August 20, 2014

In previous posts, I have alluded to the fact that studying cancer in a dish (in vitro) is not optimal (here and here). Today, I give you the next essential tool in a cancer stem cell (CSC) researcher’s toolbox: xenotransplantation.

Xenotransplantation is an impressive mouthful that simply means the transplantation of living cells, tissues, or organs from one species to another. The cells of interest have been taken from their original source (donor), and injected into another organism (recipient): in most cancer stem cell research, we’re taking human tumour cells, and injecting them into a mouse (Figure 1). The result is referred to as a xenograft. For those of you wondering, yes, this is what CSC researchers consider an ‘in vivo’ assay (experiment).

Figure 1:  Cancer stem cell researchers use xenotransplantation as in vivo models of human cancers. Human tumour cells are grown in a dish, and then injected (transplanted) into mice. The resulting mouse with a human tumour is called a xenograft.

What is a CSC researcher trying to accomplish by using this technique? By examining tumour growth of a particular sample in mice, researchers can ultimately observe what is called the “tumour-initiating capacity” of that sample, or in other words, whether or not those transplanted cells can form new tumours. We can then define a population of CSCs in a dish as ‘tumour-initiating cells’ (TICs), and present a more relevant model for studying human tumorigenesis.

Table 1: A limiting dilution experiment can be used to estimate tumour-initiating cell (TIC) frequency of different samples. Decreasing numbers of tumour cells are injected into mice. The fewer cells required for tumour growth, the higher the frequency of TICs. In this example, Sample A (red) would have a higher TIC frequency (i.e. less cells are needed to form a tumour) than Sample B (blue).

You’ve probably noticed that the main premise of xenotransplantation is very similar to that of a sphere formation assay (an in vitro method): identifying cancer cells responsible for initiating tumour growth. As such, CSC researchers tend to look for similar things with this in vivo assay. For example, researchers can perform a ‘limiting dilution’ in effort to determine the minimum cells needed to initiate tumour growth (Table 1), allowing them to calculate the frequency of TICs. A serial transplantation (like serial sphere formation experiments, as described in this post’s Figure 2) can also be done, in order to identify cells capable of initiating tumour growth over several generations (Figure 2). In fact, the ability of a cell (or group of cells) to do this is considered the gold standard of TIC identification by CSC researchers. An important tool indeed!

Figure 2: Serial injection of tumour cells is the gold standard of tumour-initiating cell (TIC) assessment. The first mouse is injected with human tumour cells. Once a tumour has formed, those tumour cells are then injected into a new set of mice. If a tumour forms in the second mouse (top), it is then injected into a third mouse, and so on. The continued ability to do this would indicate the presence of TICs. If no tumour forms in the second mouse (bottom), no more injections can be done, suggesting the absence of TICs.

Aside from assessing the TICs of various tumour samples, xenotransplantation is also important for drug and genetic studies. Researchers will measure things like tumour growth rates, final tumour size, metastatic spread of disease, survival, and biomarkers in the context of different drug treatments or genetic status.

As with most experimental models, there are several important conditions and caveats associated with xenotransplantation:

  1. Researchers need to use immunocompromised (i.e. impaired or absent immune system) mice in order for human cells to grow in the mouse. A normal mouse would have an immune response and kill off the human cells, preventing tumour growth and any experimental assessment. The drawback here is that it is difficult to assess the interplay between the immune system and cancer cells, which is extremely important in tumorigenesis.
  2. CSC researchers prefer orthotopic injections – meaning that the human cells are injected directly into the site of origin in the mouse (rather than under the skin, or into the blood stream), see Figure 3. These kinds of injections are thought to be more relevant models of tumour growth, since they are being injected back into their ‘natural’ environment (niche). Intravenous (blood stream) injections are still useful, as they can allow for the assessment of ‘homing’ properties – do these cells have a place they prefer to go and grow?Figure 3: Orthotopic injection models are used to observe tumour growth in its ‘natural’ environment (niche). Some commonly used orthotopic injection models are shown here.
  3. A mouse is not a human! Researchers always need to keep this in mind when applying observations made in mice to human disease. For example, during serial injections (Figure 2), are we selecting for TICs, or cells that are just really good at growing in a mouse? Some have suggested developing ‘humanized’ mouse models to overcome some of these issues. In this kind of model, human ‘support’ cells (e.g. fibroblasts) are injected prior to the tumour cells of interest. They are allowed to grow and develop a human-like environment into which tumour cells can be injected. While a potentially more accurate representation of human disease, this method is very time and resource intensive.
  4. The components of the model can influence the observations and subsequent conclusions! In 2008, Sean Morrison and colleagues brought a striking example of this to light. By changing the level of immune system impairment, and how the cells were injected, this group found a 1 in 4 TIC frequency in melanoma compared to 1 in 46,700 with historical conditions (in-study comparison), begging the question: which is the true TIC frequency? Not only did this shake the foundations of the CSC research community, but it also highlighted the importance of recognizing how components of a model system may artificially influence the behaviour of tumour cells outside of the human body.

I also want to point out the importance of understanding that animal experiments are required in order to move discoveries made in vitro forward towards clinical application in humans. All of these experimental techniques must first be approved by an animal ethics board, which ensures the animals are being treated humanely and not used unnecessarily.

While in vivo of modeling cancer certainly overcomes many of the limitations associated with in vitro techniques, as we’ve seen, it is not without its own set of caveats and limitations. Because of this, CSC researchers rarely use one kind of model system. In fact, the CSC tools that I have described to you are typically used together, in a hierarchy of sorts: in vitro (e.g. sphere formation) to first test a hypothesis, followed by in vivo experiments (e.g. xenotransplantation) to confirm the results in a more complex environment. These in vitro and in vivo assays are the foundations of CSC research, and can be used in a variety of ways to best answer research questions of interest. They are truly essential parts of the CSC researcher toolbox!


1 Comment

  1. Eric Cummer says:

    In search of the perfect model: Thank you for casting light upon this perplexing challenge. Eric Cummer, Toronto.

Post a Comment