Overview: Generation of Gene Knockout UNIT 19.

12
Mice
Bradford Hall,1 Advait Limaye,1 and Ashok B. Kulkarni1
1
National Institute of Dental and Craniofacial Research, National Institutes of Health,
Department of Health and Human Services, Bethesda, Maryland

ABSTRACT
The technique of gene targeting allows for the introduction of engineered genetic mutations into
a mouse at a determined genomic locus. The process of generating mouse models with targeted
mutations was developed through both the discovery of homologous recombination and the
isolation of murine embryonic stem cells (ES cells). Homologous recombination is a DNA repair
mechanism that is employed in gene targeting to insert a designed mutation into the homologous
genetic locus. Targeted homologous recombination can be performed in murine ES cells through
electroporation of a targeting construct. These ES cells are totipotent and, when injected into
a mouse blastocyst, they can differentiate into all cell types of a chimeric mouse. A chimeric
mouse harboring cells derived from the targeted ES cell clone can then generate a whole mouse
containing the desired targeted mutation. The initial step for the generation of a mouse with a
targeted mutation is the construction of an efficient targeting vector that will be introduced into
the ES cells. Curr. Protoc. Cell Biol. 44:19.12.1-19.12.17. C 2009 by John Wiley & Sons, Inc.

Keywords: gene targeting r homologous recombination r gene knockout

INTRODUCTION coat color. Therefore, the introduced stem cells

Two discoveries have been instrumental could become established into the germline of
in creating the ability to generate knockout the chimeric mice. The isolation of stem cells
mice—the isolation of stem cells and the dis- basically meant that an ES cell clone that was
covery of homologous recombination. The genetically modified in culture could eventu-
significance of these findings was demon- ally be used to derive a mouse.
strated when Mario Capecchi, Oliver Smithies, The second important step in the estab-
and Martin Evans were awarded the 2007 lishment of knockout mice was the discovery
Nobel Prize in Physiology or Medicine for of homologous recombination within mam-
their work in establishing the knockout mouse malian cells (also see Mortensen et al., 2006,
model (Vogel, 2007). The key step needed for for basic principles of this technique). Lin
making knockout mice was the isolation of et al. (1984) were first able to demonstrate that
embryonic stem cells (ES cells). The ES cells homologous recombination was feasible by
were isolated from the inner cell mass of a 3.5- reconstructing a functional thymidine kinase
day post-coitum mouse blastocyst (Evans and gene in mouse L cells. Later, Smithies et al.
Kaufman, 1981; Martin, 1981). These stem (1985) were able to modify the β-globin locus
cells could progressively grow in tissue cul- with the insertion of the neomycin-resistance
ture and were pluripotent (also see UNIT 19.1). gene. This work provided a means for specif-
By injecting the stem cells into a mouse blasto- ically targeting the DNA insert to a planned
cyst, Bradley et al. (1984) were able to gener- genetic locus, in contrast to the random inte-
ate chimeric mice. The production of chimeric gration of a construct that is used with trans-
mice proved that the stem cells were able to genic technology.
differentiate into multiple cell lineages and The convergence of these two tech-
were able to contribute to the development nologies provided the means to generate
of the mouse embryo. Germline transmission knockout mice. The first ideal genetic lo-
was also achieved by using the pluripotent ES cus to test gene targeting was for the en-
cells. After breeding the chimeric mice, the zyme hypoxanthine-guanosine phosphoribo-
resulting offspring were clearly derived from syl transferase (HPRT). Loss of the HPRT
the ES cells, as seen by the transmission of gene could be tested with treatment using Whole Organism
and Tissue
Analysis

Current Protocols in Cell Biology 19.12.1-19.12.17, September 2009 19.12.1

Published online September 2009 in Wiley Interscience (www.interscience.wiley.com).
DOI: 10.1002/0471143030.cb1912s44 Supplement 44
Copyright C 2009 John Wiley & Sons, Inc.
 6-thioguanine, while restoration of the gene Therefore, gene inactivation is the best way
could be selected in HPRT-null cells with addi- to delineate the biological role of a protein,
tion of hypoxanthine/aminopterin/thymidine and gene targeting is a direct means to dis-
(HAT) medium. In addition, the HPRT gene rupt a gene’s open reading frame and block
is X-linked so that, with male stem cells, only its expression in a mouse. Not surprisingly,
one allele needed to be targeted for drug selec- during the twenty years that gene-targeting
tion. Doetschman et al. (1987) proved that a techniques have been available, thousands of
mutant HPRT gene could be corrected through genes have been knocked out. To date, about
homologous recombination in ES cells, while 11,000 genes have been knocked out in mice,
Thomas and Capecchi (1987) demonstrated which accounts for roughly half of the mouse
targeted gene disruption. These experiments genome (Sikorski and Peters, 1997; Vogel,
led the way to targeting of nonselectable genes, 2007). Through a combination of gene target-
starting with int-2 and c-abl knockout mice ing and gene trapping, a global effort is under-
(Mansour et al., 1988; Schwartzberg et al., way to make a knockout mouse for each of the
1989). 25,000 mouse genes (Grimm, 2006).
Since the first knockouts, there has been The knockout mouse has been a valuable
an explosive growth in the number of animal tool for geneticists to discern the role of a
models derived through the technique of gene gene in embryonic development and in nor-
targeting. While mouse stocks with sponta- mal physiological homeostasis. Mice act as a
neous or radiation-induced mutations existed good analog for most human biological pro-
in the past, the mutated alleles did not corre- cesses, since humans and mice share ∼99% of
spond to the majority of biologically impor- the same genes (Capecchi, 1994). Addition-
tant cloned genes (Mak, 1998). Through ho- ally, mice are useful experimental animals be-
mologous recombination, a cloned gene could cause they are small, have relatively short life
be directly utilized to disrupt the target al- spans, and are prolific. So, for geneticists, the
lele. Currently, a targeting strategy is often targeted deletion of a gene in a mouse provides
planned shortly after the mouse gene is orig- an important means to determine the biologi-
inally cloned. Therefore, targeted recombina- cal role of a genetic allele.
tion has become well established as an impor- While useful for studying in vivo gene func-
tant tool to inactivate a gene in order to study tion, some knockout mice have also served
its function in vivo. as valuable animal models for human genetic
The process of gene targeting provides a diseases. When a human mutation is found
means to alter a specified gene in order to to disable a protein, the corresponding knock-
better discern its biological role. Through ho- out mouse can be an important resource to
mologous recombination, an engineered mu- study the underlying pathophysiology and to
tation can be directed to a designated genetic develop therapies to treat a genetic disease
locus. In this manner, a potentially important (Majzoub and Muglia, 1996). Additionally,
genomic clone can directly be utilized to cre- pharmaceutical companies obtain clues about
ate a mutation in a selected gene. Even within inhibiting a protein by first looking at the phe-
the 2.5 Gb of the mouse genome, the cellular notype of a knockout mouse (Zambrowicz and
DNA repair mechanisms are able to align a Sands, 2003). Thus, knockout mice can pro-
targeting vector with its corresponding region vide insight into a gene’s physiological role in
of homology and cause recombination into the humans.
chromosome. Rather than just inactivating a gene, how-
While the goal of transgenic technology ever, some genetic diseases result in the ex-
is to overexpress a gene to study its bio- pression of a mutated protein. Point mutations,
logical role in vivo, homologous recombina- microdeletions, or insertions are responsible
tion is typically employed to create a loss-of- for many human genetic diseases. These sub-
function mutation. The most common applica- tle mutations can also be mimicked in a mouse
tion of gene targeting is to produce knockout model using gene targeting. Instead of disrupt-
mice, where a drug-resistance marker replaces ing a gene, as in most knockout mice, homol-
an essential coding region in a genetic lo- ogous recombination is employed to swap the
cus. In the majority of cases, the importance normal copy of an exon with a mutated ver-
of a gene cannot be determined by simply sion. As long as a similar mutation can be
recognizing amino acid motifs in the pro- reproduced in the mouse protein, then the cor-
Generation of tein (Iredale, 1999). Additionally, the role of responding amino acid substitution, deletion,
Gene Knockout a gene often cannot be completely revealed or insertion can be targeted into a gene of inter-
Mice
by examining closely related family members. est to replicate the human disease. The effects
19.12.2
Supplement 44 Current Protocols in Cell Biology
of the altered protein can then be studied in the (rtTA; Utomo et al., 1999). Doxycycline is
animal model. administered to activate the rtTA which, in
With some knockout mouse models, the turn, will induce transcription of Cre. Lastly,
severity of the phenotype can preclude anal- Cre can also be delivered through injection
ysis of a gene’s role in the organogen- of a viral vector (Anton and Graham, 1995).
esis of a particular tissue. For example, When and where the Cre is expressed is con-
about 15% of all knockout mice have mu- trolled by the timing and site of injection
tations that result in developmental lethality for the virus. The amount of DNA rearrange-
(http://www.genome.gov/12514551). To cir- ment can be adjusted by varying the titer of
cumvent this problem, Cre/loxP technology the virus. The numbers of conditional knock-
has been employed to create conditional out mice have dramatically increased since
knockout mice. Derived from the P1 bacte- Cre/loxP and Flp/FRT technologies were in-
riophage, the Cre recombinase will excise any troduced into gene targeting. This prolifera-
region of DNA placed between two loxP sites tion of conditional animal models attests to
(locus of crossing over; Sternberg and Hamil- the value of the recombinases as a molecular
ton, 1981; Sauer and Henderson, 1998). The switch.
loxP site is a 34-bp nucleotide sequence that Another interesting application of gene tar-
can be genetically targeted around an essen- geting is knock-in technology, in which any
tial exon in a gene (Gu et al., 1994). The re- gene of interest can be placed under the cis-
sulting mice contain the floxed (flanked by acting regulatory elements of another gene
loxP sites) allele in all tissues, but are phe- (Cohen-Tannoudji and Babinet, 1998). Orig-
notypically wild-type. These floxed mice are inally, knock-in mice were derived as a means
then bred to Cre-expressing transgenic ani- to visualize a gene’s expression due to targeted
mals, wherein the promoter used to drive Cre recombination of the lacZ reporter gene into
expression will determine the site of the gene a genetic locus (LeMouellic et al., 1990). The
deletion. For spatial inactivation of a gene in initial coding sequence of a gene would be re-
a mouse, a cell-type-specific promoter is used placed with the lacZ marker, which is inserted
to limit Cre expression to a particular tissue. under the direction of the gene’s promoter.
Although less commonly used than The strategy of using homologous recombi-
Cre/loxP technology, the Flp recombinase also nation to knock in a reporter gene, like lacZ,
provides a similar means to rearrange a genetic allows not only for the creation of homozy-
locus. Flp (flippase) was isolated from Saccha- gous null mice for a gene, but also provides a
romyces cerevisiae and, like Cre, the recom- technique to study the targeted gene’s expres-
binase will also excise DNA flanked between sion in the heterozygous mice, which are often
34-bp sequences known, in this case, as FRT phenotypically normal. The knock-in idea was
sites (Dymecki, 1996). So, through the use of later elaborated to include the replacement of
either the Cre/loxP or the Flp/FRT systems, a gene by the sequence coding for a similar
gene expression can be disrupted in a spatial isoform of the protein (Hanks et al., 1995).
and temporal manner, and the lethality of a Similar protein isoforms could thus be tested
knockout mouse phenotype can be overcome. for redundancy. Through knock-in mice, ge-
With this versatility, mice utilizing Cre/loxP or netically similar proteins can be examined to
the Flp/FRT systems are often shared among determine if two isoforms are truly biologi-
research laboratories studying differing phys- cally distinct or if the proteins are basically
iological systems. functionally equivalent. with differing patterns
In addition to spatial excision of a floxed of expression being the only key divergence.
allele, temporal control of Cre-mediated re-
combination is also possible in a conditional VECTOR DESIGN
knockout mouse. The timing of recombina- Homologous recombination takes advan-
tion can be regulated by use of a tamoxifen- tage of a cell’s own DNA-repair machinery
inducible Cre (Feil et al., 1997; Hayashi and to replace a targeted genetic locus with ho-
McMahon, 2002). In this strategy, Cre is lig- mologous genomic sequence. The targeted
ated to a mutated ligand-binding domain of event probably occurs through strand transfer
the estrogen receptor that restricts transcrip- by the recombination RAD52 epistasis group
tion until tamoxifen is present. Cre is, there- (Rijkers et al., 1998; Vasquez et al., 2001).
fore, not expressed until tamoxifen is applied For recombination to occur at all in a cell,
either topically or through injection. Another around 2 kb of sequence homology is required Whole Organism
inducible Cre system takes advantage of the (Melton, 2002). However, 6 to 14 kb of ho- and Tissue
Analysis
reverse tetracycline-controlled transactivator mology is typical for targeting constructs. In
19.12.3
Current Protocols in Cell Biology Supplement 44
 unique restriction
enzyme site

homology
arm

neo r

homology
arm replacement plasmid
vector backbone

HSV-tk
unique restriction
enzyme site

homology arm neo r homology arm HSV-tk

Figure 19.12.1 A schematic of a replacement vector. Two homology arms flank a positive drug-
selection marker (neor ). A negative-selection marker (HSV-tk) is placed adjacent to one of the
targeting arms. A unique restriction-enzyme site is located between the vector backbone and
the homology arm. When linearized for gene targeting, the vector backbone will then protect the
HSV-tk from nucleases.

addition, linear DNA was found to be the tion of vector used in the targeting experiment
preferred substrate for the recombination pro- (Thomas et al., 1986; Capecchi, 2005). So,
teins (Hasty et al., 1992). The actual targeted the capacity to locate homologous sequence
event takes place in only a small percentage within the recipient cell genome is not a rate-
of cells, as homologous recombination occurs limiting step, since this would be influenced
at a rate ∼1000-fold lower than random in- by the input concentration of DNA. Rather,
sertions (Sargent and Wilson, 1998). During the cellular machinery performing the homol-
a stem cell experiment, only about 10–2 to ogous recombination sets the reaction rate for
10–3 of the DNA integrations are homologous gene targeting.
recombination events (Melton, 2002). There- Primarily, two classes of vectors have been
fore, a thorough screening process employ- used to generate targeted mutations. Most
ing Southern blotting or PCR is necessary to gene-targeting experiments employ replace-
identify cells with the targeted event. Electro- ment vectors (Fig. 19.12.1), which have been
poration became the preferred way to deliver particularly instrumental for efficiently gener-
the replacement vectors into a large number ating knockout mice. In the design of a re-
of stem cells. While microinjection has a bet- placement vector, the open reading frame of a
ter targeting ratio (1:15 targeted recombinants genomic clone is disrupted by the placement
to random integrants), a mass delivery sys- of an intervening drug-selection marker. Two
tem was needed to introduce the targeting con- homologous recombination events function to
structs into cells, and electroporation provided insert the targeting construct containing the
the most efficient technique (1:2400 targeting drug-resistance gene into a homologous ge-
ratio; Vasquez et al., 2001). However, since the netic locus (Fig. 19.12.2). The drug-resistance
transformation efficiency is low (10–3 ), a posi- gene works for the positive selection of
tive selection marker is needed to enrich clones cells that have integrated the targeting vec-
that have inserted the targeting vector into their tor into their chromosome. Neomycin is the
genome (Ledermann, 2000). The likelihood of most common drug used for positive selec-
recombination peaks when cells are in early to tion. Integration of the neomycin phospho-
mid-S phase, and the process occurs rapidly, transferase (neor ) gene allows for resistance
in ∼30 min after the construct is taken into the to neomycin, an aminoglycoside that inter-
nucleus (Wong and Capecchi, 1987; Cappec- feres with protein synthesis in eukaryotic cells.
Generation of chi, 1989). The rate of homologous recombi- Other drug-selection markers include resis-
Gene Knockout
Mice nation is not dependent on the input concentra- tance genes for puromycin and hygromycin.

19.12.4
Supplement 44 Current Protocols in Cell Biology
 neo r HSV-tk
targeting
vector

target locus
in chromosome

neo r
chromosome with
targeted replacement

Figure 19.12.2 Gene inactivation through a replacement vector. Homologous recombination

with a replacement vector requires a positive-selection marker (neor ), a negative-selection marker
(HSV-tk), and two targeting arms (homologous sequence is depicted with exons in light gray). A
representative target gene (with exons in dark gray) is aligned with the targeting vector. For this
example, the targeting vector is designed so that Exon 2 is substituted by the neor gene. The
replacement of Exon 2 by the neor gene is then recapitulated in the target locus as homologous
recombination exchanges genomic sequence for the homologous sequence of the targeting vector.
Two homologous recombination events (depicted through the crosses) occur via the long and
short targeting arms to introduce nonhomologous sequence (i.e., neor ) into the designated gene.
Insertion of the neor gene is selected for by treatment of cells with neomycin sulfate (G418) in
tissue culture. The negative-selection marker (HSV-tk) is not recombined into the chromosome
and is lost during gene targeting. If the targeting construct is randomly integrated anywhere in the
genome, the HSV-tk gene would be intact. Random integrants can be selected against by either
gancyclovir or FIAU treatment.

Mansour et al. (1988) established positive vector directs the disruption of an essential
and negative selection for gene targeting by coding exon (or exons) of a gene through the
the additional placement of the HSV thymi- insertion of a positive drug-resistance marker.
dine kinase (HSV-tk) gene adjacent to one of This replacement occurs via recombination at
the vector arms containing targeted homol- the two flanking homology arms. Any impor-
ogy to a genetic locus. Random integrants tant coding region that is essential for a gene’s
will usually contain an intact form of the function can be thereby targeted for deletion.
HSV-tk gene when inserted into the genome. When designing a targeting construct, a few
Cells with random integrants are killed dur- factors should be considered that could result
ing negative selection via treatment with gan- in an incomplete knockout. A gene may be
cylovir or FIAU [1-(2 -deoxy-2 -fluoro-β-D- residually expressed if there exist alternative
arabinofuranosyl)-5-iodouracil], compounds or cryptic promoters that are not disrupted in
that require phosphorylation by HSV-tk to in- the targeted allele (Müller, 1999). Differen-
hibit DNA synthesis. An ∼2000-fold enrich- tial splicing could also generate RNA species
ment of targeted clones can occur with this where the selection marker is skipped. Read-
type of vector. Even with positive and nega- through transcription of the drug-resistance
tive selection, a substantial number of ES cell gene is another way mutant mRNA that has
clones that arise in culture are false positive, some coding sequence from the targeted allele
since HSV-tk can be inactivated before recom- can appear. In gene targeting via a replace-
bination by events like partial deletions. The ment vector, a strong promoter is typically
random integrants are, therefore, insensitive used to drive the neor marker in the stem cells,
to negative drug selection. To avoid subject- but the poly(A)-addition site can sometimes
ing the ES cells to drugs like gancyclovir and be skipped during transcription of the drug-
FIAU, some targeting vectors omit using HSV- resistance gene. Therefore, as the neor gene is
tk in favor of a negative-selection marker like transcribed, downstream exons could possibly Whole Organism
the diphtheria toxin gene (Yagi et al., 1990). be spliced into the mRNA. The neor gene is and Tissue
In a gene-targeting experiment, a replacement often oriented in the opposite direction with Analysis

19.12.5
Current Protocols in Cell Biology Supplement 44
 respect to gene transcription for the targeted Subtle mutations can also be introduced
allele, to avoid any potential transcription of with vectors that take advantage of Cre/loxP
downstream exons. In addition, this orienta- technology (Fig. 19.12.3; Ferradini et al.,
tion also ensures that the strong promoter on 1996; Cohen-Tannoudji and Babinet, 1998).
the neor gene does not influence any down- Similar to a replacement vector, the target-
stream genes. Therefore, the size and location ing construct is designed with a drug-selection
of the targeted deletion should be carefully marker flanked by two homology arms. Both
considered when designing a replacement vec- positive and negative selection are required to
tor, to help avoid an incomplete knockout of isolate cells with the properly recombined mu-
the gene function. tation. Instead of replacing an entire exon with
Double replacement vectors are a variation a drug-selection marker, the goal, here, is to ex-
of the knockout vector design that is primarily change normal coding sequence in a targeted
used to target subtle mutations into a selected allele for a mutated version. One homology
genetic allele (Askew et al., 1993; Stacey et al., arm will carry the planned point mutation,
1994). Also known as tag and exchange, this micro-deletion, or insertion to be introduced
strategy requires two rounds of homologous into the targeted gene. However, in this strat-
recombination to create the desired mutation. egy, the positive drug-selection marker needs
In contrast to a replacement vector, the HSV- to be removed because it will interfere with
tk gene is placed adjacent to neor , rather than transcription of the mutated allele. Using a
at the end of the construct. Both the positive neor gene that is floxed allows a way to even-
and negative selection markers are inserted tually remove the drug-selection marker by us-
into the targeted genomic site, with the ho- ing Cre recombinase. In the construct design,
mology arms flanking both genes. In the first the neor gene is inserted into a noncoding re-
tag step, a targeted stem cell clone is isolated gion. Upon recombination, only a 34-bp loxP
with only neomycin treatment. Once isolated, site remains in the intron, creating a clean mu-
the ES cells then undergo a second round of tation (without a drug-selection marker) in the
electroporation with a targeting vector that targeted genetic locus. Without disruption by
will exchange both drug-selection markers for a drug-selection marker, the modified allele
a mutated exon. This second targeting vec- should be transcribed to produce the expected
tor is basically a genomic clone containing mutant protein, and the remaining loxP site
the desired point mutation, deletion, or inser- within an intron should not interfere with gene
tion. During this second round of gene tar- expression. Like double replacement vectors,
geting, gancyclovir is added to isolate cells this strategy is also a two-step process, except
that have lost the HSV-tk gene from homol- that Cre is used to remove the drug-selection
ogous recombination with the second vec- marker rather than a second round of homolo-
tor. The HPRT gene can be used to replace gous recombination.
both the neor and HSV-tk genes in the con- Making a targeting construct for a con-
struct design. The HPRT gene is part of a ditional knockout is essentially like mak-
purine salvage pathway that can be adapted ing a clean mutation, but, with this strategy,
as a drug marker for both positive and nega- the inserted mutation results in the addition
tive selection (Melton, 2002). HAT (hypoxan- of loxP sites into a targeted genetic locus
thine/aminopterin/thymidine) medium is used (Fig. 19.12.4). In the final targeted allele, two
for positive selection, since it forces a cell to loxP sites will flank an essential exon (or ex-
depend on the salvage pathway that needs an ons) to be excised by Cre in a conditional man-
intact copy of the HPRT gene. Negative selec- ner (Gu et al., 1994; Cohen-Tannoudji and
tion involves the purine analog 6-thioguanine, Babinet, 1998). As with the clean mutation,
which is phosphorylated by the HPRT gene a floxed neor gene is needed for positive se-
to generate a toxic compound for the cell. ES lection and will eventually be removed using
cells with a mutated copy of their endogenous Cre. This process of Cre excision will leave
HPRT gene, however, are needed for this type one of the necessary loxP sites required to flox
of strategy. So, with a double replacement vec- a targeted exon(s). One of the homology arms
tor, two rounds of targeted recombinations are from the replacement vector is then used to
required to obtain the final ES cell clones with insert the other flanking loxP on the opposite
the expected subtle mutation. Several muta- side of this exon(s). Both loxP sites must be
tions can readily be placed into the modified positioned in the introns of the targeted gene,
Generation of gene once targeted ES cell clones from the first and should not interfere with the either the
Gene Knockout electroporation are isolated. coding sequences or the promoter of the gene,
Mice

19.12.6
Supplement 44 Current Protocols in Cell Biology
 neo r HSV-tk
targeting vector
(* subtle variation)

target locus
in chromosome

neo r
chromosome with
targeted insertion

targeted locus after

Cre-mediated excision

Figure 19.12.3 Gene targeting to insert a subtle mutation. In the targeting vector, a subtle mu-
tation (depicted by the asterisk) is designed into one of the homology arms. Subtle mutations can
consist of point mutations, microdeletions, or insertions. When sequence in the target gene is
exchanged with homologous sequence from the targeting vector, the subtle mutation is introduced
into the chromosome. Homologous recombination occurs through both the long and short homol-
ogy arms. A positive drug-selection marker (neor ) is needed in this strategy to select for clones that
have undergone recombination. With a clean mutation, the neor gene is inserted into the targeted
locus, but no exons are lost as a result of recombination. The neor gene needs to be removed
so that it does not interfere with transcription of the recombined allele. The loxP sites (depicted
as black triangles) are used to flank the neor gene in order to facilitate its removal. These 34-bp
loxP sites are recognized by the Cre recombinase, which can be introduced into targeted stem
cells by transient transfection. If the loxP sites are in the same direction, the Cre recombinase will
circularize out any intervening sequence. With Cre-mediated recombination, only a single loxP
site will remain. The loxP site in the recombined allele is situated within intron sequence so that it
does not interfere with transcription of the mutant protein.

nor with any neighboring genes. After homolo- low for conditional disruption of the gene in
gous recombination, the mutated genetic locus the tissue of interest.
will initially have three loxP sites consisting of Knock-in strategies basically adapt replace-
the floxed exon(s) followed by a floxed neor ment vectors to usurp the endogenous pro-
gene. Once the targeted clones are identified, moter of a gene to transcribe an exogenous
the targeted ES cells are then transiently trans- cDNA of interest (Fig. 19.12.5; LeMouel-
fected with a Cre expression vector to remove lic et al., 1990; Hanks et al., 1995; Cohen-
the neor gene. Three scenarios arise when Cre Tannoudji and Babinet, 1998). During recom-
is used to recombine the targeted allele. In bination, the protein start site of the targeted
one situation, the targeted allele is properly re- gene is disrupted when a selected cDNA is
combined and Cre only excises the neor gene. knocked into the promoter, essentially result-
However, recombination can also result in the ing in a targeted transgenic mouse. For this
exclusion of the floxed exon(s) or both the vector design, both the cDNA to be knocked in
exon(s) and the neor gene. A second round of and the neor gene are brought into the genome
genomic screening by Southern blot or PCR is by two flanking homology arms. The 5 arm,
needed to find the ES cell clone that is miss- in this case, will typically contain homologous
ing the neor gene, but which has the floxed genomic sequence that is upstream from the
exon intact. Transcription of the targeted gene protein start site. Similar to a transgenic con-
should not be impacted when the final floxed struct, this regulatory sequence is followed by
allele is generated. The mutated mice are then a cDNA and a poly(A)-addition signal. For Whole Organism
bred to a Cre-expressing transgenic line to al- positive selection, a drug-resistance gene like and Tissue
Analysis

19.12.7
Current Protocols in Cell Biology Supplement 44
 neo r HSV-tk
targeting
vector

target locus
in chromosome

neo r
chromosome with
targeted insertion

neo r

properly recombined chromosome

Figure 19.12.4 Gene targeting to create a conditional knockout. In this strategy, the targeting
construct is designed so that a loxP site (indicated with a black triangle) is located next to an
essential exon within the homology arm. A floxed neor gene is then positioned on the opposite
side of this exon. Both the loxP site and the floxed neor gene are introduced into the target locus
through homologous recombination. Unlike a conventional knockout experiment, the targeting
vector is assembled so that no exons are lost as a result of homologous recombination. After
the stem-cell clone with the properly targeted insertion is identified, the Cre recombinase is then
introduced into the cells through a transient transfection. The Cre recombinase can produce three
types of recombinations. In the first example (shown as a circled “1”), both Exon 2 and the neor
gene are excised from the chromosome, leaving only one loxP site. In the second scenario, (shown
as “2”), only Exon 2 is removed while the floxed neor gene remains in the targeted chromosome.
In the third case (”3”), just the neor gene is excised while two loxP sites remain to flank Exon 2,
creating a floxed allele. Only the stem-cell clones with this specific recombination are injected into
blastocysts to generate floxed mice. Conditional deletion of Exon 2 in vivo is then accomplished
typically by breeding the floxed mice with a Cre-expressing transgenic line. After the removal of
neor in tissue culture and subsequent conditional deletion of Exon 2 in vivo, the final allele has
only one loxP site remaining within the intron sequence of the targeted gene (as depicted in “1”).

neor is placed downstream of the cDNA. A in not only to produce a null mouse, but also to
second 3 homology arm then follows after help track the targeted gene’s expression pat-
the neor gene. Like a replacement vector, two tern through the mutated gene. Even though
homologous recombination events function to knock-in mice are more time consuming to
replace the coding sequence of the gene with generate than transgenic mice, the knocking
the cDNA and neor drug marker in the con- in of a cDNA into a targeted genetic locus
struct. If positioned correctly, the knocked-in can be useful to overcome problems of un-
cDNA should then be in frame and expressed reliable transgene expression that sometimes
by the recombined gene. Through this de- occur from the randomness of the site of in-
sign, a gene can be knocked out as the cDNA tegration and the variability in transgene copy
of closely related isoform is simultaneously number (Jasin et al., 1996; Ledermann, 2000).
knocked in. Protein domains can be switched If the knocked-in cDNA is correctly in frame,
Generation of
Gene Knockout by homologous recombination as well. In ad- expression of the knocked-in transgene should
Mice dition, marker genes, like lacZ, can be knocked more faithfully mirror the transcription pattern

19.12.8
Supplement 44 Current Protocols in Cell Biology
 neor HSV-tk
targeting
vector

target locus
in chromosome

neor
chromosome with
targeted insertion

Figure 19.12.5 Knock-in of a cDNA through the use of gene targeting. Knock-in constructs are
similar in design to conventional knockout gene-targeting vectors, except that additional sequence
(i.e., a cDNA or protein domain of interest) is inserted into the target gene. Through homologous
recombination, this foreign sequence is introduced in frame into the target locus to be expressed
by its promoter. Essential coding sequence in the target locus is simultaneously lost during recom-
bination with the targeting construct. In this example, homologous recombination places a cDNA
under the control of the promoter of the target gene. Concurrently, the translational start sequence
in Exon 1 of the target gene is also replaced by the cDNA. The cDNA, in this example, has a
poly(A)-addition signal (pA) which will stop any further transcription downstream of the targeted
insertion. For a knock-in targeting vector, one of the homology arms must consist of genomic
sequence upstream of the planned insertion site for the cDNA. To knock in a cDNA, as shown, a
targeting vector must use promoter sequence for one of its homology arms (as depicted with the
directional arrow). A positive drug-selection marker (neor ) is still needed to select for clones that
have inserted the designated cDNA into the target gene. Two homologous recombination events
serve to insert the cDNA and neor gene into the target location while knocking out Exon 1.

for the targeted promoter. In one such exam- Another less commonly used gene-
ple, Cre was targeted to the Cd19 genetic lo- targeting strategy employs insertion vectors to
cus in order to maintain all of the regulatory disrupt a genetic locus (Fig. 19.12.6). These in-
elements for transcription in B cells (Rickert sertion vectors are designed using just one arm
et al., 1997). of homologous sequence, and a single recom-
Lastly, gene targeting with a replacement bination event is all that is required to insert
vector can be applied to generate a single-copy a drug-selection gene like neor into the tar-
transgenic mouse. The use of HPRT-null ES geted gene (Hasty et al., 1991). A restriction-
cells allows for only a single copy of a trans- enzyme site located within the homology arm
gene to be integrated at a known insertion site is used to linearize the construct, so that ho-
(Bronson et al., 1996). With this replacement- mologous recombination can occur. Instead of
vector design, homologous recombination is replacing sequence, the whole insertion vec-
needed to correct the HPRT gene. The homol- tor is integrated into the genome. In this case,
ogy arms of the targeting construct restore two the homology arm serves merely to direct the
missing exons needed for a functional copy site of integration to the targeted gene. Inser-
of the HPRT gene during recombination in tion vectors result in gene duplication during
the stem cells. During homologous recombi- homologous recombination because the entire
nation, a single copy of a transgene is car- targeting construct is inserted where the ho-
ried into the restored HPRT genetic locus. The mology arm was linearized. In most cases, the
transgene flanked by the targeting arms has its neor gene will interfere with transcription in
own promoter, selected cDNA, and poly(A)- the targeted gene to create a loss-of-function
addition site. With a reconstituted HPRT gene mutation. However, since homologous recom-
available as a drug-selection marker, clones bination with an insertion vector produces Whole Organism
with the proper homologous recombination gene duplication, mutant RNA species may and Tissue
can then be isolated with HAT medium. still be expressed through altered splicing. Analysis

19.12.9
Current Protocols in Cell Biology Supplement 44
 targeting vector

neo r

Exon 2

target locus
in chromosome

neo r
chromosome with
targeted insertion

Figure 19.12.6 Gene inactivation with an insertion vector. For an insertion vector, only one
homologous recombination event is needed for targeted insertion of DNA into a designated gene.
In this case, recombination occurs around a double-strand break that is located in the homology
arm of the targeting construct. The entire insertion vector is then incorporated into the gene,
including the plasmid backbone (represented by the thin line). A drug-selection marker like the
neor gene is still needed for positive selection, but this marker can be positioned either in the
targeting arm or in the plasmid backbone of the insertion vector. In this example, the positive drug-
selection marker is designed into the homology arm in order to replace essential coding sequence
of the target gene (as shown with the disruption of Exon 2 by the neor gene). For insertion vectors,
a knockout allele is essentially generated because the target gene is disrupted with insertion of
the neor gene and by duplication of exonic sequence.

In addition, intrachromosomal recombination STRATEGIC PLANNING FOR

can also lead to regeneration of the wild- DESIGNING A KNOCKOUT
type allele. Compared with replacement vec- TARGETING CONSTRUCT
tors, though, constructs of this type have While creating a knockout mouse is costly
a higher frequency of recombination (Hasty and labor intensive (>$12,000 and ∼1 year
et al., 1991). A variation of the insertion-vector from electroporation of the targeting vector
strategy is to create a subtle mutation through a in ES cells to the generation of homozygous
hit-and-run or in-out strategy (Vanlancius and null mice; http://www.med.umich.edu/tamc),
Smithies, 1991). With this technique, the ho- the understanding of gene functions in mam-
mology arm contains a desired mutation to be malian physiology and development have
inserted into the targeted gene. Through ho- made the efforts worthwhile. To design a vec-
mologous recombination, both the neor and tor for gene targeting, the following steps are
HSV-tk genes are inserted into the targeted suggested:
gene. After initial positive selection with neor , 1. After selecting a gene to be disrupted,
the isolated cells then undergo a second round the genomic structure of the target allele
of drug selection with gancyclovir. During this should be researched. Exon/intron sequence
next step, the duplicated gene should align information should be gathered. The size of
in such a way as to undergo intrachromoso- the gene and the chromosomal location of the
mal recombination. Both drug-selection mark- allele to be targeted should also be known.
ers will then be lost in the process. The re- Lastly, the location of most major restriction
maining recombined allele will retain the sub- enzyme sites should be identified to aid
Generation of tle mutation, and the mutated protein will be with subcloning. A good restriction-enzyme
Gene Knockout expressed. map of the genomic locus will be useful in
Mice

19.12.10
Supplement 44 Current Protocols in Cell Biology
the construction of the replacement vector. is available. A high-fidelity Taq polymerase
Once an allele is selected for targeted dele- is needed that can amplify long stretches of
tion, the flanking genomic sequence should genomic DNA. Even with a high-fidelity Taq
be examined to ensure that any possible polymerase, however, the PCR products must
neighboring genes are not disrupted during be carefully sequenced to ensure that no cod-
recombination (Olson et al., 1996). Check ing errors were generated during the reaction.
for potential open reading frames that could Lastly, a mouse 129/Sv Lambda phage ge-
be disturbed when a targeted mutation is nomic library is another option to isolate the
generated. Most genomic sequence infor- sequence for the gene to be targeted. A probe
mation can be gathered through the Mouse containing some homology to the chosen al-
Genome Sequencing Consortium (MGSC). lele is used to screen the phage library. Once a
The mouse genome sequence is freely avail- genomic clone is isolated, most of the common
able in public databases (GenBank accession restriction enzyme sites should be verified.
number CAAA01000000) and is accessible The total amount of homology needed in
through various genome browsers (http:// most successful gene-targeting experiments is
www.ensembl.org/Mus musculus/; http:// about 5 to 10 kb in length. The efficiency of
genome.ucsc.edu/, and http://www.ncbi.nlm. homologous recombination increases when in-
nih.gov/genome/guide/mouse/; Mouse creasing the length of homology in the target-
Genome Sequencing Consortium, 2002). ing construct, but this peaks after 14 kb of
Lastly, perform a thorough search to ensure sequence (Deng and Capecchi, 1992). For a
that your gene of interest has not been typical replacement vector, the total length of
knocked out already. About 700 knockouts homologous sequence will consist of roughly
have been made three or more times (Grimm, a 1- to 2-kb span of DNA for a short homol-
2006). ogy arm and a separate, larger 4- to 6-kb ge-
2. A mouse genomic fragment containing a nomic fragment as a long homology arm. The
large portion of the gene to be targeted needs short arm can be as small as 0.5 kb, however,
to be isolated to create the homology arms without affecting targeting efficiency (Hasty
in the targeting construct. A 129/Sv genomic et al., 1991). In the case of the long arm,
clone is most commonly used for constructing though, increasing the length of homology
targeting vectors, since most stem cells were from 8 up to 110 kb does not enhance the
derived from this mouse strain. Using a ge- frequency of homologous recombination (Lu
nomic clone of a mouse strain different from et al., 2003). In the construct, the short and
the ES cell strain can reduce the frequency long arms will share homology with genomic
of homologous recombination. In general, ho- sequences that flank the designated coding re-
mologous recombination is four to five times gion to be deleted from the gene to be targeted.
more efficient with targeting vectors using iso- 3. Both positive and negative selection
genic DNA (Deng and Capecchi, 1992). For genes need be chosen to construct the targeting
any given genetic locus, however, the rate of vector. The most common positive-selection
recombination will depend on the amount of marker is the neomycin phosphotransferase
nonhomologous sequence between the mouse (neor ) gene. The drug-resistance genes for
strains, as a 20-fold decreased efficiency has puromycin and hygromycin have also been
been reported with the use of nonisogenic used as positive-selection markers. When de-
DNA in the targeting vector (te Riele et al., signing a targeting vector, the sequence of the
1992). Searching a BAC (Bacterial Artificial neor gene is often inserted in the opposite ori-
Chromosome) library is one means to get the entation to transcription in the target allele.
genomic sequence needed to make a target- With negative selection, the HSV thymidine
ing vector. Due to sequencing of the mouse kinase (HSV-tk) gene is often used to select
genome, BLAST can be used to readily find against random integrants. If the targeting con-
a BAC clone that contains the selected gene struct is randomly integrated, the cells will
for targeted disruption. Genomic DNA from a mostly contain an intact copy of the HSV-tk
female 129S6/SvEvTac (Taconic) mouse, for gene that will phosphorylate gancyclovir into
example, was used to create the RPCI-23 BAC a cytotoxic drug (Mansour et al., 1988). The
library. A BAC clone will generally contain diphtheria toxin gene has also been used for
from 100 to 300 kb of genomic sequence. negative selection against random integrants
Another means to obtain the homology arms (Yagi et al., 1990). A frequently used plas-
for the targeting vector is through PCR. PCR mid for making replacement vectors is pPNT Whole Organism
can be used to amplify the homology arms, (Tybulewicz et al., 1991). This plasmid con- and Tissue
Analysis
if a source of purified 129/Sv genomic DNA tains both the neor and HSV-tk genes. All that
19.12.11
Current Protocols in Cell Biology Supplement 44
 is required is to ligate the two homology arms genes are affected during homologous recom-
into the unique restriction enzyme sites that bination.
flank the neor gene. 2. After designating a location for targeted
With the acquisition of the drug markers deletion, identify neighboring restriction-
and a genomic clone, one can proceed with enzyme sites that are convenient for generating
piecing a targeting construct together. a short homology arm (∼1 to 2 kb in length) as
well as the long homology arm (∼4 to 6 kb).
DESIGNING A KNOCKOUT A total amount of 5 to 8 kb of homology is
TARGETING CONSTRUCT typical in most targeting vectors (Hasty et al.,
(REPLACEMENT VECTOR) 1993). Both the short and long homology arms
1. Once a gene is chosen for targeted inac- may need to be subcloned into a vector such a
tivation, determine the best region of genomic pBluescript (Stratagene) in order to facilitate
sequence to disrupt through homologous re- the assembly of the final targeting construct.
combination. With a replacement vector, two Often, DNA-modifying enzymes are needed
homologous recombination events will occur with subcloning, since convenient restriction
at the targeted locus via the long and short enzyme sites may not be readily available in
homology arms. The positive drug-resistance the genomic clone. The following DNA modi-
gene is inserted into a gene while coding se- fications may need to be applied for subcloning
quence is simultaneously deleted. The length the homology arms:
of genomic sequence replaced by the drug- a. Incompatible overhangs produced from
selection marker is governed by the location a restriction enzyme digest can be blunted to
of these homology arms. So, any intervening help facilitate the subcloning of a homology
sequence situated between the two directed arm. Blunt ligatable ends can be generated
sites of recombination is omitted during the with enzymes like mung bean nuclease (to re-
process of homologous recombination. The move a 3 overhang) or Klenow polymerase
length of sequence that can be replaced by (to fill in a 3 recessed end). However, sticky
the positive-selection marker does not seem to end modification can result in reduced ligation
influence the efficiency of homologous recom- efficiency.
bination. Up to 15 kb of sequence, for exam- b. Adaptors can be purchased (from New
ple, was deleted with a replacement vector to England Biolabs, Stratagene, etc.) to in-
generate the T cell–receptor knockout mouse sert a desired restriction enzyme site onto
(Mombaerts et al., 1991). With most targeting blunt-ended DNA. Alternatively, oligonu-
constructs, however, one to a few critical ex- cleotide adaptors can be generated to insert a
ons are generally disrupted in gene-targeting restriction-enzyme site into cleaved DNA with
experiments. Small deletions of 0.5 to 1 kb in sticky ends. Oligos must be designed so that,
length are typical for most gene-targeting ex- upon annealing, sticky ends are generated that
periments, and are usually sufficient to inacti- are compatible to the cleaved DNA. The syn-
vate a gene (DeChiara, 2001). Genetic coding thesized oligos are diluted to a concentration
sequence essential for protein function should of 100 ng/μl in a buffer of 10 mM Tris·Cl/50
be selected for targeted deletion because there mM NaCl, pH 8.0. This solution is then heated
is the possibility that the inserted drug marker at 95◦ C for 15 min and cooled to room temper-
can be bypassed during transcription of the tar- ature. The annealed oligos are ligated with the
geted allele. Any resulting truncated protein cleaved DNA at a 500 to 1000 molar excess.
generated should, therefore, not be functional This high concentration helps to facilitate lig-
and would hopefully be rapidly degraded so as ation of the oligo and inhibits mere re-ligation
not to interfere with the cell’s normal physi- of the vector.
ology. Generally, deleting 5 exons provides a c. For ligation, the vector may need to be
better chance to totally disrupt the formation of modified to prevent re-ligation, particularly
protein than a strategy that targets downstream if only one restriction enzyme is being used
exons at the C-terminus (Hasty et al., 1993). to cleave the DNA. In this case, the vector
With genes smaller than 20 kb in length, some needs to be dephosphorylated to prevent re-
investigators prefer to design replacement vec- ligation. Dephosphorylation of the DNA can
tors that disrupt all the coding exons in order be achieved by incubating the cleaved vector
to absolutely ensure gene inactivation (Cheah with an alkaline phosphatase (typically either
and Behringer, 2001). For any gene-targeting calf intestinal phosphatase or shrimp alkaline
Generation of strategy, though, be sure that no neighboring phosphatase).
Gene Knockout
Mice

19.12.12
Supplement 44 Current Protocols in Cell Biology
 As mentioned above, PCR provides another cutter, this site is useful for linearizing tar-
means to clone out the homology arms, but geting constructs. An example of a targeting
careful attention should be made to ensure that construct is illustrated in Figure 19.12.1.
no coding errors were generated during the 4. While the replacement vector is being
reaction. constructed, a strategy to identify the tar-
3. The targeting construct can be assem- geted clones can be planned using all the
bled after both homology arms and the drug- collected sequence information about the ge-
selection markers have been gathered. The nomic clone. Typically, to be certain of proper
targeting construct should be ligated so that recombination, a Southern blot test should be
the long and short homology arms flank designed to identify the targeted stem cell
the positive-selection marker. A negative- clone. With a Southern blot, an appreciable
selection marker is subcloned outside of this fragment-size difference must appear between
grouping. Try to ligate the vector so that, upon the wild-type and the mutated allele. An ideal
recombination, the positive-selection marker restriction enzyme should be identified that
is transcribed in the opposite orientation with will best show a band difference from the di-
respect to the targeted gene. A unique re- gested genomic DNA. Southern blot analysis
striction enzyme site must be maintained to usually tests either the insertion or deletion of a
linearize the targeting construct to improve restriction enzyme site due to homologous re-
the efficiency of homologous recombination combination. A fragment of the genomic clone
(Hasty et al., 1992). This restriction enzyme is generally retained to make a Southern probe
site should be located outside the regions of ho- that contains sequence flanking the homology
mology, typically between the plasmid back- arms. This ensures that the Southern probe will
bone and a targeting arm. With linearization of not bind with random integrants. Therefore,
the targeting construct, the plasmid DNA will the only bands to be detected with the South-
protect the negative-selection marker (e.g., ern probe will be from either the recombined
HSV-tk gene) from nucleases (Cheah and allele or the wild-type gene. PCR can also be
Behringer, 2001). The placement of either the used to find properly targeted clones. PCR can
long or the short arm between the two drug provide a fast, high-throughput means to test
markers has varied among targeting vectors. for targeted clones. With PCR, one primer will
While some investigators prefer to locate the typically be designed within the neor gene and
HSV-tk adjacent to the long targeting arm the other primer will be located outside the
(Torres and Kühn, 1997), other vectors place short arm of the replacement vector. Screen-
the HSV-tk gene at the end of the short target- ing by PCR, however, can be tricky since long
ing arm because this design may help improve PCR products need to be amplified. In addi-
the efficiency of negative selection (Melton, tion, there is always the risk of either false
2002). positives or negatives when using PCR to find
Plasmids such as pPNT (Tybulewicz et al., a clone. Once a positive clone is identified,
1991) or the pKO Scrambler Series (Lexicon though, PCR becomes the preferred strategy
Genetics; http://www.lexgen.com) can be use- to genotype the resulting mutant mice.
ful in designing targeting constructs, since
these vectors contain both neor and thymidine
kinase genes. In addition, common restric- DESIGNING A KNOCK-IN
tion enzyme sites are positioned in locations TARGETING CONSTRUCT
that facilitate the subcloning of the homol- Designing a knock-in construct follows
ogy arms. With pPNT, for example, one ho- the same basic rules as a knockout replace-
mology arm can be subcloned into the re- ment vector. One homology arm, however,
striction enzyme sites (XbaI, BamHI, KpnI, must consist of genomic sequence upstream
and EcoRI) located between the neor and of the protein initiation site. In this case,
HSV-tk genes. Both of these drug-selection the 5 and 3 homology arms will flank both
marker genes in this vector are driven by the the knocked in cDNA and a positive drug-
3-phosphoglycerate kinase promoter (PGK; selection marker. Be sure to include a pA-
McBurney et al., 1991), with the PGK pA- addition signal for the knocked in cDNA.
addition site. PGK is a housekeeping enzyme Upon recombination, the promoter in the tar-
and the promoter is useful to drive high ex- geted gene will drive expression of the inserted
pression of these drug markers. The second cDNA (Fig. 19.12.5). For efficient homolo-
homology arm can then be placed adjacent to gous recombination, the length of nonhomol- Whole Organism
the neor gene with NotI and XhoI restriction ogous DNA (i.e., the cDNA and neor gene) and Tissue
Analysis
enzyme sites. Since NotI is a rare 8-base-pair inserted into a target locus should not exceed
19.12.13
Current Protocols in Cell Biology Supplement 44
 the length of total homology (the long and of genomic DNA needed for a targeting vec-
short arms of the targeting construct). tor follows the same rules listed for a con-
ventional knockout vector, with roughly 5 to
DESIGNING A CONDITIONAL 10 kb of homologous sequence required for
KNOCKOUT TARGETING efficient recombination. Exon/intron bound-
CONSTRUCT aries and restriction-enzyme sites need to be
With a conventional knockout vector, an mapped. Afterwards, the coding region to be
essential coding region in the targeted gene floxed should be determined. The final loxP
is replaced with a drug-selection marker dur- sites should be located within intronic se-
ing homologous recombination. For a condi- quence so that gene expression is not affected.
tional knockout mouse, however, the final tar- In addition, a loxP site should not be designed
geted allele needs to be functionally intact. into any promoter regulatory sequence, as this
In this case, the end result of gene targeting may interfere with transcription. In the floxed
is the placement of loxP sites around an es- allele, all the coding sequence should be in
sential coding region to create a floxed allele. frame. Again, check for any neighboring open
Therefore, no sequence should be omitted by reading frames that could be disrupted after
gene targeting. Basically, a continuous length Cre recombination.
of homologous genomic sequence is ideal in 2. A plasmid with a floxed neor gene should
making a vector for a conditional knockout. A be acquired to create the targeting construct. A
fragment of the genomic clone is merely cut in negative-selection marker should also be ob-
two with a restriction enzyme to derive a long tained, such as the HSV-tk gene.
and short homology arm. This design contrasts 3. Restriction enzyme sites in the genomic
with a conventional knockout, where two sepa- clone should be identified to aid in the place-
rate lengths of homologous genomic sequence ment of the loxP sites. It is best if two unique
are needed to make the targeting vector. One restriction enzyme sites can be found that
of the homology arms is then modified to po- flank a vital coding region for placement for
sition a loxP site next to an essential exon for the floxed neor gene, as well as a loxP site.
conditional deletion. As mentioned, the loxP sites should only be
While a positive drug-selection marker placed in intronic sequences. If available, the
(i.e., neor gene) is needed for initial enrich- floxed neor gene can be subcloned into one
ment of targeted clones, it must be floxed so of the unique restriction enzyme sites. In the
that it will not interfere with the final mu- second unique restriction enzyme site, a syn-
tated gene. Once the targeted clones are iden- thetic loxP site can then be ligated into the
tified, this drug marker is removed by transient vector. Oligonucleotides with the loxP site and
transfection to express Cre recombinase. If the restriction-enzyme site overhangs can be syn-
correct recombination event occurs, then only thesized, annealed, and ligated into the vec-
a loxP site will remain after Cre excises the tor. The annealed synthetic oligo should be as
neor gene. In combination with the loxP site shown in Figure 19.12.7.
in the targeting arm, the exon(s) designated for A loxP site consists of two 13-bp inverted
conditional deletion should basically be floxed repeats that flank an asymmetric 8-bp core (un-
upon Cre-mediated rearrangement. However, derlined in Fig. 19.12.7). The asymmetric core
Cre recombination can also result in the exclu- determines the directionality of the loxP site.
sion of the floxed exon(s) or both the exon(s) All loxP sites in the targeting construct must be
and the neor gene, so stem-cell clones must be in the same orientation. Otherwise, if the loxP
screened to determine the correct rearrange- sites are in opposite orientation, the floxed se-
ment (Fig. 19.12.4). quence will be inverted rather than deleted
1. A homologous 129/Sv genomic fragment from the gene. To anneal the oligos, dissolve
continuous with the gene of interest needs the DNA to form a 100 ng/μl solution in
to be isolated and characterized. The length 10 mM Tris·Cl/50 mM NaCl, pH 8.0. Combine

5′-RE Site—ATAACTTCGTATAGCATACATTATACGAAGTTAT—RE Site-3′

3′-RE Site-TATTGAAGCATATCGTATGTAATATGCTTCAATA-RE Site-5′

Generation of
Gene Knockout Figure 19.12.7 Annealed synthetic oligos with loxP site and restriction-enzyme site overhangs.
Mice
See text for additional detail.
19.12.14
Supplement 44 Current Protocols in Cell Biology
sense and antisense strands in a microcen- Capecchi, M.R. 2005. Gene targeting in mice: Func-
trifuge tube, heat at 95◦ C for 15 min, then tional analysis of the mammalian genome for the
let slowly cool to room temperature for proper twenty-first century. Nat. Rev. 6:507-512.
annealing. If needed, the vector pBS246 can Cheah, S.S. and Behringer, R.R. 2001. Contem-
be obtained to procure the loxP sites. This porary gene targeting strategies for the novice.
Mol. Biotechnol. 19:297-304.
plasmid has two loxP sites that flank a mul-
tiple cloning site (Sauer, 1993). Sequence can Cohen-Tannoudji, M. and Babinet, C. 1998. Be-
yond knock-out mice: New perspectives for the
then be ligated around the loxP sites to get the programmed modification of the mammalian
desired floxed allele. genome. Mol. Hum. Reprod. 4:929-938.
4. After proper placement of the floxed neor DeChiara, T.M. 2001. Gene targeting in ES cells.
gene and a loxP site within the homologous ge- Methods Mol. Biol. 158:19-45.
nomic fragment, this cluster should then be lig- Deng, C. and Capecchi, M.R. 1992. Reexamination
ated together with a negative-selection marker of gene targeting frequency as a function of the
like HSV-tk. Again, a unique restriction en- extent of homology between the targeting vector
zyme site must be located adjacent to the ho- and the target locus. Mol. Cell Biol. 12:3365-
mology arm so that the targeting vector can be 3371.
linearized. Doetschman, T., Gregg, R.G., Maeda, N., Hooper,
5. Eventually, after electroporation and iso- M.L., Melton, D.W., Thompson, S., and
Smithies, O. 1987. Targeted correction of a mu-
lation of targeted clones, the floxed neor gene tant HPRT gene in mouse embryonic stem cells.
should be removed by transient expression of Nature 330:576-578.
Cre. Therefore, a Cre expression vector must Dymecki, S.M. 1996. Flp recombinase promotes
be acquired for generation of the final mutated site-specific DNA recombination in embryonic
allele. One such plasmid is pBS185 (Sauer stem cells and transgenic mice. Proc. Natl. Acad.
and Henderson, 1990). This vector will ex- Sci. U.S.A. 93:6191-6196.
press Cre under the CMV promoter. Evans, M.J. and Kaufman, M.H. 1981. Establish-
6. A PCR protocol should be designed to ment in culture of pluripotential cells from
identify the floxed allele. With the sequence mouse embryos. Nature 292:154-156.
information from the genomic clone, primers Feil, R., Wagner, J., Metzger, D., and Cham-
can be designed to detect wild-type, floxed, bon, P. 1997. Regulation of Cre recombinase
activity by mutated estrogen receptor ligand-
and Cre-recombined alleles. If possible, cell binding domains. Biochem. Biophys. Res. Com-
culture work should be performed to test Cre mun. 237:752-757.
excision and PCR analysis before the actual Ferradini, L., Gu, H., De Smet, A., Rajewsky,
electroporation of the targeting construct into K., Reynaud, C.A., and Weill, J.C. 1996.
the stem cells. Rearrangement-enhancing element upstream of
the mouse immunoglobulin kappa chain J clus-
ter. Science 271:1416-1420.
LITERATURE CITED
Anton, M. and Graham, F.L. 1995. Site-specific Grimm, D. 2006. A mouse for every gene. Science
recombination mediated by an adenovirus vec- 312:1862-1866.
tor expressing the Cre recombinase protein: A Gu, H., Marth, J.D., Orban, P.C., Mossman, H., and
molecular switch for control of gene expression. Rajewsky, K. 1994. Deletion of a DNA poly-
J. Virol. 69:4600-4606. merase β gene segment in T cells using cell
Askew, G.R., Doetschman, T., and Lingrel, J.B. type-specific gene targeting. Science 265:103-
1993. Site-directed point mutations in em- 106.
bryonic stem cells: A gene-targeting tag-and- Hanks, M., Wurst, W., Anson-Cartwright, L., Auer-
exchange strategy. Mol. Cell. Biol. 13:4115- bach, A.B., and Joyner, A.L. 1995. Rescue of the
4124. En-1 mutant phenotype by replacement of En-1
Bradley, A., Evans, M., Kaufman, M.H., and with En-2. Science 269:679-682.
Robertson, E. 1984. Formation of germ-line chi- Hasty, P., Rivera-Pérez, J., and Bradley, A. 1991.
maeras from embryo-derived teratocarcinoma The length of homology required for gene tar-
cell lines. Nature 309:255-256 geting in embryonic stem cells. Mol. Cell. Biol.
Bronson, S.K., Plaehn, E.G., Kluckman, K.D., 11:5586-5591.
Hagaman, J.R., Maeda, N., and Smithies, O. Hasty, P., Rivera-Pérez, J., and Bradley, A. 1992.
1996. Single-copy transgenic mice with chosen- The role and fate of DNA ends for homol-
site integration. Proc. Natl. Acad. Sci. U.S.A. ogous recombination in embryonic stem cell.
93:9067-9072. Mol. Cell. Biol. 12:2464-2474.
Capecchi, M.R. 1989. The new mouse genetics: Hasty, P., Abuin, A., and Bradley, A. 1993. Gene
Altering the genome by gene targeting. Trends targeting, principles, and practice in mammalian
Genet. 5:70-76. cells. In Gene Targeting: A Practical Approach Whole Organism
Capecchi, M.R. 1994. Targeted gene replacement. (A.L. Joyner, ed.) pp. 1-35. Oxford University and Tissue
Sci. Am. 270:52-59. Press, New York. Analysis

19.12.15
Current Protocols in Cell Biology Supplement 44
 Hayashi, S. and McMahon, A.P. 2002. Efficient Olson, E.N., Arnold, H.H., Rigby, P.W., and Wold,
recombination in diverse tissues by a tamoxifen- B.J. 1996. Know your neighbors: Three pheno-
inducible form of Cre: A tool for temporally reg- types in null mutants of the myogenic bHLH
ulated gene activation/inactivation in the mouse. gene MRF4 Cell 85:1-4.
Dev. Biol. 244:305-318. Rickert, R.C., Roes, J., and Rajewsky, K. 1997.
Iredale, J.P. 1999. Gene knockouts. Mol. Pathol. B lymphocyte-specific, Cre-mediated mutagen-
52:111-116. esis in mice. Nucleic Acids Res. 25:1317-1318.
Jasin, M., Moynahan, M.E., and Richardson, C. Rijkers, T., Van Den Ouweland, J., Morolli, B.,
1996. Targeted transgenesis. Proc. Natl. Acad. Rolink, A.G., Baarends, W.M., Van Sloun,
Sci. U.S.A. 93:8804-8808. P.P.H., Lohman, P.H.M., and Pastink, A. 1998.
Ledermann, B. 2000. Embryonic stem cells and Targeted inactivation of mouse RAD52 reduces
gene targeting. Exp. Physiol. 85.6:603-613. homologous recombination but not resistance
to ionizing radiation Mol. Cell Biol. 18:6423-
LeMouellic, H., Lallemand, Y., and Brûlet, P. 1990. 6429.
Targeted replacement of the homeobox gene
Hox-3.1 by the Escherichia coli lacZ in mouse Sargent, R.G. and Wilson, J.H. 1998. Recombina-
chimeric embryos, Proc. Natl. Acad. Sci. U.S.A. tion and gene targeting in mammalian cells Curr.
87:4712-4716. Res. Mol. Ther. 1:584-592.
Lin, F.L., Sperle, K., and Sternber, N. 1984. Re- Sauer, B. 1993. Manipulation of transgenes by site-
combination in mouse L cells between DNA specific recombination: Use of Cre recombinase.
introduced into cells and homologous chromo- Meth. Enzymol. 225:890-900.
somal sequences. Proc. Natl. Acad. Sci. U.S.A. Sauer, B. and Henderson, N. 1990. Targeted in-
82:1391-1395. sertion of exogenous DNA into the eukaryotic
Lu, Z.H., Brooks, J.T., Kaufman, R.M., and Ley, genome by the Cre recombinase. New Biol.
T.J. 2003. Long targeting arms do no increase 2:441-449.
the efficiency of homologous recombination in Sauer, B. and Henderson, N. 1998. Site-specific
the β-globin locus of murine embryonic stem DNA recombination in mammalian cells by the
cells Blood B:1531-1533. Cre recombinase of bacteriophage P1. Proc.
Mak, T.W. (ed.) 1998. The Gene Knockout Facts- Natl. Acad. Sci. U.S.A. 85:5166-5170.
book. Academic Press, New York. Schwartzberg, P.L., Goff, S.P., and Robertson, E.J.
Majzoub, J.A. and Muglia, L.J. 1996. Knockout 1989. Germ-line transmission of a c-abl muta-
mice. New Engl. J. Med. 334:904-907. tion produced by targeted gene disruption in ES
cells. Science 246:799-803.
Mansour, S.L., Thomas, K.R., and Capecchi, M.R.
1988. Disruption of the proto-oncogene int-2 Sikorski, R. and Peters, R. 1997. Transgenics on the
in mouse embryo-derived stem cells: A gen- internet. Nat. Biotech. 15:289.
eral strategy for targeting mutations to non- Smithies, O., Gregg, R.G., Boggs, S.S., Koralewski,
selectable genes Nature 336:348-352. M.A., and Kucherlapati, R.S. 1985. Insertion of
Martin, G.R. 1981. Isolation of a pluripotent cell DNA sequences into the human chromosomal
line from early mouse embryos cultured in β-globin locus by homologous recombination.
medium conditioned by teratocarcinoma stem Nature 317:230-234.
cells. Proc. Natl. Acad. Sci. U.S.A. 78:7634- Stacey, A., Schnieke, A., McWhir, J., Cooper, J.,
7638. Colman, A., and Melton, D.W. 1994. Use of
McBurney, M.W., Sutherland, L.C., Adra, C.N., double-replacement gene targeting to replace
Leclair, B., Rudnicki, M.A., and Jardine, K. the murine α-lactalbumin gene with its human
1991. The mouse Pgk-1 gene promoter contains counterpart in embryonic stem cells and mice.
an upstream activator sequence Nucleic Acids Mol. Cell Biol. 14:1009-1016.
Res. 19:5755-5761. Sternberg, N. and Hamilton, D. 1981. Bacterio-
Melton, D.W. 2002. Gene-targeting strategies phage P1 site-specific recombination. J. Mol.
Methods Mol. Biol. 180:151-173. Biol. 150:467-486.

Mombaerts, P., Clarke, A.R., Hooper, M.L., and te Riele, H., Maandag, E.R., and Berns, A. 1992.
Tonegawa, S. 1991. Creation of a large genomic Highly efficient gene targeting in embryonic
deletion at the T-cell antigen receptor β-subunit stem cells through homologous recombination
locus in mouse embryonic stem cells by gene with isogenic DNA constructs. Proc. Natl. Acad.
targeting. Proc. Natl. Acad. Sci. U.S.A. 88:3084- Sci. U.S.A. 89:5128-5132.
3087. Thomas, K.R. and Capecchi, M.R. 1987. Site-
Mortensen, D. 2006. Overview of gene targeting by directed mutagenesis by gene targeting in mouse
homologous recombination Curr. Protoc. Mol. embryo-derived stem cells. Cell 51:503-512.
Biol. 76:23.1.1-23.5.12. Thomas, K.R., Folger, K.R., and Capecchi, M.R.
Mouse Genome Sequencing Consortium. 2002. 1986. High frequency targeting of genes to
Initial sequencing and comparative analysis of specific sites in the mammalian genome. Cell
the mouse genome. Nature 420:520-562. 44:419-428.
Generation of Müller, U. 1999. Ten years of gene targeting: Torres, R.M. and Kühn, R. 1997. Laboratory Pro-
Gene Knockout Targeted mouse mutants, from vector design to tocols for Conditional Gene Targeting. Oxford
Mice University Press, New York.
phenotype analysis Mech. Dev. 82:3-21.
19.12.16
Supplement 44 Current Protocols in Cell Biology
Tybulewicz, V.L.J., Crawford, C.E., Jackson, P.K., Yagi, T., Ikawa, Y., Yoshida, K., Shigetani, Y.,
Bronson, R.T., and Mulligan, R.C. 1991. Neona- Takeda, N., Mabuchi, I., Yamamoto, T., and
tal lethality and lymphopenia in mice with a Aizawa, S. 1990. Homologous recombination
homozygous disruption of the c-able proto- at c-fyn locus of mouse embryonic stem cells
oncogene. Cell 65:1153-1163. with use of diphtheria toxin A–fragment gene in
Utomo, A.R.H., Nikitin, A.Y., and Lee, W.H. negative selection. Proc. Natl. Acad. Sci. U.S.A.
1999. Temporal, spatial, and cell type-specific 87:9918-9992.
control of Cre-mediated DNA recombination Zambrowicz, B.P. and Sands, A.T. 2003. Knock-
in transgenic mice. Nat. Biotechnol. 17:1091- outs model the 100 best-selling drugs—will they
1096. model the next 100? Nat. Rev. 2:38-51.
Vanlancius, V. and Smithies, O. 1991. Testing an
“in-out” targeting procedure for making sub- Key References
tle genomic modifications in mouse embryonic Kmiec, E.B. (ed.) 2000. Gene Targeting Protocols.
stem cells. Mol. Cell Biol. 11:1402-1408. Humana Press, Totowa, N.J.
Gene knockout methodologies discussed in this
Vasquez, K.M., Marburger, K., Intody, Z., and
book include different strategies for gene transfer
Wilson, J.H. 2001. Manipulating the mam-
and techniques for gene targeting.
malian genome by homologous recombination.
Proc. Natl. Acad. Sci. U.S.A. 98:8403-8410.
Internet Resources
Vogel, G. 2007. A knockout award in medicine. http://www.cellmigration.org/resource/komouse/
Science 318:178-179. komouse resources targeting4beginners.shtml
Wong, E.A. and Capecchi, M.R. 1987. Homolo- This Web site was established throuhg the Cell Mi-
gous recombination between coinjected DNA gration Consortium to provide strategies for gene
sequences peaks in early to mid-S phase. Mol. targeting as well as important links useful for ac-
Cell Biol. 7:2294-2295. quiring genomic sequence information.

Whole Organism
and Tissue
Analysis

19.12.17
Current Protocols in Cell Biology Supplement 44