TNBGGA: The No Bullshit Guide to Genetic Analysis

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
<-chapter_15|Chapter 15^table_of_contents|Table of Contents^chapter_17|Chapter 17->

<typo fs:x-large>Chapter 16. %%Gene knockouts%% in multicellular model organisms</typo>

===== Totipotent mouse embryonic stem cells =====


Knocking out mouse genes, or deliberately targeting a specific gene for mutation (as opposed to creating random mutants with a chemical mutagen), is a much more complex process than making transgenic mice. To discuss this, some background information about the preimplantation mouse embryo is first needed. For about 4-5 days after fertilization, mouse embryos are free-floating in the uterus (which means we can surgically remove them) and many of the cells that will eventually form the mouse remain totipotent, meaning that they have the potential to differentiate into any kind of mouse cell type (Fig. 16.1). This has been shown in various dramatic ways. For instance, if the four-cell embryo is dissected and each cell implanted into a different foster mother, four identical mice will be born. More interestingly, if cells from two genetically different pre-implantation embryos (e.g., embryos destined to produce mice with different fur colors) are simply mixed together (the cells will naturally stick together) and implanted into a foster mother, a single chimeric mouse will be born. Essentially the two types of totipotent cells mix together and produce an animal that has a mixture of two types of cells in its body. The ability of these genetically different totipotent cells to mix together in the preimplantation embryo is crucial for the mouse gene knockout technology. 

<figure Fig1>
{{ :preimplatation_embryonic_development.jpg?400 |}}
<caption>
Preimplantation mouse embryonic development. Many cells through embryonic day 4.5 remain totipotent. Source: Mihajovic and Bruce (2017). Open Biology 7:170210. https://doi.org/10.1098/rsob.170210. Licensing: [[https://creativecommons.org/licenses/by/4.0/|CC BY 4.0]].
</caption>
</figure>


In order to make a directed genetic change in a specific mouse gene we exploit homologous recombination. However, this is much harder to do in mammalian cells than bacteria and yeast. In yeast, when a linear fragment of dsDNA is introduced into a yeast cell, that DNA is integrated into the yeast genome by homologous recombination about 90% of the time, such that the incoming DNA fragment replaces the endogenous gene on the chromosome. In mammalian cells, DNA that is introduced into a cell almost always integrates at a non-homologous site (see [[chapter_15|Chap. 15]]), and the frequency of homologous recombination at the endogenous locus is very low - about 10<sup>-3</sup> to 10<sup>-5</sup> (or 0.1% - 0.001%). To have a chance at finding rare events like this, we need to have an experimental setup where we can: 
  - look through thousands of independent integration events; and
  - a way to be able to identify the specific integration event we want, namely an integration event that took place by homologous recombination.

The first crucial technological advancement for creating mouse knockouts was being able to grow the totipotent cells from preimplantation embryos in culture in the lab; these cells are called mouse embryonic stem cells (ES cells). Cells from the inner cell mass of a preimplantation embryo at the blastocyst stage can be removed and cultured in a dish without the cells losing their totipotency (Fig. {{ref>Fig2}}); even after being cultured in the lab for many years these cells could still be introduced back into a preimplantation embryo and go on to make all the tissues of a mouse. 

<figure Fig2>
{{ :es_cell_culture.png?400 |}}
<caption>
Culturing totipotent mouse ES cells. The targeting construct is described in more detail in Fig. {{ref>Fig3}}.
</caption>
</figure>
 
This gave scientists an opportunity to genetically manipulate ES cells in a Petri dish, in an analogous way as we might genetically manipulate yeast or //E. coli// cells in a Petri dish. Importantly, this further gives scientists the opportunity to isolate clones as colonies and also the ability to grow a large number of cells to screen for rare events such as homologous recombination (it is far easier and cheaper to grow thousands of cells in a dish than it is to grow thousands of mice in cages). Once you have identified mouse ES cells (originally from a gray-furred mouse) that have been genetically altered the way you wish, these cells can be used to generate a living animal that is fully derived from these totipotent ES cells. 

===== Selecting for homologous recombination in ES cells =====

The second crucial technological advancement for creating mouse knockouts was the invention of a clever scheme to select for ES cells that integrate DNA via homologous recombination. Recall that we first introduced the concept of screens and selections in [[chapter_08#screen_selection|Chap. 08]]. A selection is simply a strategy for only allowing rare events you want to find to survive in an experiment. For generating mouse knockouts, a double selection scheme is used - a positive selection and a negative selection (Fig. {{ref>Fig3}}). We first create a targeting construct. If we wanted to knock out gene $X$, for instance, our target construct would have the following features:

  * We need two DNA fragments with sequences that match portions of gene $X$. Usually we will pick one fragment that is near the start of gene $X$, and another fragment that is some distance downstream from the first fragment. We can obtain these DNA fragments in a variety of ways, but PCR is probably the easiest way to do this. We call these two fragments the left homology arm and the right homology arm (sometimes just the left arm and right arm). Note that in [[chapter_15|Chap. 15]], we discussed gene knockouts in yeast and mentioned that 30-50 bp is sufficient for homology arms in a yeast knockout experiment. In mouse ES cells, the homology arms must be much longer - usually at least around 1000 bp.
  * In between these two fragments we will insert a gene called $neo^R$ (we also include all regulatory sequences needed for $neo^R$ to be expressed in ES cells). ES cells expressing $neo^R$ will be resistant to a drug called G-418. Wildtype ES cells will be killed by G-418.  
  * To the "right" of the right homology arm, we will add another gene called $tk^{HSV}$. This is a gene from herpes simplex virus (HSV) that codes for an enzyme called thymidine kinase. (We also include all regulatory sequences needed for $tk^{HSV}$ to be expressed in ES cells.) ES cells that express $tk^{HSV}$ do not die unless they are exposed to the drug ganciclovir.

<figure Fig3>
{{ :ko_targeting.png?400 |}}
<caption> 
Design of the targeting construct for a gene knockout and knockout double selection strategy.
</caption>
</figure>

Using this targeting construct, we can select for homologous recombination in three steps. 

  * First, we transfect  our targeting construct into ES cells. Instead of CaCl<sub>2</sub> or LiCl treatment we used to make //E. coli// and yeast cell competent, we instead use a technology called electroporation where cells are briefly shocked with a high-voltage pulse of electricity to make their membranes porous so that they can take up foreign DNA((This technique is called electroporation. It also works for yeast and //E. coli//.)).
  * Second, we use G-418 to positively select for all ES cells that have taken up our targeting construct; all ES cells that did not take up the targeting construct will die (another way we say this is that $neo^R$-expressing cells are selected for). At this point, the targeting construct is randomly integrated in most surviving cells, and homologous recombination at gene $X$ has only occurred in rare instances.
  * Third, we use ganciclovir to treat the ES cells that survived the second step. Because homologous recombination depends on the left and right homology arms and because $tk^{HSV}$ is outside of the homology arms, the only ES cells that can survive this step are ones that have undergone homologous recombination via the homology arms and excluded the $tk^{HSV}$ gene from the chromosome. ES cells in which the targeting construct has randomly integrated into the genome will include the $tk^{HSV}$ gene and will undergo negative selection (another way we say this is that $tk^{HSV}$-expressing cells will be selected against). 

ES cells that survive this multistep process now have a significant portion of gene $X$ replaced with the $neo^R$ gene - you have knocked out gene $X$ in ES cells (Figs. {{ref>Fig3}} and {{ref>Fig4}}). 

<figure Fig4>
{{ :double_selection.png?400 |}}
<caption> 
Double selection of ES cells with homologous recombination and implantation into foster mothers. 
</caption>
</figure>
===== Creating knockout mice =====

The next step is to create a mouse in which every cell in the mouse contains the genetic alteration you just created in ES cells. The first thing to note is that the ES cells in which we knocked out gene X are from a mouse strain that has gray fur. We will inject these ES cells into a new blastocyst embryo that comes from a mouse strain that has white fur (Fig. {{ref>Fig4}}). Usually, we will do this for a few dozen embryos, which are then implanted into foster mothers. Some pups born from these foster mothers will have both white and gray fur - they are chimeras. We then take male chimeric pups and breed them to white females. Since the gray fur phenotype is dominant, any chimeric father that has germ cells that formed from our modified ES cells will produce pups with all gray fur; these pups are then founders for a knockout mouse line (Fig. {{ref>Fig5}}).

<figure Fig5>
{{ :chimeric_ko_mice.png?400 |}}
<caption> 
Figure 16.5. Generation of chimeric knockout mice.
</caption>
</figure>


===== Creating a mouse model =====
 

We can apply the general strategy for making gene knockouts described above to create mice that contain a knockout allele of the mouse α-globin gene and the mouse β-globin gene. Together with the transgenes for expressing human α-globin and human β-globin in mice, we generated in [[chapter_15|Chapter 15]], we now have a way of creating a humanized mouse - that is, a mouse that only expresses human hemoglobin. See Questions and exercises below.

Some final notes on our story: this humanized mouse did indeed represent an excellent model of sickle cell disease, which was used to explore therapies that are very difficult to carry out on humans. For instance, these mice were used to explore the effectiveness of new drugs in reducing the tendency of RBCs to sickle. Moreover, these mice have been used to test out gene therapy approaches to treating the disease. Although stem cell therapies have rendered some of these treatments obsolete, we use this story to illustrate the power of reverse genetics as a general strategy to study gene function or genetic diseases.

We also note that while this chapter discusses the technique for how to make gene knockouts in mice, the idea of “reverse genetics”, i.e., knocking out a specific gene of interest (as opposed to randomly mutagenizing genes and looking for interesting mutants) is a general strategy for studying gene function. Within the technical limits of different experimental organisms, this general strategy can be applied to just about any commonly studied model genetic organism (such as //E. coli//, yeast, or Drosophila), although the precise technical details of how knockouts can be generated for each organism will of course be somewhat different. A new tool developed around 2012 called CRISPR (briefly discussed in [[chapter_17|Chapter 17]]) has made this gene knockout approach for studying gene function much more accessible to organisms other than the traditional model organisms.  

===== Questions and exercises =====

Conceptual question: How many different genetic parents does the chimeric mouse shown in Fig. {{ref>Fig5}} above have?

Conceptual question: What will the genotype for gene $X$ be for the all-gray founder pups described in Fig. {{ref>Fig5}}?

Exercise 1. (Challenge Question) Assuming that all the mouse genes and transgenes described from [[chapter_16|Chap. 15]] and this chapter are unlinked to each other and that everything is autosomal (i.e., nothing is sex-linked), how could you breed a humanized mouse that only expressed human hemoglobin? You start out with individual knockout and transgenic lines. Let's also assume you have an easy way to genotype everything (usually in these kinds of experiments you would use PCR for genotyping). Keep in mind that the globin genes are essential, and that the knockouts are homozygous lethal. In each step, what is the likelihood of getting progeny that have the genotypes you need?