General Considerations for the Growth and Maintenance of Embryonic Stem Cells
Our embryonic stem cell line was isolated from 129X1/SvJ blastocysts and is called SSC10. This ES line has been proven to generate germline chimeras with high efficiency. SSC10 cells should be grown exclusively on MEF feeder layers. The use of feeder layers and LIF, together with careful culturing in high quality medium, acts to prolong the embryonic phenotype of the cell line. Since long term tissue culture will select for abnormal cells, the SSC10 line has been carefully expanded and subsequently frozen in liquid nitrogen to maintain its integrity.
After the expansion and freezing of SSC10 cells, they are tested and shown to be free of mycoplasma, have a normal male karyotype, are MAP tested (i.e. injected directly into mice to determine whether they are contaminated with known pathogens), and are blastocyst injected and go germline. The feeder layer used for the ES cells are primary murine embryonic fibroblasts (MEFs); cell lines have also been used as feeders (for example, see the Stratagene Website in our 'Links' section), but we have always been most satisfied with MEFs. These cells are easily isolated from E13.5 embryos.
MEFs are a primary cell line and therefore have a limited life span in culture. Fresh isolates are made on a regular basis and then frozen to ensure that cells are continuously available (see "protocols" section of this web site). MEFs can also be purchased from Specialty Media (800) 543-6029. The MEFs should be mitotically inactive when they are used as feeders for the ES cells. This can be accomplished by either gamma irradiation (3000 rads) or by treatment with the drug mitomycin C (10mg/ml in medium for 2 hours followed by extensive washing in drug-free medium [1]). These inactivated MEFs are then plated to give a uniform monolayer of cells onto which the stem cells are seeded.
Embryonic stem cells grow rapidly, dividing every 18 to 24 hours. The cells are kept at relatively high densities to ensure that a high rate of cell division is maintained as this seems to minimize the level of spontaneous differentiation. The cultures should be re-fed daily, and they need subculturing every 2 days. When passaging, it is important to trypsinize the cells well to ensure that a single cell suspension is generated. If the ES cell line fails to grow, either the cells have been plated too sparsely or the MEF feeder layer was of poor quality. It is important not to let the ES cells outgrow and cause the media to become highly acidic. After electroporation and subsequent selection, it is important to carefully pick stable clones for later expansion and DNA analysis. Clones (which contain many small nests of cells) should increase in size during selection but not undergo differentiation. The cells will pile up and maintain a discrete colony on top of the feeder cells. The cells within a clone will adhere tightly to one another, making it difficult to see individual cells; a tight border should surround the clone. The ES clones will grow well on the MEF feeder layer, forming various shapes, such as oblong or triangular. Clones that exhibit cellular heterogeneity, such as endodermal cell formation on the outside free surface of the colony, or a spreading of the cells at the periphery and general flattening, should not be selected. Disaggregated clones should be frozen slowly using the cryoprotectant dimethyl sulphoxide (DMSO) at a concentration of 10%. Clones will survive for years if stored under liquid nitrogen. For further information on Embryo-derived stem cell lines see Robertson, EJ: Teratocarcinomas and Embryonic Stem Cells: a practical approach. IRL Press Limited 1987, Chapter 4, p. 71-112.
(1) Robertson, EJ: Teratocarcinomas and embryonic stem cells: a practical approach. IRL Press Limited 1987, Chapter 2, p. 26. Transfection of Targeting Constructs into Embryonic Stem Cells
Vector Design, An Introduction
In this outline, we will share some of the expertise that we have gained
in the design of knock-out and knock-in vectors, and share strategies for
producing homologously recombined embryonic stem cells. Some of the points
that we will make are not published. However, this information comes from
a large collective experience of people doing these experiments here at Washington
University and elsewhere.
- General considerations for the design of targeting vectors
- A. Length of targeting arms.. Published studies
by one group suggest that the minimum size for a targeting arm is ~0.5kb,
however, this sized targeting arm will produce very low efficiency homologous
recombination. We generally use a minimal targeting arm length on one side
of 1 kb, with at least 2-3 kb on the other side. There is no reason why the
two targeting arms can't be of approximately the same length. We generally
make both the targeting arms ~2 kb in size. There may be a significant advantage
in using substantially longer targeting arms; however, large constructs are
more difficult to manipulate. In our experience, targeting arms of ~2 kb
on each side will yield homologous recombination efficiencies of 1-2%, which
are adequate for most purposes.
We frequently PCR our targeting arms using genomic DNA derived from SSC10 (129X1/SvJ) ES cells, or from cloned genomic DNA, if it is from a 129X1/SvJ source (see the section on SSC10 cells below). A few PCR errors seem to cause few (if any) problems for efficient homologous recombination. PCR certainly makes the ends of fragments easier to manipulate for insertion into vectors containing PGK-Neo.
- B. Single vs. double selection techniques. Most
targeting constructs in the past have used both PGK-Neo and HSV-TK to perform
+/- selection. However, gancyclovir treatment of ES cells is quite toxic,
and in our experience, it negatively affects the ability of ES cells to go
germline. We are willing to accept the lower rate of homologous recombination
in exchange for a better rate of germline transmission. Therefore, we use
single-selection vectors with PGK-Neo exclusively in our laboratory. Several
different kinds of PGK-Neo cassettes are available. Now, we are routinely
using PGK-Neo that is flanked by Lox-P sites, so that the PGK-Neo selectable
marker cassette can be removed by CRE-recombinase to avoid potential problems
with "neighborhood effects" of the PGK-Neo cassette. Recent work from our
laboratory and others have suggested that this neighborhood effect is very
problematic in multigene clusters (Pham et al, PNAS 93: 13090, 1996; Hug et al, MCB
16: 2906, 1996). If you know that your gene is in a multigene cluster, you
will need to anticipate the potential neighborhood effects of retained PGK-Neo
cassettes. Lox-P flanked PGK-Neo cassettes and a high performance CRE-recombinase
expression plasmid (pTURBO-CRE) are available from our laboratory (see below).
- C. Detection of homologous recombination events. Although some people have successfully used PCR to detect homologous recombinants, we recommend against this practice. We strongly recommend that you develop a unique probe that is external to the targeting sequences themselves and use it to screen using Southern analysis. It does not matter whether this probe is upstream or downstream from the targeting construct, it only matters that it is completely external, and that it contains no repetitive DNA elements. This probe should be used in conjunction with Southern analysis of each ES clone to determine whether or not a targeting event has occurred. In addition to defining a homologous recombination band, Southern analysis also allows you to assess the molarity of mutant bands, which is difficult to do by PCR. Basically, if the wild-type and mutant bands are not equal in intensity, you must be suspicious that your targeted clone is contaminated with wild-type ES cells. If it is, the wild-type cells will have an advantage when they are injected into blastocysts, and you will get chimeric males that throw mostly wild-type agouti pups.This can be a huge problem. If the targeted bands look correct on the Southern, but the molarity is suspect, then we strongly recommend that the cells be re-selected in G-418, or even replated by limiting dilution to subclone the ES cells of interest.
- D. Generation of chimeric mice. There are several
services available for making chimeric mice from your targeted ES clones
available within the Medical School. Please note that you must have an "active"
animal protocol that covers the chimeric mice you will receive before injections
can be contracted to a specific lab. Additional Commercial and Academic Injection Services
can be accessed in the "Links" section of our website.
WUMS Microinjection Service Options
- Mouse Genetics Core
http://mgc.wustl.edu- Pathology Microinjection Core
http://pathimm.wustl.edu- Ron McCarthy
McCarthy_R@wustl.edu
Tel: 314 454-7990
Removal of Selectable marker cassettes with Cre-Lox mediated recombination
As noted above, the retention of selectable marker cassettes in targeted loci can have effects on the expression of neighboring genes (and even on the targeted gene itself), even at long distances. For this reason, our PGK-Neo selectable marker cassette is flanked by Lox P recombination sites. When Cre-recombinase is expressed in ES cells that contain the selectable marker cassette, the PGK-Neo cassette will sometimes be deleted, leaving behind a single Lox P site in the targeted locus. To our knowledge, the presence of this site rarely, if ever, affects expression of the gene in the targeted locus. However, this is a possibility that should be considered as your targeting vector is designed.Cre-recombinase can be expressed either transiently or stably in ES cells, or in a tissue-specific fashion in transgenic mice. We and others have used transient transfection of embryonic stem cells with Cre vectors like pMC-Cre. In our hands, this has been quite inefficient, yielding deletion of the selectable marker cassette in perhaps 1-2% of clones. We have therefore attempted to increase excision efficiency by building a Cre cassette that is expressed at higher levels, and that also contains a nuclear localization signal (NLS) to improve trafficking to the nucleus, where this enzyme acts. We generated a Cre cDNA with an NLS at the amino terminus, and inserted this cassette into an expression plasmid created by Mark Sands here at Washington University (pCAGGS). This plasmid (pTURBO-CRE) contains a CMV enhancer upstream from a chicken b-actin promoter driving the untranslated region of the first exon, first intron, and part of the second exon of the beta-actin gene. The cDNA is inserted at a unique Eco RI site, and is followed by a rabbit beta-globin polyadenylation site (see map below). This expression cassette has been shown to produce extraordinarily high levels of mRNA from the beta-actin promoter in a variety of transient and stable expression assays. We inserted the NLS-Cre cassette into this construct, and have used this construct in transient transfection assays of targeted ES cells to remove the LoxP flanked PGK-neo cassette. After electroporation, complete dispersal of the ES cells is extremely important , since the cells must be replated in the absence of G418 (the positive selection cassette will be removed if the transfection is successful); each colony must be derived from a single ES cell. If colonies do not grow from single cells, these colonies will contain mixtures of PGK-neo positive and negative cells, a problem that will be detected in most cases by the southern blot analysis of the transfected clones.
Using this transient transfection technique, we have found that up to 50% of the transfected colonies become G418 sensitive, and nearly all of these G418 sensitive colonies have undergone Cre-mediated excision of the PGK-neo cassette, as demonstrated by Southern blot analysis. The efficiency of PGK-neo cassette removal is so high that we now pluck only 24 clones for analysis. The methods for transient transfection of ES cells with pTURBO-CRE can be found in the Protocols section of this website.
Finally, Lox P flanked regions can be removed from the germline of mice by breeding them with mice that express the cre-recombinase expressed from a transgene. Several groups have used this strategy to obtain tissue-specific deletions of genes by targeting expression of Cre-recombinase using transgenic targeting sequences. In the past, the removal of the Lox P flanked sequences by Cre in the whole animal has been inefficient, but efficiency rates are improving with better transgenes and better versions of Cre. Westphal and colleagues (Proc Natl Acad Sci 93: 5860, 1996) have shown that Lox P flanked cassettes can be efficiently removed in developing mouse embryos using a transgene that expresses Cre under the control of the adenovirus EIIa promoter. These mice are available from Westphal's lab at the NIH. Janet Rossant and Andy McMahon have recently written a topical review of tissue specific Cre-expressing mice (Genes and Development 13:142-145, 1999), which you should consult before considering this approach. Finally, a voluntary database of cre-expressing mice has been established by Andras Nagy, which we would encourage you to consult and support (http://www.mshri.on.ca/nagy/cre.htm).
Vector Maps
Genbank Accession Number AF335419
For additional p-neo-LoxP plasmid 1339, please see Genbank Accession Number AF335420Material Transfer Agreement
- Access the Material Transfer Agreement.
Genbank Accession Number AF334827Material Transfer Agreement
- Access the Material Transfer Agreement.
The Origins and Use of SSC10 Embryonic Stem Cells
SSC10 embryonic stem cells were derived from 129X1/SvJ mice. A recent paper by Elizabeth Simpson and colleagues (Nature Genetics 16:19-27, 1997) at the Jackson Laboratory has carefully evaluated genetic variation among all of the 129 substrains, and their importance for understanding targeted mutations in mice. I strongly recommend that you examine this paper before starting your experiments. The wide variety of embryonic stem cell lines that you can obtain from different investigators can be highly divergent within the 129 background. You must choose these cell lines with care, since they are not all equivalent. There are a number of allelic differences between the 129 substrains that cause minor histocompatability differences, and there are many differences in the backgrounds of the 129 substrains (and even in ES cell lines obtained from these strains). Of note, a careful evaluation of the 129X1/SvJ substrain (Jackson Laboratory #000691), from which SSC10 cells were made, reveals no allelic differences between the mice and the SSC10 cells. Therefore, no obvious mutations have been acquired during the preparation of our cell line. A large number of laboratories (several hundred) have now used SSC10 ES cells.
References for ES cell lines and ES Core use
Our Targeting Record (as of 07/15/08)
Since the Core opened in 1996, 619 constructs (made in 130 different laboratories) have been electroporated, and 109,788 clones have been selected, expanded, and evaluated by Southern. 1257 clones have been targeted homologously (1.2% of all clones) using our strategy of single selection, isogenic DNA in the targeting arms, and arm lengths of >2.0 kB. ES cell clones from 290 of these constructs have been injected into blastocyts. 268 out of these 290 constructs have yielded chimeras (268 / 290 = 92%). The status of 238 chimeras is known. 170 have had documented germ line transmission of the mutant allele (170 / 238 = 71%), 68 did not go germline, 30 are still breeding or were not bred.