Gene Targeting In ES Cells

Embryonic Stem Cell

ES cells, mouse embryo fibroblast (MEF) feeder cells, tissue culture media, growth factors and selection reagents will be supplied as well as all necessary tissue culture and transfection facilities. Transfection and selection for recombinant clones will be carried out by the investigator. DNA analysis of selected clones will be carried out in the investigator's own lab.
Early passage, germ line proven RI ES cells from (129/Sv x 129/Sv-CP)F1mice, established by Andeas Nagy are available. These cells are competent for aggregation and blastocyst injection.

ES Cell Line Libraries

Having decided to embark on altering a specific part of the genome, it is advisable to check whether a targeted ES clone for your gene of interest already exists, and is available from one of the ES cell consortia.

IGTC 

The International Gene Trap Consortium (IGTC) represents all publicly available gene trap cell lines, which are available on a non-collaborative basis for nominal handling fees. At present, they store more than 120.000 ES cell clones. Therefore, it is likely that your gene might have been hit by one or more of these ES cell clones. You can search and browse the IGTC database for cell lines of interest using accession numbers or IDs, keywords, sequence data, tissue expression profiles and biological pathways. There is also a gene trap tutorial, which is useful to review.

KOMP

The KOMP Repository is the official archive and distribution center for the Knockout Mouse Project (KOMP), a major 5-year trans-NIH initiative designed to generate null alleles in C57BL/6 embryonic stem (ES) cells for most genes not already available as knockout mice.

MMRRC

The Mutant Mouse Regional Resource Centers (MMRRC) 

EUCOMM

European Conditional Mouse Mutagenesis Program (EUCOMM)

Jax MGI

A search in the Jax Mouse Genome Informatics (MGI) database will help you find any existing targeted mutations or transgenes of your gene of interest. In the Gene/Marker box type in the name of the gene, then press Search. Links to all the mutants for this gene will appear on the resulting datasheet under Phenotypes (in the left hand margin).
Our unit has produced several germ line transmitting chimeras using ES cells from such resources.
Note that if the resource does not provide full pathogen testing, you may be required, in addition to the price of the line, to pay for a pathogen screen, which costs approximately $560.

ES Cell Lines

We currently have a number of ES cell lines available for targeting experiments. The choice of cell line will depend primarily on which background you wish to generate your mutation on, and which DNA is available to be used as a basis for building the targeting construct.
Most analyses of mutations in mice are best studied on a defined inbred strain, generally C57BL/6. For technical and historical reasons most knockouts to date have been derived using ES cells from the 129 inbred strain. This is true also of targeted mice generated at the WIS.
However the 129 strain has a number of drawbacks:

  • It has low fertility and is difficult to maintain. It also carries various anatomical and behavioral abnormalities.

  • Backcrossing to C57BL/6 is a lengthy and costly process, and even after 10 generations (2 years) of backcrossing to C57BL, the resulting mice will still retain 129 derived sequences closely linked to the target gene. These 129 sequences have been known to confound phenotype analysis data (Schalkwyk et al).

  • 129 BAC clones are very expensive to obtain relative to C57BL/6 clones, for use in generating a targeting construct. See BACPAC Resources Center (BPRC).

For a comprehensive review on the question of targeting in 129 versus C57BL/6 backgrounds please consult the following reference (Seong et al). It should be noted that there are reports that ES cells derived from the C57BL/6 mouse strain have a higher rate of genomic instability and aneuploidy then 129 derived lines (Hughes et al).

Available 129 strains:

  • R1 - derived from (129X1/SvJ x 129S1)F1 mice from the Nagy lab

  • W4 - derived from the 129S6/SvEvTac from Taconic

Available C57BL/6 strains:

Designing a Construct

Meticulous and detailed design of the targeting construct is fundamental to the success of any targeting experiment. The planning stage is one of the most critical steps in making a targeted mouse. It is essential that every detail in design (restriction sites for linearization, PCR and Southern screening strategies, removal of selection markers etc.) is taken care of in the planning stage before ever touching an Eppendorf tube.
There are many different ways of targeting a particular locus. Targeting can be used to generate knockouts, knockins or conditional alleles. The strategy employed will reflect the goal of the experiment. A detailed survey on different targeting designs is available in Nagy, A et.al. (ed). Manipulating the Mouse Embryo 3rd Edition, 2003, Cold Spring Harbor Laboratory Press. This volume covers most aspects of gene targeting, ES cell production, culture and manipulation, mouse chimeras, mouse cloning, assisted reproduction strategies and a summary of mouse development at the molecular level. It is available from the library, and is recommended reading for anybody planning to carry out a gene targeting project.
Below are several rules, which apply to the construction of most targeting vectors:

1. Isogenic DNA:

The degree of identity of sequence between the targeting vector and the target locus will influence the efficiency of homologous recombination. Polymorphisms between different mouse strains will reduce the targeting frequency, therefore it is important to construct the homology arms of the targeting vector from DNA isogenic to the ES cells used in the experiment.
If you are using PCR to derive your homology arms it is recommended that you use commercially available DNA from the Jackson labs, which matches the ES cell line you will be using.
R1 ES cells are derived from (129X1/SvJ x 129S1)F1 mice (Nagy lab). Genomic DNA from 129X1/SvJ mice (stock # 00691) is available from the Jackson Labs Mouse DNA Resource.
Bruce 4 ES cells are derived from C57BL/6J mice. Genomic DNA from the C57BL/6J mice (stock # 000664) is available from the Jackson Labs Mouse DNA Resource.
If you are constructing the vector using recombineering, the ES cells used in targeting should match the strain from which the BAC clone was derived.
For targeting in the 129 background, the RPCI - 22 (129S6/SvEvTac) Mouse BAC library from BacPAc resources is most widely used. We have available the matching ES line W4, made from the 129S6/SvEvTac strain for targeting.
For targeting in the C57BL/6 background, the RPCI-23 Mouse (C57BL/6J Female) BAC library from BacPAc resource, and the RPCI-24 Mouse (C57BL/6J Male) BAC library from BacPAc resource, are the most widely used. We have available the Bruce 4 ES line from the C57BL/6J strain for targeting.

2. Arms of homology:

These are sequences homologous to the area to be targeted. They are placed in the target vector flanking a positive selection marker gene. In the event of homologous recombination, sequences between the two arms of homology will be replaced by the selection marker. The greater the amount of sequence match, the more likely the targeting is to succeed. The two arms, termed the 5’ and 3’ arms, should have a minimum combined length of 6-8 kb, though longer is preferred. Ideally, the long arm of the vector should be a minimum of 5 kb, although 7-9 kb is preferred.
The short arm should never be less than 1 kb. This will be used for PCR amplification in screening, so it cannot be too long. A short arm of over 2 kb may make PCR screening very difficult.

3. Unique restriction site:

Prior to transfection the vector is linearized outside of the arms of homology, so provision must be made for a unique recognition sequence for a restriction enzyme, at an appropriate place. (See Knock In Construct.) Increasing the length of the homology arms may make it difficult to introduce a unique recognition sequence. The recognition site can be introduced by including its sequence in the 5’ oligo primer for the selection marker amplification.

4. Selection markers:

Positive Selection marker: A positive selection marker (SM) for integration of the construct is included in the vector, usually between the two arms of homology. The intervening genomic sequences will be replaced by the SM in a homologous recombination event. Usually the neomycin resistance gene Neo under the PGK1 promoter, which is active in ES cells, is used. Neo confers resistance to the drug G418, an antibiotic which inhibits protein synthesis. If a second targeting event is required, and neo cannot be used, because the cells have already acquired neo resistance, puromycin or hygromycin can be used. (Note; if using a drug other than neo make sure feeder fibroblasts are resistant to the drug of use. We have available DR4 fibroblasts which are resistant to G418, puromycin, hygromycin and 6-thioguanine for Hprt1 resistance.)
The neo coding sequence and its PGK1 promoter can have unintended consequences on the targeted gene and on adjacent genes if they remain in the targeted locus. The neomycin resistance gene contains cryptic splice acceptors and donors which can be utilized by transcripts from the targeted gene. In addition, the function of genes adjacent to the targeted gene can be altered by the neomycin gene inserted at the targeted locus. The selection cassette is generally put in the opposite orientation to the targeted gene, so there will be a minimal effect of the exogeneous promoter.
It is recommended that the targeting vector is designed with recognition sequences for a site-specific DNA recombinase (loxP or FRT) flanking the neo gene, so that the selection marker can be removed. It may be removed, either after gene targeting by transient transfection of an expression plasmid for the appropriate recombinase (Cre or FLP) into the targeted ES clone, or it may be removed in the targeted mouse by mating with Cre or FLP expressing mouse.
We recommend removal in the mouse, since this does not require a further round of transfection and selection in the targeted ES clone, which can reduce its germ line potential. We have available general Cre (pgk-Cre) and FLP deletors (FLP1eR). See Moross Repository.
If a conditional construct using the Cre-lox system to activate the targeting is planned, it is recommended to use FLP-FRT system to excise the selection marker. This avoids the complication of three lox sites during excision.
Negative Selection marker: A negative selection marker may be included in the construct. The most common one is the tk (thymidine kinase ) gene, which if incorporated into the genome, will induce sensitivity to the drugs gancyclovir or FIAU. If the tk gene is placed outside the arms of homology, it will be lost during a homologous recombination event and the clone will be resistant to drug selection. However, in a random insertion event the tk gene will be incorporated, rendering the clone drug sensitive. This should enrich for homologous recombination events in neo-resistant clones. More recently the DTA gene (diptheria toxin A) which inhibits protein sysnthesis, has become the negative selection marker of choice, since it requires no additional drug selection.
In our experience, we have generally not noted substantial enrichment using tk selection, and because of possible effects of drug selection on germ line competence of the cells, we do not recommend using tk selection. However a DTA cassette may be included.

5. Screening Strategy:

Southern Blot: A strategy for identifying the targeted locus by Southern blot analysis must be developed during the vector design.
Two DNA probes, one 5’ to the targeting vector, and one 3’ to the targeting vector must be able to detect a change in fragment size, resulting either from the introduction or elimination of a restriction enzyme site by homolougous recombination. For more details see Southern Analysis. Not all restriction enzymes work efficiently in the crude genomic DNA extracts used in colony screening. A table of restriction enzymes efficiency is available.
It is essential to carry out a test blot, before transfection of the ES cells with the targeting vector. This is in order to verify that the probes and enzymes planned for Southern blot analysis, work well. DNA and Southern blotting must be prepared from ES cells grown on 96 well plates using the identical protocol to be used in screening of neo resistant clones. A single, clean band of predicted size, representing the wild type non-recombined allele should be visible on a clean blot. We recommend doing the test screen before beginning to build the targeting vector.
PCR: It is possible to screen for homologous recombination by PCR. In this case strategy is to amplify a novel sequence created only by the homologous recombination event and not by either wild type DNA or random integration. Two primers are used, one that binds to the neo cassette and the other that binds to an endogenous genomic sequence just beyond the short arm. Amplification will only take place if the two primers are juxtaposed by homologous recombination. In order to calibrate the PCR conditions, it is necessary to build and transfect a control vector, which includes the positive selection cassette in the same direction and site as the targeting vector, followed by a an extended version of the short arm which includes more downstream endogenous genomic sequences. The control vector does not need to include the negative selection cassette and the long arm. The control vector is transfected into ES cells and DNA from neo-resistant colonies is used as a template. The system must be fully calibrated before transfection of the targeting vector. This can take considerable time.
It is accepted practice that clones positive for homolgous recombination by PCR are reconfirmed positive, by Southern blotting using both 5’ and 3’ probes.
Although PCR may be simpler, it suffers from a significant degree of false positive and false negative results. Where homologous recombination frequency is low, it is possible to miss all recombination events. We generally recommend screening by Southern blotting only.
We also recommend probing clones positive for homolgous recombination with an internal neo probe. This is to rule out the possibility of secondary random integration events by the targeting vector, within the same clone.

6. Conditional Alleles:

In many cases, following disruption of a target gene, the deficiency of the gene leads to embryonic lethality, precluding the analysis of gene function in later developmental stages or in the adult. This problem can be overcome by creating conditional knockout animals, allowing a gene to be inactivated in a tissue- specific or temporal-specific fashion. Typically, a conditional knockout allele is made by inserting loxP or FRT sites into two introns of a gene, flanking the exons to be deleted. Expression of Cre or FLP recombinase in the animal carrying the conditional knockout allele catalyzes recombination between the loxP sites or the FRT sites respectively, and inactivates the gene. By expressing Cre or FLP recombinase from a tissue-specific promoter, the gene can be inactivated in a tissue-specific fashion, or it can be inactivated in all tissues by crossing with a deletor strain, which expresses Cre or FLP ubiquitously. A range of Cre-expressing mice are available from the WIS mouse strains databases.

The main objectives in designing a conditional construct are that the expression of the target gene should not be disturbed by the presence of the loxP sites, but should become inactivated (or otherwise modified) upon Cre-mediated recombination. A conditional construct contains a selectable marker gene (neo) flanked by two regions of homology to the target gene, and a gene for negative selection, if desired. A short homology arm of 1 kb or less is required only if the transfected ES clones will be tested by PCR. In contrast to a conventional gene targeting construct, the selection marker does not disrupt or replace one of the exons. It is placed into an intron and flanked usually, by two FRT sites, next to a single loxP site. (See loxP/FRT). A second loxP site is placed into one of the homology arms such that both loxP sites are flanking one or more exons of the target gene (See Conditional Construct scheme). The loxP sites should be placed at least 200 bp from the exon boundaries to reduce the risk of interference with splicing signals. Generally a region of 1–3 kb comprising one or more exons of the target gene is chosen for deletion. Although it is generally easier to inactivate a gene functionally by the excision of many exons, the deletion of a large loxP-flanked segment is less efficient, as recombinase efficiency decreases the further apart the loxP sites are. For a detailed discussion of conditional mutations see Kuhn and Schwenk.

A conditional targeting scheme is recommended for all knockout projects, as this offers far greater flexibility compared with a conventional knockout scheme, and avoids the potential consequences of embryonic lethality.

Generating Your Construct

Today there are two main ways to generate targeting constructs:

  • conventional PCR cloning, whereby long-range PCR is used to amplify a large genomic region spanning the exons or region to be targeted, of the gene of interest. A positive (neo) selectable marker gene is introduced at a suitable restriction site within target region, and a negative selection marker gene may be introduced at the end of one of the homologous arms. Desired restriction sites may be introduced via the PCR primers.
  • Recombineering technology, an alternative procedure developed more recently. Read further for an outline of this technology, whose methodology is highly recommended.

Basic schemes for generating knockout vectors, conditional knockout vectors and knockin vectors are outlined under Recombineering Schemes.
Once the targeting vector has been completed we recommend sequencing of essential elements such as the reporter sequence (if used), the eukaryotic promoter of the selection marker, Lox and FRT sites etc.