Blog Section

Mice in HD Research

Huntington’s disease research and drug testing frequently involve the use of mouse models. However, when different scientists refer to an HD mouse model they may not be referencing the same model. A variety of HD mouse models exist and are used regularly. This chapter will describe the most common HD mouse models, how they differ, and the ways in which they mimic the disease.

This article makes reference to both the mouse and the human huntington gene. The mouse huntington gene is a homolog of the human huntington gene, but there are subtle differences between them. For example, the mouse huntington gene has fewer CAG repeats than the normal human huntington gene, explaining why spontaneous mouse models of the disease do not exist. Despite this difference, the mouse huntington gene is 81% similar to the human huntington gene at the DNA level, showing that the two are true homologs. Different mouse models have been generated since the isolation of the HD gene in 1993. There are three general types of mouse model: knockout, transgenic, and knock-in.

Knockout models were the first models to be generated. These are models where the gene coding for the huntingtin protein (the mouse huntington gene) is removed or interrupted so that the DNA can not be transcribed. Although these models provide valuable scientific insights, they do not actually represent the disease because mice that are nullizygous for the huntington gene die during embryogenesis.

Transgenic models are made when the mutant human huntington gene, or a fragment of that gene, is inserted into the nuclei of a model organism. In transgenic models the gene insertion is not targeted to a specific location in the animal genome, meaning that where the gene inserts itself cannot be predicted. For a mouse model, this means that the mouse will express both the two normal copies of the mouse huntington gene as well as the human mutant gene or fragment that was inserted into its genome. It also means that the expression of the inserted mutant human huntington gene will not be controlled by the homologous promoter region of the mouse huntington gene. This frequently leads to much higher protein expression than normal endogenous levels.

Knock-in mice have either part of or the entire human mutant huntington gene inserted in place of part of or the entire endogenous mouse gene. Knock-in mice, therefore, carry the expanded CAG repeat mutation in the same place in the genome that it would appear if it were to develop naturally. This is the most faithful model in the sense that the mutant gene is located in the appropriate genomic context and will have the normal promoter region associated with the huntingtin protein. Knock-in mouse models can be either homozygous or heterozygous for the huntingtin mutation.

To learn more about the process of creating genetically modified model animals click here.

The remainder of this chapter considers the following topics:

More about knock-out mice^

Although knock-out mouse models are not a viable model of HD, studying these early models did increase understanding of the disease. Showing that a complete absence of the huntington gene, and consequently the huntingtin protein, was embryonic lethal revealed that Huntington’s disease is not a loss-of-function disease. The huntingtin protein is necessary for successful embryogenesis and when it is entirely absent the organism will die before birth. This suggests that the mutant huntingtin protein in individuals with HD must fulfill some of the same functions as the normal huntingtin protein.

Since most individuals with HD are heterozygous for the disease, they do still have one copy of the normal huntington gene and can express some levels of the normal huntingtin protein. However, HD homozygotes do develop normally and have an age of disease onset comparable to HD heterozygotes, confirming that the mutant protein fulfills some of the normal huntingtin protein functions. This suggests that HD is a dominant “gain of function” disease as HD patients are able to live with either one or two mutant HD alleles.

Some scientists suggest that despite the finding that HD is a gain-of-function disease, it is still possible that some loss of normal huntingtin protein function contributes to the disease. One study supporting this idea found postnatal brain degeneration in mice in which huntingtin was selectively inactivated in the brain and testes after early embryogenesis. Furthermore, another study found that normal huntingtin, but not mutant huntingtin, increases the production of brain-derived neurotrophic factor, which is necessary for the survival of striatal neurons in the brain. Although both of these studies support the conclusion that loss of normal huntingtin function contributes to the pathogenesis of HD, most researchers believe that HD should be considered a gain-of-function disease.

More about transgenic mice^

The earliest HD transgenic mouse models and still some of the most frequently used are from the line designated R6. In this model, a fragment encoding exon 1 of the human mutant huntingtin gene is inserted into the mouse genome. Four different R6 lines were established, with CAG repeat lengths ranging from 115-156. The R6/2 mouse, one of this line, is probably still the most commonly used HD mouse model. All four lines use the IT15 promoter. In humans it only takes a CAG repeat length of 36 or more to develop HD, but mouse models seem to need much larger repeat lengths to develop an HD-like phenotype, possibly due to their shorter lifespan.

Three of the R6 lines ubiquitously expressed the mutant huntingtin protein and exhibited an abnormal phenotype. The level of mutant huntingtin protein expression varied among these three R6 lines. The level of protein expression is usually represented as a percentage of the normal endogenous mouse huntingtin protein level. The R6/1 line had the lowest expression, measured at about 31% of endogenous mouse huntingtin levels. The R6/2 and R6/5 lines had an average expression of 75% and 77% endogenous mouse huntingtin levels. These lines displayed motor abnormalities such as chorea-like movements, weight loss, seizures, frequent urination, unusual vocalizations, and sudden deaths of unknown cause.

Newer transgenic mouse models now express longer fragments or the full length human huntington gene. One example is the YAC line. Mice from this line express the full length human huntington gene with a CAG repeat length of up to 128. These newer transgenic mouse models tend to show progressive motor abnormalities as well as neuronal loss in the striatum. These symptoms all closely parallel human symptoms in HD, and provide evidence that these mice might accurately model many aspects of the disease.

More about knock-in mice^

The first HD knock-in models did not display overt motor deficits like those observed in the first transgenic mouse models. Instead, they displayed some behavioral abnormalities such as increased aggression. The earliest models usually had between 70-80 CAG repeats. One reason that the early models did not reveal a phenotype resembling HD may be that they had a low number of CAG repeats for mouse models, which tend to have very high CAG repeat lengths.

Later knock-in models were more promising as they did show abnormal motor abnormalities, as well as cellular, molecular and neuropathological abnormalities suggestive of an HD phenotype. They also tended to reveal behavior abnormalities at earlier ages. These later knock-in models tended to have many more CAG repeats than the earlier models (from 94-140 CAG repeats).

Knock-in models are not all the same. Besides differences in CAG repeat length, there are also differences in how much of the mutant huntington gene is inserted and in the background strain of the mouse. In some models, only the sequence coding for the polyglutamine tract in exon 1 was replaced, while in others the entire first exon was replaced. The models that replace all of exon 1 with the mutant version should be more accurate representations of the human condition, because many of the proteins that interact with mutant huntingtin do so in the region immediately surrounding the polyglutamine tract. In other words, when a larger portion of the gene is inserted and it is expressed at an endogenous level, the mouse imitates the genomic context of the huntingtin mutation as well as the mutation itself, and consequently then phenotype may be more consistent with human HD. Most significantly, this means that any results found with these mice, either about relevant proteins, molecular pathways involved, or pre-clinical drug studies, are much more likely to be relevant to human HD.

Why do transgenic mice frequently display a more obvious disease phenotype?^

Transgenic mouse models sometimes seem to display a phenotype more similar to human HD than the phenotypes displayed by knock-in models. This might seem confusing since the knock-in models are theoretically more accurate representations of the disease genotype.

One reason for this unexpected result is that transgenic mice usually only express a fragment of the human mutant huntington gene. The mutation is then expressed as a truncated, or shortened, protein fragment. In vitro studies have founded the truncated protein to be more toxic than the full length protein, which could explain the more severe disease phenotype of these models.

In some transgenic models, this result could also occur because of unnaturally high expression levels of mutant huntingtin protein. Recall from above, for example, that HD line mice sometimes express as much as five times more mutant huntingtin protein than endogenous huntingtin protein. This overexpression could exaggerate the disease phenotype.

How are animal models used?^

Mouse models allow researchers to study aspects of the disease that would be impossible to do with human subjects. For instance, researchers can focus on changes in the brain or in cell-to-cell interactions during early stages of the disease with animal brain tissue. Mouse models are also a crucial part of early testing of potential therapeutics. Because mice have a shorter lifespan than humans, testing can be carried out much faster. Additionally, in the earliest stages of drug development it would be unethical to subject humans to the risks associated with the new drugs being tested.

Unfortunately, the use of mouse models in drug development does have some drawbacks. Mouse models of HD are not exactly the same as the human disease and there are also differences in how a drug will affect one species compared to another. Hopefully, as scientists develop more sophisticated models that better represent the genetics of human HD, model organisms will become more relevant for drug development and therapeutic testing. But it is always vitally important to keep in mind that drugs that look promising in animal models may not prove successful when applied to humans.

Further Reading^

– Adam Hepworth, 11-21-08