODE OF ACTION OF ? ?C31 INTEGRASE
?C31 integrase is a site-specific recombinase originating
in a bacteriophage of Streptomyces species [9,10]. It is a
member of the serine-catalyzed recombinase family that in-
cludes the ?? and Tn3 resolvases of E. coli, as well as the
integrases from the phages R4, TP109, and many others.
These integrases are similar in structure to the serine resol-
vases and invertases in the amino terminal portion of the
proteins. However, the serine integrases contain much larger
carboxy terminal regions (300-500 amino acids) and are
therefore called large serine recombinases . Serine re-
combinases are distinct, both mechanistically and evolution-
arily, from the other group of site-specific recombinases,
which are tyrosine-catalyzed. This latter group includes the
commonly used Cre and FLP enzymes, as well as lambda
In its host cell, ?C31 integrase catalyses the precise uni-
directional integration of the ?C31 phage genome into the
bacterial genome . Integration is mediated through two
~30 bp attachment sequences, termed attB and attP, that are
~50% identical. The recombination reaction is unidirectional
and requires no host cofactors. Upon integration, two hybrid
sites, attL and attR, are generated, which are not substrates of
?C31 integrase . The lack of a requirement for cofactors
to mediate recombination is a feature of both Cre and FLP,
but not of lambda integrase. However, the recognition sites
for Cre and FLP are identical, resulting in the recombined
site being a substrate for recombination. Thus, these en-
zymes are better suited for sequence deletion than for inte-
gration. The ?C31 integrase system integrates the entire
therapeutic plasmid, including the undesired plasmid back-
bone and antibiotic resistance gene. However, these se-
quences can be removed if desired, for example by using the
Cre-lox system. The properties of unidirectional integration,
relatively large recognition sites, and the lack of a require-
ment for cofactors make ?C31 integrase attractive for use in
mammalian gene therapy.
Our laboratory has demonstrated that ?C31 integrase
functions efficiently in many different mammalian cell types
. This enzyme, when expressed from the CMV promoter,
was able to mediate precise recombination between attB and
attP sites when present on the same or on different non-
integrated plasmids. The minimal sizes of the att sites were
determined to be 34 bp for attB and 39 bp for attP .
When the ?C31 attP site was placed into the genome of a
human cell line by random integration, ?C31 integrase was
able to complete recombination between attP and an attB-
bearing plasmid . The reverse reaction, placing the attB
site in the genome and attempting to integrate an attP-
containing plasmid, was much less efficient.
Because ?C31 integrase could recognize and recombine
an attB plasmid with a genomically placed attP site, we in-
vestigated whether attB could also recombine with sites that
shared sequence similarity with the native att sites. To test
this hypothesis, we transfected cells with a plasmid encoding
?C31 integrase, along with a second, or donor, plasmid car-
rying either the attP or attB sequence and a selectable drug
resistance gene. When the donor plasmid contained the attP
sequence, integration above background was not observed.
However, when the donor plasmid carried the attB sequence,
integration was observed at levels 5-10 fold above back-
ground (12). These experiments suggested that ?C31 inte-
grase was unable to utilize sites similar to the native attB site
for integration, but could make of use of sites with sequences
similar to the native attP site (pseudo attP sites). The paucity
of pseudo attB sites is beneficial for ?C31 integrase as a
gene therapy tool. If ?C31 integrase were to recognize and
recombine both pseudo attP and pseudo attB sequences effi-
ciently, rearrangements of the genome might occur at unac-
ceptable levels. The apparent absence of pseudo attB sites
may be a result of the ?C31 phage having evolved to inte-
grate very specifically at the native attB site, to avoid killing
its bacterial host through off-target integration. Attempts
have been made to modify ?C31 integrase activity and speci-
ficity [16,17]. However, to date none of these modified
forms have proven to be superior to wild-type ?C31 inte-
grase in practical applications. Likewise, an inducible form
of ?C31 integrase has been developed , but has not been
adopted in therapeutic applications.
USE OF ? ?C31 INTEGRASE IN IN VIVO GENE
of ?C31 integrase. The initial studies examining the function
of ?C31 integrase for use in gene therapy in animals were
published in 2002. Olivares et al.  demonstrated delivery
of naked plasmid DNA to mouse liver hepatocytes and se-
quence-specific integration of a human factor IX gene into
the mouse genome. Since that time, numerous studies have
verified the effectiveness of ?C31 integrase as a gene trans-
fer agent in gene therapy-related experiments [20-28]. To
test ?C31 integrase in an animal model, the newly developed
technique of hydrodynamic tail vein injection was utilized
for liver delivery [24, 25]. In this method, plasmid DNA is
Table 1. Some Recent Applications of the ? ?C31 Integrase
Table 1 provides an overview of some recent applications
Application Reference; number
Factor IX gene therapy in mouse liver Keravala et al., 2011; 31
Impact of hydrodynamic injec-
tion/?C31 on tumor latency
Woodard et al., 2010; 40
Longevity of ?C31 integrase expres-
sion in mouse liver
Chavez et al., 2010; 36
?C31 integrase-mediated integration in
Aneja et al., 2007; 41
Effectiveness of ?C31 in skeletal mus-
cle precursor cells
Quenneville et al., 2007; 42
Biosaftey of ?C31 in cord-lining
Sivalingam et al., 2010; 43
?C31 integrase-mediated integration in
mouse neural progenitor cells
Keravala et al., 2008; 45
Generation of iPS cells using ?C31
Ye et al., 2010;
Karow et al., 2011; 46, 48
Therapeutic Applications of the PhiC31 Integrase System Current Gene Therapy, 2011, Vol. 11, No. 5 3
diluted in a large volume (2 ml) of buffer and rapidly in-
jected over 3-6 sec into the tail vein of an adult mouse. This
volume is approximately equal to the blood volume of the
mouse and results in the creation of transient high pressure
within the liver. This pressure causes a subpopulation of he-
patocytes to take up the plasmid DNA, resulting in up to
10% or more of the hepatocytes becoming transfected [29-
30]. When using the ?C31 integrase system for liver gene
therapy, equal concentrations of two separate plasmids have
been co-injected into the mouse liver by hydrodynamic tail
vein injection. One plasmid encodes the ?C31 integrase
gene, driven by the cytomegalovirus immediate early pro-
moter. This plasmid does not become integrated into the ge-
nome. The second, or donor, plasmid contains an attB site
for genomic integration, as well as the therapeutic gene of
interest Fig. (1).
to develop a gene therapy treatment for a genetic disorder.
Hemophilia is an appealing candidate because only low lev-
els (1-5% of normal) of the missing clotting factor are re-
quired to achieve a therapeutic benefit and prevent spontane-
ous bleeding episodes. In recent experiments, factor IX
knockout mice were used as subjects, and attB-containing
human factor IX and ?C31 integrase plasmids were co-
injected by the hydrodynamic injection method . Plasma
factor IX levels were monitored over a 168-day period and
were found to stabilize at just under 1,000 ng/ml by two
months post-injection. Factor IX activity levels in treated
mice stabilized at about 10% of normal, well above the
therapeutic threshold. When the factor IX plasmid was in-
jected alone (no ?C31 integrase present), factor IX levels
stabilized at about 100 ng/ml, and there was little or no
measurable factor IX activity in the plasma of the mice by
two weeks post-injection. Therefore, long-term factor IX
expression required ?C31 integrase-mediated integration. A
parallel approach using a factor VIII expression cassette and
?C31 integrase has also been effective in a mouse model for
hemophilia A .
Because ?C31 integrase evolved to function in a pro-
karyotic environment, it was unknown where in mammalian
cells ?C31 integrase protein would localize and how this
might affect its function in mammalian hepatocytes or other
cells. To address this issue, we used an HA-tagged version of
In our lab, hemophilia B was chosen as the first attempt
the ?C31 integrase and immunofluorescence to determine
the localization of the ?C31 integrase protein within the
HeLa human cell line . Results from these experiments
showed that wild-type ?C31 integrase localized to the cyto-
plasm, and not to the nucleus where the genomic DNA was
located. We hypothesized that forced localization to the nu-
cleus might increase the number of cells undergoing a ?C31
integrase-catalyzed transgene integration. Thus, we added
the SV40 nuclear localization signal (NLS) sequence to the
?C31 integrase coding sequence. Using immunofluores-
cence, the ?C31 integrase-NLS protein was indeed found to
localize to the nucleus. Unexpectedly, however, addition of
the SV40-NLS resulted in an 80% decrease in intermolecular
integration efficiency when compared to the wild-type ?C31
integrase, as tested in a HeLa cell colony-forming assay .
By contrast, the addition of a NLS to ?C31 integrase was
found to increase the efficiency of intramolecular recombina-
tion when tested in reporter cell lines [34,35].
When the ?C31 integrase-NLS plasmid was co-injected with
an attB-containing factor IX expression plasmid into mouse
liver, no increase in plasma factor IX levels were detected,
compared to wild-type ?C31 integrase. Therefore, nuclear
localization was not beneficial for in vivo liver gene therapy,
possibly because nuclear localization of the bulk of the ?C31
integrase protein is not rate-limiting for integration effi-
ciency. Since cytosolic ?C31 integrase might gain access to
the genomic DNA during mitosis due to the breakdown of
the nuclear membrane, we tested whether cell division was
necessary for integration . Hepatocytes were labeled
with iododeoxyuridine (IdU) following hydrodynamic injec-
tion. IdU becomes incorporated into newly synthesized DNA
and therefore marks cells that have undergone cell division.
Thirteen weeks post-injection, the livers of treated mice were
harvested and subsequently analyzed for the presence of
cells that were positive for IdU, factor IX, or both. The re-
sults indicated that 71% of factor IX-positive cells were IdU
negative. These factor IX-positive cells were not due to the
presence of unintegrated plasmid, because a bacterial colony
forming assay showed no factor IX plasmid in the liver at
this time point . This result strongly suggested that ?C31
integrase-mediated plasmid integration did not require cell
division in the mouse liver.
The nuclear-localized integrase was also tested in vivo.
Fig. (1). Mechanism of ? ?C31-mediated integration. In nature, ?C31 integrase mediates the precise integration of the ?C31 phage genome
into the Streptomyces chromosome. In mammalian cells, ?C31 integrase can integrate an attB-containing plasmid into genomic pseudo attP
sites, i.e. endogenous sequences that share similarity with the native attP sequence.
4 Current Gene Therapy, 2011, Vol. 11, No. 5 Chavez and Calos
potential use in liver gene therapy was its safety profile in
the liver. Under the artificial conditions of tissue culture, an
elevated level of chromosomal rearrangement and large dele-
tions flanking integration sites were reported with ?C31 in-
tegrase expression [4, 37-39]. This has lead to concerns
about the safety of ?C31 integrase. To determine if hydrody-
namic injection alone or in conjunction with ?C31 integrase
could contribute to tumorigenesis, a transgenic mouse in
which C-MYC was conditionally expressed in the liver un-
der the control of the Tet-Off system was utilized . In
this study, mice received water containing doxycycline to
repress C-MYC expression from birth. At 7-8 weeks after
birth, doxycycline was withdrawn to allow continuous ex-
pression of C-MYC within the liver. One week after with-
drawal of doxycycline, the mice were hydrodynamically
injected with saline alone, an integrating luciferase plasmid
plus or minus ?C31 integrase, or with ?C31 integrase alone.
Following injection, the mice were monitored for tumor for-
mation. When large tumors were detected, the mice were
sacrificed and the livers harvested for analysis. Mice receiv-
ing no injections had a tumor latency of 154 days. Mice that
received hydrodynamic injections of saline only, had a tumor
latency of only 105 days. This decrease in tumor latency
indicated that the effect of hydrodynamic injection alone
could speed up tumor formation . To test if the presence
of DNA itself had an effect on tumor formation, we per-
formed hydrodynamic injection of the empty expression vec-
tor pCS plus a luciferase expressing plasmid. There was no
significant difference between this group and the saline only
group, suggesting that DNA itself was not tumorigenic. In-
terestingly, when the injection contained either active or
catalytically inactive ?C31 integrase, tumor latency was de-
termined to be the same 153 days as the untreated group,
suggesting that integrase did not increase the frequency of
tumors and even had a protective effect. Furthermore, none
of the tumors examined had integrations of the luciferase
plasmid as determined by polymerase chain reaction. There-
fore the initial tumor cell did not arise from a recombination
event facilitated by ?C31 integrase . Overall, the study
provided no evidence to suggest that the presence of ?C31
integrase or integration mediated by the enzyme caused any
increase in tumor formation.
Even though chromosomal rearrangements due to ?C31
integrase expression in vivo have not been reported to date,
prolonged expression of ?C31 integrase is considered unde-
sirable, as it may promote chromosomal instability and/or
initiate an immune response to the integrase protein. We set
out to determine how long ?C31 integrase was expressed,
both within the cells of the mouse liver following hydrody-
namic delivery and in cultured cells . It was determined,
by both Western blot and immunofluorescence, that expres-
sion of ?C31 integrase was detectable within the liver of
mice two hours post-injection. Somewhat surprisingly, inte-
grase expression was undetectable by these two methods at
48 hours post-injection. Plasmid integration was observed at
the mpsL1 genomic hot spot three hours after hydrodynamic
injection and remained detectable by PCR throughout the six
month duration of the experiment. These experiments dem-
onstrated how quickly ?C31 integrase can mediate perma-
nent genomic modification. The retention time of the ?C31
Another aspect of ?C31 integrase that was relevant for its
integrase plasmid within the liver was also determined. The
integrase plasmid was detectable by Southern blot from one
hour to 32 hours post-injection within the liver. However, the
integrase gene was detectable by PCR, albeit at low levels,
over the six-month course of the experiment. When tested in
tissue culture cells, ?C31 integrase expression ranged from
three days in HeLa cells to one week in HEK 293 cells. The
extended in vitro ?C31 integrase expression, compared to in
vivo expression, may contribute to the reported chromosomal
instabilities in cultured cells.
?C31 integrase can also function in mouse lung tissue.
This activity was demonstrated in a murine alveolar epithe-
lial cell line (MELE12) through an intramolecular recom-
bination assay . It was then shown that when complexed
with polyethylenimine (PEI) and injected intravenously,
?C31 integrase could mediate genomic integration and long-
term expression of a luciferase-containing plasmid within the
lungs of mice. Genomic integration was later verified by
PCR analysis, which showed that seven of the fifteen mice
treated had experienced integration at the mpsL1 mouse hot
spot within the lungs .
EX VIVO USE OF ? ?C31 INTEGRASE
The ?C31 integrase system is increasingly being used to
introduce therapeutic genes stably into cultured cells that
will later be transplanted into the animal. One study investi-
gated the effectiveness of ?C31 integrase in skeletal muscle
precursor cells . Both mouse and human muscle precur-
sor cells were co-transfected with ?C31 integrase and a GFP
plasmid, leading to long-term in vitro GFP expression .
After G418 selection, the muscle precursor cells were in-
jected into the tibialis anterior muscle of an irradiated mouse
and were shown to fuse with the existing muscle fibers. Ex-
pression of a GFP mini-dystrophin fusion protein was also
demonstrated in vitro. Muscle precursor cells carrying this
construct were able to engraft into mouse muscle after injec-
tion. A GFP- full-length dystrophin fusion plasmid was also
constructed and was integrated into both mouse and human
muscle precursor cells. Upon transplantation into the tibialis
anterior muscles of mice, the modified cells fused with the
existing muscle fibers, and dystrophin expression was re-
stored . Thus, ?C31 integrase can be used to modify
muscle precursor cells with a therapeutic gene, while not
compromising the ability of the cells to engraft.
Another recent paper evaluated the genotoxicity of ?C31
integrase in cord-lining epithelial cells (CLECs) .
CLECs isolated from the outer membrane of human umbili-
cal cords were found to express the pluripotency markers
Oct-4 and Nanog. The cells were found to transfect well and
to have a ?C31 integrase-mediated integration efficiency of
3.0%. Forty-four documented independent integration events
were analyzed and found to map to 18 distinct genomic loci.
Genomic integration loci were determined by plasmid rescue
on pooled populations of clones, and like a previous study
performed in this manner , it was concluded that 8p22
was the most frequently recovered pseudo attP site. We note
that when individual clones, rather then pools were analyzed
, it was observed that the most common ?C31 integrase-
mediated integrations take place in endogenous retrovirus
elements, which make up approximately 8% of the human
Therapeutic Applications of the PhiC31 Integrase System Current Gene Therapy, 2011, Vol. 11, No. 5 5
genome . The most frequently used single-locus human
hot spot was an intronic locus at 19q13.31 .
unmodified and ?C31 integrase-modified CLECs were found
to display no differences in the expression of over 96.6% of
transcripts . High-resolution genome copy number
analysis was conducted on these two populations, and copy
gain in two loci and loss in one other were identified. This
result was in contrast to the multiple copy number changes
seen in many types of cancers. Spectral karyotyping of ?C31
integrase-modified cells revealed that 4 out of 90 metaphase
spreads had translocations present in a polyclonal popu-
lation. However, no chromosomal abnomalities were
detected in over 210 metaphase spreads from eight clonal
lines of CLECs . Fluorescence in situ hybridization of
over 200 ?C31 integrase modified cells showed that >85% of
the cells had one or two integrations per cell. This feature
allows one potentially to select single integrant cells with
safe integration sites for clinical use. The ?C31 integrase-
modified cells were also shown not to have altered prolife-
rative behavior or to be tumorigenic in NOD-SCID mice. A
human factor VIII expression plasmid was then integrated in
the CLECs with ?C31 integrase. After implantation of the
modified CLECs in factor VIII knockout mice, plasma factor
VIII levels of 3% of normal were detected three days post-
The transcriptional profile of polyclonal populations of
in mouse neural progenitor cells. When a donor plasmid con-
taining both an attB and a luciferase reporter gene were in-
troduced into mouse neural progenitor cells along with a
?C31 integrase-expressing plasmid, sequence-specific inte-
gration was detected by PCR at the mouse genomic integra-
tion hot spot mpsL1 . Additionally, long-term in vitro
luciferase expression was detected within the cells for the
duration of the experiment (56 days). The gene-modified
cells displayed no loss of the ability to proliferate or to dif-
ferentiate into neurons and astrocytes.
?C31 integrase has recently been used to generate
induced pluripotent stem cells (iPS) from mouse embryonic
fibroblasts, as well as from human amniotic fluid cells .
To generate mouse iPS cells, plasmids contaning the four
Yamanaka reprogramming factors  were integrated into
the genome of mouse embryonic fibroblast cells using the
?C31 integrase system. iPS colonies were detected after 14-
20 days in culture. When characterized, these cells displayed
common markers of pluripotency, such as expression of
Nanog and SSEA1, and formed teratomas in immuno-
deficient mice. To reprogram the human amniotic fluid cells,
the same reprogramming plasmids were integrated, again
using the ?C31 integrase system. iPS colonies were first
visible nine days post-transfection and were picked on days
35-40. When tested, the amniotic fluid-derived iPS cells
displayed the characteristic pluripotency markers, gene
expression profiles, and methylation patterns. The iPS cells
were also able to differentiate into the three germ layers of
endoderm, mesoderm, and ectoderm. Integration sites from
both cell types were obtained by plasmid rescue and were
determined to be in seemingly safe genomic locations.
A similar strategy for generating iPS cells was also
developed in the Calos lab . The ?C31 integrase system
Long-term transgene expression has also been achieved
was used to integrate a reprogramming cassette into the
genome of mouse embryonic fibroblasts and adipose-derived
mesenchymal stem cells. Several iPS cell lines were isolated,
and integration sites were characterized at the sequence
level. The iPS clones were demonstrated to be pluripotent by
in vitro and in vivo methods, including the generation of
chimeric mice. An advantage of this system is that the
reprogramming cassette was flanked by loxP sites, which
allowed the reprogramming factors to be excised using Cre
recombinase . Excision of the reprogramming factors
resultd in an iPS cell clone free of most of the exogenous
DNA, and this may be a safer, more clinically acceptable cell
for use in a cell therapy treatment.
demonstrated that hydrodynamic delivery of the ?C31
integrase system leads to robust long-term expression of
transgenes within the livers of mice. If an appropriate liver
delivery method is developed that is safe and effective in
large animals, there is the potential for clinical translation of
these studies for diseases such as hemophilia. The ?C31
integrase system has also been shown to be effective in a
variety of stem cells. An exciting application is use of the
system to generate iPS cells. Patient-specific iPS cells could
potentially be genetically corrected with a second recom-
binase and differentiated into the disease-appropriate cell
type. While this ex vivo application removes many of the
DNA delivery barriers of in vivo gene therapy, extensive
further studies will be needed to determine effective
differentiation and implantation procedures for different cell
types and diseases.
As reviewed here, numerous studies have now
National Institutes of Health, the Jain Foundation, the Mus-
cular Dystrophy Association, and the California Institute for
Regenerative Medicine to MPC. CLC was supported in part
by a Dean’s Postdoctoral Fellowship from Stanford Univer-
sity School of Medicine. MPC is an inventor on Stanford-
owned patents covering ?C31 integrase.
We gratefully acknowledge research grants from the
att site = Attachment site
attB = Bacterial attachment site
attP = Phage attachment site
bp = Base pair
GFP = Green fluorescent protein
NLS = Nuclear localization signal
IdU = Iododeoxyuridine
CLECs = Cord-lining epithelial cells
iPS cells = Induced pluripotent stem cells
Li Z, Dullman J, Schiendlmeier B, et al. Murine leukemia induced
by retroviral gene marking. Science. 2002;296:497-.
6 Current Gene Therapy, 2011, Vol. 11, No. 5 Chavez and Calos
Hacein-Bey-Abina S, Von Kalle C, Schmidt M, et al. LMO2-
associated clonal T cell proliferation in two patients after gene
therapy for SCID-X1. Science. 2003;302:415-9.
Davis D, Stokoe D. Zinc finger nucleases as tools to understand
and treat human diseases. BMC Med. 2010;8:42.
Chalberg TC, Portlock JL, Olivares EC, et al. Integration
specificity of phage phiC31 integrase in the human genome. J Mol
Biol. 2006;357:28 - 48.
Thyagarajan B, Calos MP. Site-specific integration for high-level
protein production in mammalian cells. Therapeutic Proteins:
Methods and Protocols. 2005;308:99-106.
Calos MP. The phiC31 integrase system for gene therapy. Current
Gene Therapy. 2006;6:633-45.
Ginsburg DS, Calos MP. Site-specific integration with phiC31
integrase for prolonged expression of therapeutic genes. Non-viral
vectors for gene therapy, Second edition: Part II. 2005;54:179-87.
Groth AC, Calos MP. Phage integrases: biology and applications. J
Mol Biol. 2004;335:667-78.
Kuhstoss S, Rao RN. Analysis of the integration function of the
Streptomycete bacteriophage PhiC31. J Mol Biol. 1991;222:897-
Rausch H, Lehmann M. Structural analysis of the actinophage
PhiC31 attachment site. Nucleic Acids Research. 1991;19:5187-9.
Smith MCM, Thorpe HM. Diversity in the serine recombinases.
Molec Microbiol. 2002;44:299-307.
Grindley NDF, Whiteson KL, Rice PA. Mechanisms of site-
specific recombination. Annual Review Biochem. 2006;75:567 -
Thorpe HM, Smith MCM. In vitro site-specific integration of
bacteriophage DNA catalyzed by a recombinase of the
resolvase/invertase family. Proc Natl Acad Sci USA. 1998; 95:
Groth AC, Olivares EC, Thyagarajan B, Calos MP. A phage
integrase directs efficient site-specific integration in human cells.
Proc Natl Acad Sci USA. 2000;97:5995-6000.
Thyagarajan B, Olivares EC, Hollis RP, Ginsburg DS, Calos MP.
Site-specific genomic integration in mammalian cells mediated by
phage phiC31 integrase. Molecular and Cellular Biology.
Keravala A, Lee S, Thyagarajan B, et al. Mutational derivatives of
phiC31 integrase with enhanced efficiency and specificity.
Molecular Therapy. 2009;17:112-20.
Liesner R, Zhang W, Noske N, Ehrhardt A. Critical amino acid
residues within the phiC31 integrase DNA-binding domain affect
recombination activities in mammalian cells. Hum Gene Ther.
Sharma N, Moldt B, Dalsgaard T, Jensen T, Mikkelsen J.
Regulated gene insertion by steroid-induced phiC31 integrase.
Nucleic Acids Res. 2008.
Olivares EC, Hollis RP, Chalberg TW, Meuse L, Kay MA, Calos
MP. Site-specific genomic integration produces therapeutic factor
IX levels in mice. Nature Biotechnology. 2002;20:1124-8.
Chalberg TC, Genise HL, Vollrath D, Calos MP. PhiC31 integrase
confers genomic integration and long-term transgene expression in
rat retina. Investigative Ophthalmology & Visual Science.
Ehrhardt A, Xu H, Huang Z, Engler JA, Kay MA. A direct
comparison of two nonviral gene therapy vectors for somatic
integration: in vivo evaluation of the bacteriophage integase
phiC31 and the Sleeping Beauty transposase. Molecular Therapy.
Held PK, Olivares EC, Aguilar CP, Finegold M, Calos MP,
Grompe M. In vivo correction of murine hereditary tyrosinemia
type I by phiC31 integrase -mediated gene delivery. Molecular
Ishikawa Y, Tanaka N, Murakami K, et al. Phage phiC31 integrase-
mediated genomic integration of the common cytokine receptor
gamma chain in human T-cell lines. J Gene Medicine. 2006;8:646 -
Keravala A, Portlock J, Nash JA, Vitrant DG, Robbins PD, Calos
MP. PhiC31 integrase mediates integration in cultured synovial
cells and enhances gene expression in rabbit joints. J Gene Med.
2006;8:1008 - 17.
Ortiz-Urda S, Keene D, Lin Q, Calos MP, Khavari P. PhiC31
integrase-mediated nonviral genetic correction of junctional
epidermolysis bullosa. Human Gene Therapy. 2003;14:923-8.
Ortiz-Urda S, Lin Q, Green CL, Keene DR, Marinkovich MP,
Khavari P. Injection of genetically engineered fibroblasts corrects
regenerated human epidermolysis bullosa skin tissue. J Clin Invest.
Portlock JL, Keravala A, Bertoni C, Lee S, Rando TA, Calos MP.
Long-term increase in mVEGF164 in mouse hindlimb muscle
mediated by phage phiC31 integrase after nonviral DNA delivery.
Human Gene Therapy. 2006;17:871 - 6.
Quenneville SP, Chapdelaine P, Rousseau J, et al. Nucleofection of
muscle-derived stem cells and myoblasts with phiC31 integrase:
stable expression of a full-length-dystrophin fusion gene by human
myoblasts. Mol Ther. 2004;10:679-87.
Liu F, Song YK, Liu D. Hydrodynamics-based transfection in
animals by systemic administration of plasmid DNA. Gene
Zhang G, Budker V, Wolff JA. High levels of foreign gene
expression in hepatocytes after tail vein injections of naked plasmid
DNA. Human Gene Therapy. 1999;10:1735-7.
Keravala A, Chavez, CL, Hu, G, Woodard, LE, Monahan, PE,
Calos, MP. Long -term phenotypic correction in factor IX knockout
mice by using PhiC31 integrase-mediated gene therapy. Gene
Therapy. 2011; in press.
Chavez CL, Keravala A, Chu JN, et al. Long-term expression of
human coagulation factor VIII in a tolerant mouse model using the
phiC31 integrase system. Hum Gene Ther. 2011.
Woodard LE, Hillman, RT, Keravala A, Lee, S, Calos, MP. Effect
of nuclear localization and hydrodynamic delivery-induced cell
division on phiC31 integrase activity. Gene Therapy. 2010;17: 217-
Andreas S, Schwenk F, Kuter-Luks B, Faust N, Kuhn R. Enhanced
efficiency through nuclear localization signal fusion on phage
PhiC31-integrase: activity comparison with Cre and FLPe
recombinase in mammalian cells. Nucleic Acids Res. 2002 Jun
Zhang MX, Li ZH, Fang YX, et al. TAT-phiC31 integrase mediates
DNA recombination in mammalian cells. J Biotechnol. 2009 Jun
Chavez CL, Keravala A, Woodard LE, et al. Kinetics and longevity
of phiC31 integrase in mouse liver and cultured cells. Hum Gene
Ther. 2010 Oct;21(10):1287-97.
Ehrhardt A, Engler J, Xu H, Cherry A, Kay M. Molecular analysis
of chromosomal rearrangements in mammalian cells after phiC31-
mediated integration. Human Gene Therapy. 2006;17:1077-94.
Liu J, Jeppesen I, Nielsen K, Jensen TG. PhiC31 integrase induces
chromosomal aberrations in primary human fibroblasts. Gene
Therapy. 2006;13:1188 - 90.
Liu J, Skjorringe T, Gjetting T, Jensen TG. PhiC31 integrase
induces a DNA damage response and chromosomal rearrangements
in human adult fibroblasts. BMC Biotechnol. 2009;9:31.
Woodard LE, Keravala A, Jung WE, et al. Impact of hydrodynamic
injection and phiC31 integrase on tumor latency in a mouse model
of MYC-induced hepatocellular
Aneja MK, Imker R, Rudolph C. Phage phiC31 integrase-mediated
genomic integration and long-term gene expression in the lung
after nonviral gene delivery. J Gene Med. 2007 Nov;9(11):967-75.
Quenneville S, Chapdelaine P, Rousseau J, Tremblay J. Dystrophin
expression in host muscle following transplantation of muscle
precursor cells modified with the phiC31 integrase. Gene Therapy.
2007 Nov. 30;14:514-22.
Sivalingam J, Krishnan S, Ng WH, Lee SS, Phan TT, Kon OL.
Biosafety assessment of site-directed transgene integration in
human umbilical cord-lining cells. Mol Ther. 2010 Jul;18(7):1346-
Weiss RA. The discovery
Keravala A, Ormerod BK, Palmer TD, Calos MP. Long-term
transgene expression in mouse neural progenitor cells modified
with phiC31 integrase. J Neurosci Methods. 2008 Aug
Ye L, Chang JC, Lin C, Qi Z, Yu J, Kan YW. Generation of
induced pluripotent stem cells using site-specific integration with
phage integrase. Proc Natl Acad Sci U S A. 2010 Nov
carcinoma. PLoS One.
of endogenous retroviruses.
Therapeutic Applications of the PhiC31 Integrase System Current Gene Therapy, 2011, Vol. 11, No. 5 7 Download full-text
Takahashi K, Yamanaka S. Induction of pluripotent stem cells from
mouse embryonic and adult fibroblast cultures by defined factors.
Cell. 2006 Aug 25;126(4):663-76.
Karow M, Chavez CL, Farruggio AP, et al. Recombinase strategy
to make and modify iPS cells. 2011. Stem Cells, in press.
Received: February 18, 2011 Revised: July 15, 2011 Accepted: July 19, 2011