View the examples by clicking the buttons above.

 

Example 1: Combining a knockout, a tissue specific knockout, and a human gene transgene

For this example we want to generate a mouse where the endogenous mouse Gene L is knocked out, the human form of GENE L is expressed, and the endogenous mouse Gene M is knocked out in specific cells using the Cre-lox system.  We have the following four mouse strains to begin with.

1. Homozygous Gene L knockout(KO), viable and fertile as homozygotes
2. Homozygous Gene M loxp-flanked (flox), viable and fertile as homozygotes
3. Hemizygous for Cre Tg, viable and fertile
4. Hemizygous for Human (Hu) GENE L transgene, viable and fertile (in this example, we do not know if homozygotes are viable or fertile)

All four genes/transgenes are on different chromosomes, so all are segregating independently.  The four strains are NOT on the same genetic background.  Therefore, we will need our experimental and control mice to come from the same colony.

Let’s work through the four steps for this example….

Step1:  Genotype

The genotype of the experimental mice we will need is as follows:

Homozygous (HOM) for Gene L KO / Homozygous (HOM) for Gene M flox / Hemizygous (HEMI) for Cre transgene (Tg) / HEMI for Hu GENE L Tg.

The control mice we will need are:

HOM for Gene L KO / HOM for Gene M flox / Noncarriers (NCAR) for Cre transgene (Tg) / HEMI for Hu GENE L Tg.

Step 2:  Final Cross/Colony

Now we determine the breeding schemes of our final cross for each mutation or transgene individually,

Gene L KO:  We need homozygotes for all our mice, and homozygotes are viable and fertile.  Therefore, for this mutation, we can breed HOM x HOM.

Gene M floxed:  We need homozygotes for all our mice, and homozygotes are viable and fertile.  Therefore, we can also breed this mutation HOM x HOM.

Cre Tg:  We need hemizygotes for the experimental mice and noncarriers for the control mice.  We can breed HEMI x NCAR (female x male) or NCAR x HEMI (reciprocal) for this transgene.

Hu GENE L Tg:  We need hemizygotes for all our mice.  Since we do not know if homozygotes are viable or fertile for this transgene, we will have to breed HEMI x NCAR or the reciprocal (NCAR x HEMI).

Now that this is understood, we can line up all four individual crosses.  The two transgenes can be crossed in either direction (passed through either the female or the male), but we set them here in the opposite direction, as we do not need to combine all four alleles in the same breeders.  We next combine the individual genotypes on the left of the “X” into one breeder, and combine the genotypes on the right of the “X” into the other breeder.

HOM Gene L KO          X    HOM Gene L KO

HOM Gene M flox        X    HOM Gene M flox

HEMI Cre Tg                    X    NCAR Cre Tg

NCAR Hu GENE L Tg   X    HEMI Hu GENE L Tg

HOM Gene L KO / HOM Gene M flox / HEMI Cre / NCAR Hu GENE L Tg    X    HOM Gene L KO / HOM Gene M flox / NCAR Cre Tg / HEMI Hu GENE L Tg 

Or the reciprocal cross, referring to the two hemizygous alleles and which sex carries the cre or GENE L transgene.

From this cross, 25% (1/4) of the mice produced will be experimental mice and 25% (1/4) will be controls.  The breeder genotypes will also be produced (one is the same genotype as the controls). See figure 2.

Fig 2.  Punnett square crossing HEMI Cre / NCAR Hu GENE L Tg  X  NCAR Cre Tg / HEMI Hu GENE L Tg.

Fig2:  Punnett square crossing HEMI Cre Tg / NCAR Hu GENE L Tg  X  NCAR Cre Tg / HEMI Hu GENE L Tg.  One of the four squares are HEMI Cre / HEMI Hu GENE L Tg, which indicates that 1/4 (25%) of the progeny will be HEMI Cre (Cc) / HEMI Hu GENE L Tg (Ll)  The Gene L KO and  Gene M flox are not included in the Punnett square as they are both mated as HOM X HOM, which will give 100% HOMs.

Step 3:  Identify Common Alleles 

Lining up the alleles for the final cross of breeders as shown in step 2, allows us to easily see that all of our mice need to be homozygous for both the Gene L KO and Gene M flox.  Therefore they are the common alleles that we will fix for homozygosity in our final breeders.

Step 4:  Design Breeding Scheme To Get From Initial Strains to Final Colony

Since we need to generate two different genotypes for our final breeders, we will perform two parallel series of crosses concurrently, one for each genotype.  We then bring them together at the end to produce the mice for experiments.  All of the following crosses can be set up in both directions, but we will list only one direction for relative simplicity.

Generation 1:

We will start by setting up three crosses.

• HOM Gene L KO    X   HEMI Cre Tg
• HOM Gene M flox  X   HEMI hu GENE L Tg
• HOM Gene L KO     X   HOM Gene M flox

There is flexibility for the first two crosses at this stage.  We could have crossed the Gene L KO to the Hu GENE L Tg and the GENE M flox to the Cre Tg instead.  None of the downstream expected Mendelian ratios would change if we used those crosses instead.  The progeny from the third cross will help in the second generation crosses, as you will see below.

The progeny we will produce from these three crosses and their expected Mendelian ratios are:

• HET Gene L KO    /  HEMI Cre Tg  (1/2 or 50%)
• HET Gene M flox /  HEMI hu GENE L Tg (1/2 or 50%)
• HET Gene L KO   /  HET Gene M flox  (100%)

Generation 2:

In the second generation, we will cross the progeny from the first two crosses of the previous generation to the progeny from the third cross in order to fix one of the common alleles for homozygosity.

• HET Gene L KO    /  HEMI Cre Tg                 X  HET Gene L KO  /  HET Gene M flox
• HET Gene M flox /  HEMI hu GENE L Tg  X  HET Gene L KO  /  HET Gene M flox

The progeny we will select from those two crosses and their expected Mendelian ratios are:

• HOM Gene L KO /  HET Gene M flox / HEMI Cre Tg  and  HOM Gene L KO  /  HET Gene M flox / NCAR Cre Tg  (1/16, 6.25% for both genotypes)
• HET Gene L KO / HOM Gene M flox /  HEMI hu GENE L Tg and HET Gene L KO / HOM Gene M flox  /  NCAR hu GENE L Tg (1/16, 6.25% for both genotypes)

Note that in cross 1 we are looking for mice homozygous for the Gene L KO and in cross 2 we need mice homozygous for the Gene M flox allele.

Generation 3:

 As shown in the prior generation, we have now fixed one of the common alleles for homozygosity in each of these parallel crosses.  The next step is to fix the second common allele for homozygosity as shown in the following two crosses:

• HOM Gene L KO  /  HET Gene M flox / HEMI Cre Tg  X  HOM Gene L KO  /  HET Gene M flox / NCAR Cre Tg
• HET Gene L KO / HOM Gene M flox /  HEMI hu GENE L Tg X  HET Gene L KO / HOM Gene M flox  /  NCAR hu GENE L Tg

Note that the first cross is set up to re-establish homozygosity for Gene M and the cre is segregating. The second cross will re-establish homozygosity for Gene L and the human GENE L transgene is segregating.  The progeny we will select from these two crosses and their expected Mendelian ratios are:

• HOM Gene L KO /  HOM Gene M flox / HEMI Cre Tg (1/8, 12/5%)
• HOM Gene L KO / HOM Gene M flox /  HEMI hu GENE L Tg (1/8, 12.5%)

These are the genotypes we need to cross together to produce our experimental and control mice!  It took three generations of breeding to get to this point.  Assuming 3 months per generation, we can estimate 9 months to get here.

Generation 4 and beyond:

Now we set up the following cross to expand the colony and start producing mice for experiments!

HOM Gene L KO  /  HOM Gene M flox / HEMI Cre Tg   X  HOM Gene L KO / HOM Gene M flox  / HEMI Hu GENE L Tg

In this example, we need the Hu GENE L Tg in all the mice, but we do not know if homozygotes for that transgene are viable and fertile.  Once we have this colony established, we can consider attempting to make that transgene homozygous.  Alternatively, this is something that could have been considered in the planning phase of the breeding strategy. An independent test colony could have been established during the first cross to see if mice homozygotes for GENE L survive and breed in the absence of the other alleles. If the homozygotes are in fact viable and fertile, with no unwanted phenotypes, additional crosses can be performed to generate breeders homozygous for the GENE L transgene. Once established, the genotypes of the breeders would be the following;

HOM Gene L KO  /  HOM Gene M flox / HEMI Cre Tg / HOM Hu GENE L Tg   X  HOM Gene L KO / HOM Gene M flox  / NCAR Cre Tg/ HOM Hu GENE L Tg

These breeders would then produce 50% experimental and 50% control mice, making the colony much more efficient.  The strategy for creating triple homozygotes is explained in the next section.

Example 2:  Triple Homozygous KO Mice 

For this example we want to generate a mouse that is homozygous for knockout alleles of three genes; Gene A, Gene B, and Gene C.   We have the following three mouse strains to begin with;

1. Homozygous Gene A KO, viable and fertile as homozygotes
2. Homozygous Gene B KO, viable and fertile as homozygotes
3. Homozygous Gene C KO, viable and fertile as homozygotes

All three genes are on different chromosomes, so they are segregating independently.  The three strains are all on the same genetic background (let’s use C57BL/6J). 

Let’s work through the four steps for this example….

Step1:  Genotype

The genotype of the experimental mice we will need for this example is;

HOM Gene A KO / HOM Gene B KO / HOM Gene C KO

The control mice we will need are wild type (WT) for all three alleles;

HOM Gene A WT allele / HOM Gene B WT allele / HOM Gene C WT allele

Step 2:  Final Cross/Colony

For this example we need mice that are homozygous for three KO alleles and mice that are homozygous for three WT alleles.  Generating both genotypes from a single colony would require an enormous colony and would be very inefficient.  In order to do it this way,

we would need to breed female and male mice that are heterozygous for all three KO alleles. The colony would produce only 1/64 (1.5%) of triple homozygous mice for our experimental group and a similar 1/64 for the control wild type mice.  These frequencies are simply not feasible for most researchers.

Since all three individual starting strains are on the same genetic background ( C57BL/6J ), we can use C57BL/6J inbred mice as our controls.  We will then focus our breeding solely on generating the experimental mice.  We could maintain a separate colony for the controls, or purchase them as needed.  Purchasing them saves cost, frees up space for other breeding colonies and experiments, and significantly diminishes wasted resources.

While the genotypes we will generate for our final cross may be more obvious in this case, let’s go through it again.  We begin by working backward as in the first example and look at each allele independently.

Gene A KO:  We need homozygotes and homozygotes are viable and fertile.  Therefore, for this KO allele we can breed HOM X HOM

Gene B KO:  We need homozygotes and homozygotes are viable and fertile.  Therefore, for this KO allele we can breed HOM X HOM

Gene C KO:  We need homozygotes and homozygotes are viable and fertile.  Therefore, for this KO allele we can breed HOM X HOM

Now we line up all three individual crosses and combine.

HOM Gene A KO  X  HOM Gene A KO

HOM Gene B KO  X  HOM Gene B KO

HOM Gene C KO  X  HOM Gene C KO

HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO  X   HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO

From this cross 100% of the progeny will be our experimental mice, which in this case are the same genotype as our breeders.  Hence, we will produce the experimental mice and regenerate new breeders at the same time.

Step 3:  Identify Common Alleles 

In this example, we have no alleles that are common between the experimental and control mice.

Step 4:  Design Breeding Scheme To Get From Initial Strains to Final Colony

For this scenario, we will need three concurrent crosses every generation.  We will start by crossing each of the individual strains to each other, then begin fixing one KO allele at a time for homozygosity until all three KO alleles are homozygous.

Generation 1:

We will start by setting up three crosses.

• HOM Gene A KO X  HOM Gene B KO
• HOM Gene A KO X  HOM Gene C KO
• HOM Gene B KO X  HOM Gene C KO

The progeny we will produce from those three crosses and their expected Mendelian ratios are:

• HET Gene A KO /  HET Gene B KO  (100%)
• HET Gene A KO /  HET Gene C KO  (100%)
• HET Gene B KO /  HET Gene C KO  (100%)

Generation 2:

In the second generation we will cross the progeny from each of the first three crosses back to one of its original homozygous KO parental strains.  There is flexibility at this stage, we just need to use each original homozygous KO strain once.

• HET Gene A KO /  HET Gene B KO  X  HOM Gene A KO
• HET Gene B KO /  HET Gene C KO  X  HOM Gene B KO
• HET Gene A KO /  HET Gene C KO  X  HOM Gene C KO

The progeny we will select from those three crosses and their expected Mendelian ratios are:

• HOM Gene A KO /  HET Gene B KO  (1/4, 25%)
• HOM Gene B KO /  HET Gene C KO  (1/4, 25%)
• HET Gene A KO /  HOM Gene C KO  (1/4, 25%)

Generation 3:

In the third generation we will cross all three combinations of progeny from the previous generation together to generate mice that are homozygous for one knockout allele AND heterozygous for the other two knockout alleles

• HOM Gene A KO /  HET Gene B KO  X  HET Gene A KO  /  HOM Gene C KO 
• HOM Gene A KO /  HET Gene B KO  X  HOM Gene B KO  /  HET Gene C KO   
• HET Gene A KO /  HET Gene C KO  X  HOM Gene B KO  /  HET Gene C KO 

The progeny we will select from those three crosses and their expected Mendelian ratios are:

• HOM Gene A KO /  HET Gene B KO / HET Gene C KO  (1/4, 25%)
• HET Gene A KO /  HOM Gene B KO  / HET Gene C KO  (1/4, 25%)
• HET Gene A KO /  HET Gene B KO  /  HOM Gene C KO  (1/4, 25%)

Generation 4:

Now we can cross any and all of the above progeny from the third generation together.  All combinations will produce HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO  with the same expected frequency 6.25% (1/16).  Note that because the triple homozygote frequency is very low and you need to identify both females and males, you will need to set up as many breeders as possible and screen a large number offspring.

Since it took four generations of breeding to get to this point, we can estimate that it will take a year to get here.

Generation 5 and beyond:

Now we breed HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO  X  HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO  to expand the colony and produce mice for experiments.

Example 3:  Quadruple Knockouts from Existing Triple Knockouts

In this example, we will assume the triple knockout mice we generated in the previous example have become a fantastic model that opened up new areas of research and lead to new research questions.  Now we want to add an additional knockout for Gene D to the existing triple knockout mice.  The strains we have to start with are…

1. HOM Gene A KO /  HOM Gene B KO  /  HOM Gene C KO,  viable and fertile 
2. HET Gene D KO, homozygotes are viable but are not fertile

All four genes are on different chromosomes, so they are segregating independently.  The two strains are all on the same C57BL/6J genetic background. 

Step1:  Genotype

In this case, we will compare the quadruple knockout mice to the original triple knockout mice.  The genotype of the experimental mice we will need for this example are:

HOM Gene A KO / HOM Gene B KO / HOM Gene C KO / HOM Gene D KO

The control mice we will need are:

HOM Gene A KO / HOM Gene B KO / HOM Gene C KO / HOM Gene D WT

Step 2:  Final Cross/Colony

We look at each allele individually.

Gene A KO, Gene B KO, and Gene C KO:  We need homozygotes, and homozygotes are viable and fertile.  Therefore, for this the Gene A, Gene B, and Gene C KO alleles we can breed HOM X HOM

Gene D KO:  We need homozygotes for both the knockout and wild type alleles.  As the homozygotes do not breed, we have to breed HET X HET

Now we line up all four individual crosses and combine.

HOM Gene A KO  X  HOM Gene A KO

HOM Gene B KO  X  HOM Gene B KO

HOM Gene C KO  X  HOM Gene C KO

HET Gene D KO  X  HET Gene D KO

HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO / HET Gene D KO  X  HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO /  HET Gene D KO

From this cross 25% ( ¼ ) of the progeny will be our experimental mice.  The control mice are the same genotype as the triple knockout mice we began the project with.  We could use those existing mice as our controls as the genetic backgrounds will be matched.  However, we will also generate the control genotype from this cross at a frequency of 25% ( ¼).  Therefore we can produce both the experimental and control mice from the same colony with no loss of efficiency.  We will also regenerate new breeders from this cross at a frequency of 50% ( ½).

Step 3:  Identify Common Alleles

Homozygotes for the Gene A KO, Gene B KO, and Gene C KO are needed in the experimental and control mice.  They are the common alleles

Step 4:  Design Breeding Scheme to Get From Initial Strains to Final Colony

Since we already have the triple homozygous knockout mice, we will use them for the first two generations of breeding.  We will NOT need to use multiple crosses for each generation, highlighting the advantage of already having a colony fixed for these alleles.

Generation 1:

We start by crossing the existing triple knockout mice to the heterozygous Gene D KO mice.

• HOM Gene A KO /  HOM Gene B KO  /  HOM Gene C KO  X  HET Gene D KO

The progeny we will need from this cross and the expected Mendelian ratio is:

• HET Gene A KO /  HET Gene B KO  /  HET Gene C KO  /  HET Gene D KO  50% ( ½)

Generation 2:

Next, we cross those quadruple heterozygotes back to the original triple homozygous mice.

• HET Gene A KO /  HET Gene B KO  /  HET Gene C KO  /  HET Gene D KO  X HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO 

The progeny we will need from this cross and the expected Mendelian ratio is:

• HOM Gene A KO /  HOM Gene B KO  /  HOM Gene C KO / HET Gene D KO 25% (1/16)

However, don’t forget we need females and males to proceed, so be sure to set up as many breeders as possible.

With only two generations of breeding, we can estimate it will take us about 6 months to get here.  That is the advantage of starting with the triple homozygotes!

Generation 3 and beyond:

Generation 2 will produce the final genotype we need, so we set up the final cross below to expand the colony and produce the experimental and control mice required.

HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO / HET Gene D KO  X 

HOM Gene A KO  /  HOM Gene B KO  /  HOM Gene C KO / HET Gene D KO

Example 4: Model Generation Technologies and IVF Increase Efficiency

The efficiency of generating new mouse models by adding additional mutations or transgenes to existing complex mouse models can be significantly increased by utilizing technologies like CRISPR or pro-nuclear microinjection.

For example, let’s look at adding a new mutation to the existing NSG™ ( NOD.Cg-Prkdc^scid Il2rg^tm1Wjl/SzJ ) mouse strain.  The NSG™ mice are a commonly used immunodeficient mouse strain and are maintained as homozygous for the Prkdc^scid (scid) and homozygous/hemizygous for the Il2rg^tm1Wjl  knockout allele (Il2rg is X-linked, so mutant males are hemizygotes). 

In this example we will generate a mutation in Gene Q that changes an amino acid to match the human protein sequence (Hu Gene Q) directly in the NSG™ mouse strain using CRISPR.  We expect that homozygotes for the Hu Gene Q will be viable and fertile.

Step1:  Genotype

The genotype of the experimental mice we will need are:

HOM Hu Gene Q / HOM scid / HOM or HEMI for Il2rg KO

The control mice we will need are:

HOM Gene Q WT / HOM scid / HOM or HEMI for Il2rg KO

Step 2:  Final Cross/Colony

The genotype of mice needed for the controls is the same as the NSG™ mouse strain.  Since we are making the Hu Gene Q mutation directly into the NSG™ mice, we can use the NSG™ mice as controls.  Therefore, we focus only on generating a colony to produce the experimental mice.

Hu Gene Q:  We need homozygotes.  We are expecting that homozygotes will be viable and fertile, and will work off that assumption.  However, any newly generated mutation or transgene can have unexpected phenotypes.  If we discover viability or fertility issues with the Hu Gene Q mutation, we will have to adjust accordingly at that time.

Scid mutation:  We need homozygotes, which are viable and fertile.

Il2rg KO:  We need homozygotes and hemizygotes (X-linked!), which are viable and fertile.

Line up all four individual crosses and combine.

HOM Hu Gene Q  X  HOM Hu Gene Q 

HOM scid               X   HOM scid             

HOM Il2rg KO       X   HEMI Il2rg KO

HOM Hu Gene Q / HOM scid / HOM Il2rg KO   X   HOM Hu Gene Q / HOM scid / HEMI for Il2rg KO

All of the progeny from this cross will be experimental mice or replacement breeders.

Step 3:  Identify Common Alleles

Homozygotes for the scid mutation and the Il2rg KO are needed in the experimental and control mice.  They are the common alleles. 

Step 4:  Design Breeding Scheme to Get From Initial Strains to Final Colony

We are generating the new Hu Gene Q mutation directly in the NSG™ mice.  As the NSG™ mice are already homozygous for the scid mutation and the Il2rg KO, we do not need to be concerned about fixing them for homozygosity.  That is already done. The ability to cross back to the NSG™ mice will make this easier and far more efficient.

Generation 1:  Founder mice

By using CRISPR to generate the Hu Gene Q mutation directly in the NSG™ mice, the founder mice produced that carry the correct mutation will be;

HOM or HET Hu Gene Q / HOM scid / HOM or HEMI Il2rg KO

Generation 2:  N1

We then cross the founder mice back to the NSG™ mice. This is an important step needed to breed out off-target events and ensure germline transmission of the new allele.  Two founder mice should never be bred together.

HOM or HET Hu Gene Q / HOM scid / HOM Il2rg KO X   HOM Gene Q WT / HOM scid / HEMI Il2rg KO

The progeny we will need from this cross are:

HET Hu Gene Q / HOM scid / HOM or HEMI Il2rg KO

It is not possible to calculate the expected Mendelian ratios at this stage, as there can be mosaicism of the new mutant allele in the founders.  Since we do not know what percentage of the gametes from the founders will carry the new mutation, we cannot determine how many of the progeny will be heterozygotes for the Hu Gene Q.

Generation 3:  N2

In many cases we will not have enough N1 mice to attempt to fix the Hu Gene Q mutation to homozygosity.  Crossing back to the NSG™ mice again will help to expand the colony if necessary.  If there are not enough N1 progeny, we would breed as follows;

HET Hu Gene Q / HOM scid / HOM Il2rg KO  X  HOM Gene Q WT / HOM scid / HEMI Il2rg KO

The progeny we will need from this cross and the expected Mendelian ratio is:

HET Hu Gene Q / HOM scid / HOM or HEMI Il2rg KO  50% ( ½)

If there are enough N1 progeny, we could skip this step, and jump to the next cross.

Generation 4:

 Here, we cross mice HET for the Hu Gene Q mutation to generate homozygotes.

HET Hu Gene Q / HOM scid / HOM Il2rg KO  X  HET Hu Gene Q / HOM scid / HEMI Il2rg KO

The progeny we will select from this cross and the expected Mendelian ratio is:

HOM Hu Gene Q / HOM scid / HOM or HEMI Il2rg KO  25% ( ¼).

With four generations of breeding to get here, it would take about a year.

 Generation 5 and beyond:

Now we breed to expand the colony and produce mice for experiments.

HOM Hu Gene Q / HOM scid / HOM Il2rg KO  X  HOM Hu Gene Q / HOM scid / HEMI Il2rg KO

Use of IVF to Decrease Timeline:

In vitro fertilization (IVF) can dramatically accelerate this process.  Sperm from N1 or N2 HET Hu Gene Q / HOM scid / HEMI Il2rg KO males can be used for IVF with oocytes collected from NSG™ female mice (HOM Gene Q WT/ HOM scid / HOM Il2rg KO).  This can be scaled to produce as many HET Hu Gene Q / HOM scid / HOM or HEMI Il2rg KO 50% ( ½) mice as need to cross together to produce enough HOM Hu Gene Q / HOM scid / HOM or HEMI Il2rg KO  mice needed for experiments and further breeding, thereby saving at least 1-2 breeding generations or 3-6 months.

Please be aware that for CRISPR, pro-nuclear microinjection, and/or IVF to be practical, a large colony of the strain is required.  Also, these technologies do not always work consistently in all genetic backgrounds or strains.  Therefore, utilizing these technologies with strains that have not already proven successful increases the risk of such projects.