Genetics, Pathology, & Potential Future Advances in the Study of Huntington's Disease
IN THIS ARTICLE
Huntington’s disease is a progressive neurodegenerative disorder that affects around five people in every 100,000. It is caused by an increase in a polyglutamine region of the Huntingtin protein, resulting in a toxic gain of function mutation. However, the exact mechanism of pathogenesis, or even the exact role of the normal Huntingtin protein are not comprehensively understood. It is known that Huntingtin is ubiquitously expressed and has interactions with many different cellular components. Mutant Huntingtin has been extensively studied and its effects on cells noted but questions regarding its toxicity remain to be answered. Much research has been directed towards finding treatments for the condition and while some show progress, the scientific community is still some way away from a definitive therapy to slow or prevent the disease. This review will critically assess research that has been conducted into Huntington’s disease and propose possible future work.
Huntington’s disease (HD) is an autosomal dominant repeat disorder with complete penetrance and prevalence of around 4-8 per 100,000 (Harper, 1992). It is an expansion disorder, caused by an increase in the number of CAG (glutamine) repeats in the coding region of the gene for Huntingtin (Htt) protein on chromosome 4p16.3. This large polyglutamine (polyQ) expanse results in the production of mutant Huntingtin (muHtt), in a gain of function dynamic mutation (Landles & Bates, 2004). This new protein is pathogenic and causes neurodegeneration, leading to cognitive decline and progressive chorea.
The condition was first described by George Huntington (Huntington, 1872). He noticed that there was a high prevalence of the then unexplained condition amongst certain families- up to this point, sufferers had been thought to be possessed due to the chorea. In 1911, further research by Charles Davenport confirmed the genetic link and showed it was an autosomal dominant disorder (Davenport, 1911). A major breakthrough came in 1993 when the gene responsible was located at 4p16.3 (Huntington's Disease Collaborative Research Project, 1993). This paper showed where the mutation occurred, providing an insight into how the disease manifested in patients.
Non-Huntington’s sufferers have 36 or fewer repeats of CAG in their Htt gene, resulting in the production of normal cytoplasmic Htt (Strachan & Read, 2011). Once the polyQ region exceeds 36 repeats, it becomes unstable and the protein produced has different properties. Generally, a repeat length of between 36-39 will result in a delayed onset form of the disease, with less severe symptoms and a slower progression. Repeats of 40 or more CAGs are associated with almost complete penetrance by the age of 65 years (Langbehn, Brinkman, Falush, Paulsen, & Hayden, 2004).
2. Outline of Huntington’s Disease
2.1 Clinical Description
Huntington’s disease is characterised by both motor and cognitive disorders of varying severity, which usually onset between the ages of 30 and 50 (Roos, 2010). The disease progresses for around 17-20 years and is eventually fatal. The most common cause of death due to Huntington’s is pneumonia, followed by cardiovascular disease (Heemskerk & Roos, 2012).
Motor symptoms include progressive chorea- involuntary, jerky movements. These initially begin in the extremities and gradually progress throughout the body, occurring when the sufferer is awake (Roos, 2010). The most common choreatic movement involves extension of the lower back muscles. As well as the chorea, the neuronal decline causes a loss of balance, leading to unstable walking. Sufferers are also affected by bradykinesia and akinesia. Dystonia is one of the first presenting symptoms and results in an abnormal posture presenting. Ultimately, these symptoms increase in severity and in the end stages of the disease patients struggle to stand and walk, resulting in them needing increasing levels of care.
Psychiatric disorders onset very early in the disease, usually before motor symptoms manifest. Depression is the most common, followed by guilt and anxiety (Julien, et al., 2007). Because of how early these symptoms appear (up to 20 years before motor symptoms in some cases, (Rosenblatt, 2007)) and the strong similarities in psychiatric symptoms in families (Lovestone, Hodgson, Sham, Differ, & Levy, 1996), it was initially unclear as to whether they were attributed to physical changes in the brain or simply genetics, somehow related to being at risk of the disease but with no neurodegeneration. A study by Julien et al. (2007) investigated this and found that lifetime psychiatric symptom presentation was no different in those with or without the mutation. The study did however find an increased incidence of depression amongst HD sufferers. Therefore, while psychiatric disturbances seem to have no relation to carrying the mutant gene, the HD gene confers an above average risk of depression.
Cognitive decline is a final major progressive symptom. This too can be present before motor signs and varies in its severity regardless of the overall progression of the disease. The decline is in ‘executive functions’, which are required to distinguish between relevant and useless information and to be able to organise and plan the patient’s life. Speech is generally unaffected, as is semantic memory, but other memory functions are gradually lost.
Unexplained weight loss is present in all patients from an early period. While it was first assumed this was due to the chorea, this has been shown to be untrue (Roos, 2010). It is more likely that the weight loss is due to the neuronal changes resulting in decreased appetite, slower functioning and difficulty with preparing and eating food. Recent research has shown that the longer the CAG repeat length, the faster the rate of weight loss in a patient (Aziz, et al., 2008) and has also associated hypothalamic neuronal loss and the subsequent metabolic and endocrine changes with a loss in body mass (Aziz, Swaab, Pjil, & Roos, 2007).
A final, secondary symptom is circadian rhythm disturbances. Human patients show disrupted sleep patterns, which have knock-on effects on their wellbeing. Mouse models of the disease show a progressive shift from nocturnal to daytime activity, eventually leading to a complete loss of circadian rhythm (Morton, et al., 2005). This symptom is due to a morphological change in the suprachiasmatic nucleus of the hypothalamus (responsible for maintaining circadian rhythm); specifically, a large reduction in both vasoactive intestinal polypeptide neurons (85%) and a reduction in arginine vasopressin neurons (33%) (van Wamelen, et al., 2013). These are key circadian rhythm neurotransmitters and mRNA levels for them were unchanged, showing the problem lies with their release, not synthesis.
2.2 Juvenile Onset Huntington’s Disease
A second, less common form of the condition is juvenile onset, defined as presenting before the age of 20. This accounts for ~10% of cases and is associated with a much larger than average repeat length (Walker, 2007). The key symptomatic difference commonly involves an absence of chorea. Instead, the sufferers exhibit rigidity. The disease also progresses much more quickly and is associated with a high incidence of seizures (Walker, 2007). Because this condition can begin from as early as two years and repeat length increases through generations, children of parents with the disease may display symptoms before them.
2.3 Method of Elongation
It is thought that the repeat sequence becomes elongated during DNA replication. Lagging strands of DNA are synthesised as a number of overlapping Okazaki fragments. Because of this, a specialised enzyme called FEN1 removes the overhang, allowing the fragments to correctly anneal. A mutated enzyme could join overlapping fragments, leading to the repeat expansion (Strachan & Read, 2011). Other research suggests that instead of occurring during DNA replication, it is the base excision repair system, involved in fixing single strand breaks due to oxidative damage, that causes expansion. The BER system can generate single strand overlaps which may then form repeats (Fig. 1) (McMurray, 2010). In human HD fibroblasts, which usually show no expansion when cultured in vitro, expansion can be induced by peroxide exposure, which damages DNA by oxidation (Kovtun, et al., 2007). This study also showed an in vivo escalating oxidation/excision system, which explains the age-dependant expansion seen through generations of sufferers of HD.
Figure 1. Showing the BER system. Oxidised, damaged DNA (1) is targeted by the BER enzyme complex (2), excising a single base pair gap which is then lengthened. The new strand is then synthesised (3). As shown in (4), strand displacement occurs during this process, forming a flap. This flap cannot be removed by FEN1 as commonly happens due to the stabilisation of the flap into a trinucleotide repeat hairpin (5). As shown by the red dot in the hairpin, A-A mismatches further stabilise the structure as the mismatch repair enzymes MSH2 and 3 bind. This hairpin may then be incorporated into the DNA (7) and this repeat process eventually leads to a ‘toxic oxidation cycle’ (8), where the repeat sequence progressively expands. Figure from (McMurray, 2010).
A final mechanism lies in the nucleotide excision repair system, consisting of two subsystems; global genome repair (GGR) and transcription-coupled repair (TCR). GGR is unlikely to be involved, as loss of the system in knockout mice has little effect on HD progression (Dragileva, et al., 2009). TCR involvement is more likely, as TCR knockout mice show a slower progression of contraction within repeat length (contractions are associated with a more instable repeat) (Lin & Wilson, Trascription induced CAG repeat contraction in human cells is mediated in part by transcription coupled nucleotide excision repair, 2007). However, the loops formed in this system are too small to form repeat sequences, meaning it is more likely that the BER system is involved in the CAG repeat expansion seen in HD.
2.4 Familial Inheritance and Anticipation
A child of a heterozygous sufferer of HD has a 50% chance of inheriting the disease. HD was the first condition for which premanifest diagnosis information could be obtained, allowing for genetic counselling of potential sufferers and families expecting children. For genetic counselling, as long as the applicant meets none of the defined exclusion criteria, an initial consultation followed by a blood test 4-6 weeks later is available (International Huntington Association, 1994). For prenatal diagnosis, chorionic villi sampling can take place between the 10th and 12th week of pregnancy, as long as the parents are aware of their own genetic status. More recently, in vitro fertilisation has become involved. At the eight cell stage, one cell is removed and tested for HD and only embryos that do not have the disease will be implanted (Roos, 2010). Genetic testing is relatively simple, as the only diagnostic marker analysed is the length of the CAG repeat in the Htt gene.
Interestingly, despite the efficacy of the test, fewer than 5% of at risk individuals are actually tested (Walker, 2007). This is attributed to the lack of an effective treatment- those who undergo testing are generally making family or career choices. Around 1% of those who are diagnosed as sufferers attempt suicide (Almqvist, Bloch, Brinkman, & Crauford, 1999), meaning the exclusion criteria must be rigorously examined before testing and those who test positively must be adequately educated to prevent them being misled about the disease.
Huntington’s disease shows a genetic phenomenon called anticipation, a process where the age of onset in children is earlier than that of the parents (Strachan & Read, 2011). This earlier onset is usually accompanied by an increase in the severity of symptoms. A study by Ridley et al. (1988) found that generally the change in age of onset is equal regardless of whether the disease was inherited from the mother or father but in a small proportion of paternal inheritance cases (6%), the age of onset is lowered by around 20 years. This specific pedigree was associated with early motor rather than psychiatric symptoms. They further postulated that the reason for the small number of sufferers of this subset was due to the fact that constantly reducing the age of onset would eventually lead to such an early disease progression that the sufferer would be unable to reproduce, preventing the line continuing. Anticipation means it is essential that children of sufferers of the disease are tested as early as possible, in order to begin counselling and treatments.
3. The Huntingtin Protein
The IT15 (interesting transcript 15) gene located on chromosome 4p16.3 codes for the Huntingtin protein (Huntington's Disease Collaborative Research Project, 1993). Huntingtin itself is a large, 350kDa 3000 amino acid protein. It is mainly structured of large, 50 amino acid long repeats, termed HEAT repeats (from Huntingtin, Elongation factor 3, protein phosphatase 2A, TOR kinase) (Andrade & Bork, 1995). These repeats are found in a number of different proteins and form hairpin helical structures. The repeat units then assemble into a superhelix with a continuous hydrophobic core seen in figure 2 (Li, Serpell, Carter, Ruiensztein, & Huntington, 2006).
Before continuing, an interesting point to note is that as the function of Htt is currently only theorised, there is no function-based assay available. This means studies that claim to have purified Htt cannot be sure that this is actually the material they have obtained.
The N terminus of Htt contains the crucial polyQ region. When this is expanded beyond the safe number of 36 repeats, the resulting conformational change causes multiple knock-on effects, including the formation of truncated protein (Ross & Tabrizi, 2011). To fully understand the effects of the mutant protein, knowledge of the role of wild type Huntingtin (wtHtt) is essential.
Figure 2. HEAT repeats. A shows a single helix turn helix HEAT repeat (helical hairpin). Both the helices (blue) and the coil regions (red) can vary in length. B shows the hydrophobic residues (green) of this structure. C shows a 2880-residue protein consisting of HEAT repeats, which can be assumed to be similar to the main structure of Htt. The superhelix structure is clearly visible. The study by Li et al (2006) that created this protein pointed out that as every red region in this protein was flexible, this showed why Htt was susceptible to proteolysis along its entire length. Figure from (Li, Serpell, Carter, Ruiensztein, & Huntington, 2006).
3.1 The Roles of Wild Type Huntingtin Protein- Development
Much research has been conducted into the cellular function of Htt. Generally, when a new protein is discovered, it is sequenced to look for recognisable domains (e.g. ATP binding), which would provide some clue to function. However, when Htt was first analysed, it became clear that it had no recognisable domains (Landles & Bates, 2004). As well as this, the protein is expressed in many different subcellular compartments.
Research into the protein’s function has shown that it is both complex and important. A study by Rigamonti et al. (2000) found that wild type Huntingtin protected CNS neurons from apoptotic stimuli such as death receptor activation. Apoptotic cellular death is controlled by caspase proteins and wtHtt acted as a caspase-3 inhibitor in some mechanism upstream of its activation pathway. It was also shown that this protective effect required the portion of the protein between amino acid 63-548 to be present (Rigamonti, et al., 2000). Huntingtin is crucial for early embryonic development, as knock-out mouse models for their Huntingtin homolog protein (Hdh) show embryonic lethality within 8 days (Duyao, et al., 1995). This is attributed to increased apoptosis (Riener, Dragatsis, Zeitlin, & Goldowitz, 2003), linking to the protective role of Htt.
In hypomorphic Hdh mutations (protein reduced to around 1/3 of normal levels), early embryonic lethality is avoided but is replaced by severe developmental difficulties. These include defective neuron development and macroscopic malformations of the brain (White, et al., 1997). Interestingly, these characteristics are not shown in mice that express normal levels of the mutant form of Htt, indicating muHtt does not lose its neurogenic function. This confirms that HD is a gain of function mutation.
Due to the early embryonic death of Htt knockout mice, study of the long term effect of a lack of the protein is difficult. Another model organism that can be investigated is Drosophilia melangosta which also has an Htt homolog, known as dHtt (Li, Karlovich, Fish, Scott, & Myers, 1999). Unlike mouse models, dHtt knock out flies survive gestation and are born as normal. They also show a normal rate of development and no morphological differences but have reductions in mobility and fertility later in life (Zhang, Feany, Saraswati, Littleton, & Perrimon, 2009). The reason for the relatively normal development has been hypothesised to either be down to a compensatory mechanism, or defects that do not appear under normal conditions and would instead become apparent when the flies were stressed (Zheng & Joinnides, 2009).
3.2 The Roles of Wild Type Huntingtin Protein- Cellular Transport
Huntingtin has roles in maintaining the function of the cytoskeleton, which are reliant on interactions with HAP1 (Huntingtin-associated protein), which itself interacts with proteins such as kinesin and dynein (microtubule transporters, opposite directions) and dynactin (responsible for binding dynein and kinesin to the molecule requiring transport) (Caviston & Holzbaur, Huntingtin as an essential integrator of intracellular vesicle trafficking, 2009).
There are three possible models for the role of Htt in transport. It could be a switch, with phosphorylation of the protein acting to change transport direction along microtubules. Colin et al. (2008) showed serine phosphorylation at position 421 resulted in increased amounts of anterograde transport along cellular microtubules, whereas the unphosphorylated residue was associated with retrograde transport. The exact reason for this is unknown, but it is likely that the phosphorylated Htt protein has an increased affinity for acting as a scaffold protein between kinesin and dynactin (due to the direction kinesin travels in), whereas the dephosphorylated version enhances interaction between dynein and dynactin.
The second model of Htt in transport involves HAP40. When associated with Htt, HAP40 has been shown to affect whether vesicles are associated with actin filaments, used for short-range transport, or microtubules, used for long range transport (Pal, Severin, Lommer, Shevchenko, & Zerial, 2006). Htt/HAP40 complexes increase the association of vesicles with actin and as HAP40 is upregulated in HD (Pal, Severin, Lommer, Shevchenko, & Zerial, 2006), this could suggest a disease mechanism- defective long range cellular transport as a result of a greater than normal vesicle actin. Aberrant transcription of HAP40 may be why it is upregulated in HD.
The final potential model is that Htt only interacts with dynein and depletion of wtHtt results in defective and bi-directional dynein-bound vesicle transport, along with detachment of vesicles from transporters (Caviston, Ross, Antony, Tokito, & Holzbaur, 2007). This would result in cellular recycling pathways becoming compromised, potentially leading to a build-up of toxic material or depletion of crucial cellular components, again suggesting a role in the pathogenesis of HD.
Overall, there is abundant evidence that Htt is implicated in cellular transport. It is very likely that all three of the above models are correct and Htt is intimately involved in all stages of the transport process; forming a complex to bind vesicles to transporters, selecting the direction of travel and then detaching the vesicle at the target site. This complexity would explain why lowered levels of wtHtt have such a significant effect on the body.
3.3 The Roles of Wild Type Huntingtin Protein- Transcription
Huntingtin has been shown in multiple studies to have direct interactions with transcription factors and transcription co-activators such as NFκB, p53 and CBP (Stefan, et al., 2000)(Takano & Gusella, 2002). As well as this, wtHtt interacts with the transcription activator Sp1, which itself is essential for recruitment of TFIID to the nucleus (Dunaw, et al., 2002). TFIID is a complex of proteins including TATA binding protein, used to bind to the TATA sequence present in the promoter sequences of certain eukaryotic genes. wtHtt binds to both Sp1 and TFII130, a component of TFIID, once again suggesting a role for Htt as a scaffold protein.
Other research has suggested that wtHtt is a transcription repressor. Htt binds to the N-CoR and Sin-3a receptor complex, used to prevent transcription of certain genes targeted by ligand activated nuclear receptors, such as retinoic acid receptor (Boutell, et al., 1999). It is clear Htt is involved in gene transcription, possibly promoting the synthesis of some genes and preventing synthesis of others to have an unknown combined effect.
The most interesting role of Htt in transcription involves the production of a molecule called BDNF (brain derived neurotrophic factor). This acts on neurons to promote growth, survival and differentiation, forming new synapses (Huang & Reichardt, 2001). The BDNF gene is silenced by a molecule called REST/NRSF (RE1 silencing transcription factor/neuron restrictive silencer factor). REST/NRSF binds to the neuron restrictive silencing element (NRSE) in the BDNF promoter, recruiting a protein complex to inhibit synthesis (Fig. 3). Htt binds to this molecule, sequestering it (Zuccato, et al., 2003). Resultantly, levels of BDNF synthesis are increased and neuron development is stimulated.
Figure 3. The REST/NRSF system. In the case of wtHtt (A), REST/NRSF is bound by the Htt molecule. This prevents it from binding to the NRSE, thereby preventing formation of the repressor complex, allowing synthesis of the BDNF gene. In the case of HD (B), muHtt does not effectively bind REST/NRSF, causing it to accumulate in the nucleus, repressing transcription of the BDNF gene. Figure adapted from (Cattaneo, Zuccato, & Tartari, 2005)
This suggests a crucial pathogenic role of muHtt- as there is such catastrophic neuron loss in disease sufferers and this molecule is responsible for healthy neuron survival, clearly there is some link. Heterozygous Hdh +/- knockout mice show reduced levels of BDNF mRNA (Zuccato, et al., 2003) and homozygous -/- Hdh knockout mice show a greatly reduced level of BDNF mRNA, along with progressive neurodegeneration and sterility (Dragatsis, Levine, & Zeitlin, 2000). Further to this, it has been shown that muHtt does not bind NRSF effectively (Zuccato, et al., 2003). This leads to the accumulation of NRSF in the nucleus, repressing BDNF synthesis. Therefore, both a lack of wtHtt and defective binding ability of muHtt are responsible for a loss of BDNF, a key factor in development of HD.Continued on Next Page »