We Found the Gene! Huntington’s Disease After the Cheering

In 1872, George Huntington, a third-generation physician on Long Island, penned a vivid, still accurate description of a disorder that he, his father, and his grandfather had observed in their patients. Huntington identified the disease as a type of chorea, a nervous disorder marked by incessant, uncontrollable muscle twitches and sometimes called St. Vitus’s Dance (“chorea” comes from the same root as “choreography”). This, however, was chorea with a difference:

It is attended generally by all the symptoms of common chorea, only in an aggravated degree, hardly ever manifesting itself until adult or middle life, and then coming on gradually but surely, increasing by degrees, and often occupying years in its development, until the hapless sufferer is but a quivering wreck of his former self.

Huntington not only described the clinical features and characteristic adult onset of the disorder, but also its transmission within families, and the effect it has on them:

The hereditary chorea, as I shall call it, is confined to certain and fortunately a few families, and has been transmitted to them, an heirloom from generations away back in the dim past. It is spoken of by those in whose veins the seeds of the disease are known to exist, with a kind of horror, and not at all alluded to except through dire necessity, when it is mentioned as “that disorder.”

Huntington also noted the “tendency to insanity and suicide” in sufferers: “As the disease progresses the mind becomes more or less impaired, in many amounting to insanity, while in others mind and body gradually fail until death relieves them of their sufferings.” Indeed the behavioral and cognitive symptoms are usually much more devastating to the patient and family than the movement disorder, and are the reason the disease is no longer called simply “Huntington’s chorea.”

Huntington wrote in the predawn of genetics, shortly before Gregor Mendel published his first description of inheritance patterns in peas. This did not prevent Huntington from rendering an exact account of its inheritance:

When either or both the parents have shown manifestations of the disease, and more especially when these manifestations have been of a serious nature, one or more of the offspring almost invariably suffer from the disease if they live to adult age But if by any chance these children go through life without it, the thread is broken and the grandchildren and the great-grandchildren of the original shakers may rest assured that they are free from the disease… Unstable and whimsical as the disease may be in other respects, in this it is firm, it never skips a generation to again manifest itself in another; once having yielded its claims, it never regains them.

Huntington could not know that he was observing only the most recent in a succession of victims stretching back, on the East Coast of the United States, to just two ancestors, born in Suffolk, England, who had emigrated to Salem, Massachusetts, in 1630. For more than 300 years, the disease has manifested itself in each of the 12 generations of these families.

Finally, Huntington portrayed the inexorability of disease progression in a description as accurate today as 130 years ago:

I have never known a recovery or even an amelioration of symptoms in this form of chorea; when once it begins it clings to the bitter end. No treatment seems to be of any avail, and indeed nowadays its end is so well known to the sufferer and his friends, that medical advice is seldom sought. It seems… to be one of the incurables.

As it was in Huntington’s era, so it remains now: for a patient who inherits the Huntington’s gene, disease progression is foreordained, relentless, and entirely untreatable.

At the same time, this simple causation and stark prognosis make it one of the most tantalizing of all disorders for those who hope to understand and cure neurodegenerative diseases. The current grim lack of Huntington’s disease treatment is countered by optimism about the future: most people working in the field believe the knowledge gained from a decade of research on the gene will—eventually—lead to rational treatment aimed directly at the gene or gene product and, at the same time, provide crucial insights into the pathogenesis and treatment of other, more common neurodegenerative diseases, including Alzheimer’s and Parkinson’s diseases.

More than for any other neurodegenerative disorder, the tools of molecular genetics have led to fundamental discoveries about the Huntington’s disease gene and its protein product. Several excellent transgenic animal models have been developed that show cellular and behavioral pathology matching the human disease. One model has even suggested that the symptoms of Huntington’s disease may be reversible— that patients may be able to get better. But, while these models have led to an ever-deeper understanding of the pathogenic cascade that leads to neurodegeneration and clues to developing treatments, no successful treatment has yet emerged.

Two decades ago, in the earliest days of gene discovery, finding the gene for a disease was assumed, at least by many on the periphery of research, to be the biggest hurdle on the road to treatment and cure. The example of Huntington’s disease has shown in retrospect how naive this view was, and how long, tortuous, and frustrating that road can be. Why didn’t the discovery of the Huntington’s disease gene rapidly lead to treatments? Is rational, effective therapy really on the horizon for Huntington’s disease? Or is that tantalizing prospect a mirage?

The Genetics: One Gene, One Disease

To understand the genetics of Huntington’s disease, begin by recalling that each of us possesses two copies of every gene, termed alleles—one inherited from each parent—that are present on pairs of homologous chromosomes. This combination of genes, called our genotype, controls the development of our particular form of the trait, our phenotype. Huntington’s disease is due to a single aberrant allele, which displays the autosomal dominant pattern of inheritance. It is “autosomal” because it is not on either sex chromosome and thus neither its inheritance nor its expression are sex-dependent, and “dominant” because possession of a single copy is enough to ensure development of the disease. A child of a person with Huntington’s disease stands a 50 percent chance of inheriting the culprit allele and, if he does, will inevitably develop Huntington’s disease.

While the original language of “dominance” and its opposite, “recessiveness,” implied a struggle between the two alleles, advances in understanding molecular genetics through the twentieth century made it clear that alleles themselves do not interact in the nucleus of the cell to determine which will control the phenotype. Rather, in most cases, both alleles are available for transcription (“reading” by the cell machinery), and both are transcribed to make the protein they encode.

In some cases an allele is so defective it cannot be transcribed, or its protein product cannot function and is destroyed. This defect is known as a “loss-of-function” mutation. Usually a cell can compensate for this loss through excess production from the good allele or other means, and so the defective allele remains phenotypically silent—it is recessive.

In other cases, an allele is altered so that it or its protein is no longer subject to normal regulatory control, or the protein interacts with new partners in the cell, or interferes with other cellular systems. This type of defect is known as a toxic gain-offunction mutation, and it typically acts in a dominant manner.

By the 1970s, scientists were confident the Huntington’s disease gene had a toxic gain-of-function mutation, but little else was clear. Nothing was known about the normal gene or the function of the normal protein it presumably encoded. Nothing was known about the location of the gene on the chromosomes, nor about the molecular nature of the mutation or how it caused the inexorable progression of Huntington’s disease. Answers to all of these questions awaited the development of the tools of recombinant DNA, tools that first appeared in the late 1970s and continue to be developed and improved today.

The Hunt for the Gene

Among the earliest recombinant DNA techniques were those used to locate genes on chromosomes. The tools were a set of polymorphic DNA “markers,” spaced across the 46 human chromosomes, and a set of powerful DNA-cutting enzymes. The markers are not themselves disease genes; they are noncoding DNA regions that come in a variety of forms (polymorphisms) that vary from person to person. The enzymes recognize and cleave precise DNA sequences. If two forms of a marker differ in sequence at a cutting site, one will be cleaved and the other will not, generating different-sized sets of DNA fragments. Visualizing these differences allows a researcher to determine which form of the marker a particular person has inherited. By matching the inheritance of a specific form of a marker with the inheritance of a disease, a researcher could hope to determine which chromosome the disease gene was carried on, and approximately where on the chromosome it lay.

Today this feat is routine (and completely automated), but 20 years ago it represented the cutting edge of genetic technology and involved difficult and painstaking labors by a small army of researchers working at the lab bench. When James Gusella and his colleagues at Massachusetts General Hospital set out to find a marker that co-inherited with Huntington’s disease, they had no idea where to begin, and faced potentially years of randomly testing hundreds of markers. They could not even look back on the success of others to encourage them, because they were the first group to try using polymorphic markers to find a disease gene.

As it turned out, Gusella’s group was lucky; they found their Huntington’s disease marker on only the 12th try. The marker, and therefore the gene, was found somewhere near the tip of the short arm of chromosome 4. The discovery, announced in 1983, was greeted with enormous enthusiasm and with optimism that the gene itself would quickly be identified. As a report in Science put it at the time, “What the finding means is that scientists now know exactly where to look for the gene. It’s only a matter of time until they isolate it.”

But the tools and techniques that would allow researchers to locate the gene more precisely were still being developed in the early 1980s, and it was a full decade before the gene was finally found.

Meanwhile, the original marker, and others discovered later, at least allowed genetic testing for gene inheritance in families with the disease. Along with its many benefits, this opportunity was also fraught with psychological peril. In the absence of effective treatments, many people at risk continue to avoid testing at least until they must make choices about having children.

From Gene to Protein

The long wait finally ended in March of 1993 with the announcement by Gusella and 58 co-authors on two continents that they had found the Huntington’s gene and identified the mutation responsible for the disease. The gene was a newcomer to science, not some already familiar player suddenly unmasked. Nothing about the sequence of its 10,000 nucleotides (that is, its particular lineup of DNA building blocks) suggested what its protein product might normally do in the brain or the body.

The mutation, however, was a familiar type; it was a CAG trinucleotide repeat, a run of cytosine-adenine-guanine triplets inserted into the gene. DNA is a long string of nucleotides that come in four types (the fourth is thymine), whose sequence dictates the order of amino acid building blocks in proteins. CAG codes for the amino acid glutamine, and the mutant protein was predicted therefore to contain a polyglutamine tract. While the normal allele usually has only a handful of these triplets (and never more than 35), the disease allele has from 37 to 100 of them. Huntington’s disease thus joined Fragile X syndrome, myotonic dystrophy, and spinobulbar muscular atrophy in the category of polyglutamine diseases. (The list has since grown to nine in all, with more almost certain to be discovered.)

But with every new discovery came new questions, many still unanswered. One of the most basic questions is what is the role of normal, unmutated huntingtin? It is clearly required at least during development, as “knocking out” the version of the gene found in mice causes early embryonic death. Beyond that, almost nothing is known, although several interacting partners of huntingtin have been discovered and recent work suggests it may have a role in vesicle trafficking, the movement of small sacs of materials within the cell.

Given the intensity of the research effort, the lack of knowledge of huntingtin’s normal function may seem surprising. There are good reasons for it. First, the basic task is very hard and the tools available are very crude. Determining the normal role of a protein in a cell using standard techniques is somewhat akin to determining the normal function of a sparkplug in an engine—after running the car through a junkyard crusher.

Second, and perhaps more centrally, very few researchers believe the key to Huntington’s disease pathogenesis lies in understanding huntingtin’s normal role, and so comparatively little effort has been devoted to it. Instead, attention has focused on the toxic action of the mutated protein, a toxicity thought to derive from some completely new interaction, almost certainly involving the polyglutamine tract encoded by the gene’s extra CAGs.

Why Is Polyglutamine Toxic?

The evidence supporting polyglutamine toxicity is strong. Perhaps most convincing is the sheer number of polyglutamine diseases. While each is due to a mutation in a different gene, all lead to neurodegeneration, albeit of different subsets of neurons (In Huntington’s disease, the basal ganglia and cortex are most significantly affected, leading to the characteristic effects on movement and cognition). Additional evidence is found in the direct relation between the number of trinucleotide repeats and disease severity, and the inverse relation between repeat number and age of onset: Longer repeats cause worse disease, and at a younger age.

How polyglutamine tracts cause cell death is the central question of Huntington’s disease, one to which hundreds of researchers have devoted millions of hours over the past decade. At least half a dozen strong hypotheses have been promoted; most still have champions and good, as-yetunrefuted (if incomplete) evidence to support them.

The earliest hypothesis to generate excitement came in 1996, when a group at Duke University showed that mutant huntingtin bound itself to a central enzyme involved in the cell’s production of energy, GAPDH (glyceraldehyde phosphate dehydrogenase). Furthermore, another expanded polyglutamine protein, from an even rarer disease known as DRPLA (dentatorubro-pallidoluysian atrophy), also bound to GAPDH. This immediately suggested that a defect in energy production underlies pathology in both diseases and that perhaps cells were dying either directly or indirectly from lack of energy. In its favor, this hypothesis explained why brain cells were so susceptible: While other body cells can use alternative enzyme pathways to metabolize fats, brain cells feed only on glucose and have an unforgiving requirement for GAPDH.

Despite its appeal, the hypothesis soon fell out of favor. Researchers discovered that while energy production was indeed impaired in affected neurons, it did not appear to involve GAPDH function, but rather the mitochondria (more about this later). Furthermore, mutant huntingtin was found to interact with dozens of other cell proteins, at least in the test tube, weakening the case for a special effect mediated through this single pathway.

Enter the Animals; Cue the Inclusions

Even as the GAPDH story evolved, powerful new tools from molecular biology were allowing even finer manipulation of the Huntington’s gene. Transgenics—the introduction of a gene from one organism into the DNA of another—changed forever the way genetic research was done. Before transgenics, studying human disease in an animal required either finding a pre-existing animal model or using X-rays or mutagenic agents to make one. In either case, the mutation in the model was unlikely to match the disease mutation very closely, making tenuous the relevance of experimental results in such models.

With transgenic techniques, in contrast, the entire mutated gene, or any desired piece of it, can be transferred wholesale into the model organism, and its expression can be precisely controlled by stitching in the right “promoter” (a gene region that dictates the timing, pace, and location of gene transcription).

With good animal models, research on the cellular pathogenesis of Huntington’s disease has exploded. Mice receiving the mutant gene develop neurodegenerative disease reminiscent of Huntington’s, opening the door for detailed studies of neuronal death pathways in vivo, and at the same time providing a model in which to test a wide range of therapies. The gene has since also been transferred into fruit flies, whose retinal neurons degenerate under its influence, and even into yeast, which is well suited for detailed biochemical studies.

Examining the brains of Huntington’s disease transgenic mice, Gillian Bates at Guys Hospital in London, and Marian DiFiglia at Massachusetts General Hospital, made an unexpected discovery. The nuclei of dying neurons contained inclusions—clumps— packed with mutant huntingtin protein linked to a ubiquitous cell protein termed, logically enough, ubiquitin. They quickly found similar inclusions in the brain cells of Huntington’s disease patients. Both results were published in late 1997. The significance of these inclusions, the researchers believed, was enormous: “Because brain regions affected in HD contain [these inclusions] …the formation of these structures is directly implicated in HD pathogenesis.”

Exactly how these inclusions did their dirty work was still unknown, of course, but there seemed little doubt that inclusions somehow gummed up the works and led straight to neuronal death. And the discovery of these clumps led straight to hopes for therapy: Would anti-aggregation drugs slow or prevent Huntington’s progression?

Only a year later—a mark of how rapidly research in this field has begun to advance—another group, also based in Boston, refuted Bates’s and DiFiglia’s conclusions. Michael Greenberg and colleagues showed that blocking the linkage of huntingtin and ubiquitin prevented clumping, but increased cell death. The researchers argued that, far from being toxic, “the formation of intranuclear inclusions may be part of a cellular strategy for degrading or inactivating toxic forms of huntingtin and, thereby, protecting the cell.” In this view, protein aggregates are something like graveyards: They do not cause death, they are just a sign of it. Without them to accumulate the bodies, the cell would be even worse off than it is.

Together with parallel conclusions in another polyglutamine disease model, these results threw a bucket of cold water on the inclusion hypothesis. While some leading researchers still maintain protein aggregates are central to the Huntington’s disease process, many others think that, barring some surprising demonstration that contradicts Greenberg’s results, that case is closed. The implications of these studies have gone beyond polyglutamine diseases, as well. Cytoplasmic inclusions called Lewy bodies are the pathological hallmark of Parkinson’s disease, and had long been assumed, though never proved, to be implicated in the degeneration of substantia nigra neurons in that disease. Increasingly these too are seen as the flotsam of a ship about to go under, rather than the blast that sinks it. Many groups originally intrigued by inclusions have turned their attention away to focus on what they see as more likely pathogenic candidates.

Plugged Proteasomes, Porous Membranes

Greenberg discovered that by blocking ubiquitination of huntingtin, he prevented aggregation but increased cell death. By turning the spotlight on ubiquitin, Greenberg highlighted what many now think is an important pathway in Huntington’s disease pathogenesis.

Ubiquitin is not just any protein partner: Like a toe tag on a corpse, ubiquitin marks a defective protein for a trip to the cell’s crematorium—a barrel-shaped molecular machine called the proteasome that renders proteins back into amino acids, simultaneously ridding the cell of worn-out molecules and recycling their parts.

What mischief? One intriguing hypothesis relies on structural arguments about the polyglutamine itself. Its best-known proponent, until his death in early 2002, was Max Perutz. Trained as a physicist, Perutz is revered as the father of molecular biology for his elucidation in 1959 of the structure of hemoglobin, for which he received the Nobel Prize three years later. In the 1990s, Perutz turned his attention to the polyglutamine diseases. He argued that the crucial similarity among the diseases was the threshold of approximately 37 to 41 glutamine repeats. Below that, no disease occurred, above it, it always did. Perutz’s research explained this threshold by showing that 40 glutamines were needed for a polyglutamine chain to curl around to form a stable tube.

What damage might these tubes cause? Some research has shown that purified polyglutamine tubes can insert themselves directly into artificial membranes created in a laboratory, where they wreak havoc with electrical potentials across the membrane. These results were only laboratory curiosities until recently, when Tim Greenamyre and colleagues at Emory University showed that mitochondria from Huntington’s disease patients and transgenic mice were defective in their ability to regulate the cross-membrane calcium potential. (Mitochondria are small, membrane-bound power generators in the cell, which use calcium to maintain a charge separation across their membranes.) Prolonged loss of this potential triggers apoptosis (cell suicide). What Greenamyre found was that long, but not short, polyglutamines directly associated with the mitochondrial membrane were preventing a balance in calcium and causing cell death.

The implication is that mutant huntingtin may curl up into tubes and insert itself into the mitochondrial membrane just as it does in the artificial membranes, disrupting the mitochondria’s ability to regulate calcium. Eventually, perhaps from accumulated dysfunction, perhaps from some acute stressful event, the cell death response is triggered.

An Embarrassment of Riches, a Poverty of Proof

As exciting and intriguing as the polyglutamine tube scenario is, however, this is still only one of many competing hypotheses. Another, for which there is much new evidence, suggests that the polyglutamine tract directly interferes with the regulation of a suite of genes, including genes for nerve growth factors, neurotransmitter receptors, and other processes that are altered in Huntington’s disease cells. Changes in the expression of these genes might cause the cell to be more sensitive to stress, less responsive to life-sustaining growth factors, or more likely to undergo apoptosis. The number of genes thought to be influenced by mutant huntingtin is large and will surely grow, but just as surely, not all these genes will prove equally central in pathogenesis. The unanswered question is whether any are, or whether these gene regulatory changes are epiphenomena in a cell dying for some other reason.

A similar horse-and-cart question surrounds the involvement of caspases, enzymes that propel a sick cell along the apoptotic pathway. Mutant huntingtin activates caspases, but it is not clear whether this is an early event or a late one.

Here, then, are some of the questions, hypotheses, and puzzles awaiting further research, a kind of embarrassment of scientific riches that has sprung forth in the wake of the triumphant discovery of the Huntington’s disease gene. Although that discovery was first cheered more than 20 years ago, Allan Tobin, director of the Hereditary Disease Foundation, which underwrites some of this research, describes the field as still in its early years. “It will be mature,” he says, “when we can falsify some of these hypotheses—when we can say, ‘I don’t believe that anymore.’” As of early 2003, then, the field of Huntington’s disease research remains immature. It is possible that all these hypotheses are partly correct, that each describes only one part of a still-hidden whole. This is the predicament of research on the edge of knowledge; the truth remains obscure, and all one has is the imperfect evidence.

Whither Treatments?

In the meantime, experimental treatments are being developed or contemplated that are aligned with each pathogenic model, and the relative success of each treatment will likely advance or retard the relative standing of each model. A case in point is minocycline, an antibiotic that is also a caspase inhibitor, that has slowed disease progression in mice. Only two years after this discovery, clinical trials are in progress, a testament both to the lightning pace of Huntington’s disease research and to the dearth of any other treatments for the disease.

Some treatments have been tried, and have failed, already. Transplanting fetal cells into the brain’s striatum yields limited motor benefit, but doesn’t halt neurodegeneration and is useless for cortical losses to cognition, which is clinically more significant than motor impairment. Despite promising results in mice, neither creatine nor coenzyme Q10 is effective in patients; these nutritional supplements are thought to have antioxidant and mitochondria-boosting properties. Do they address the wrong problem, or are they just too little too late?

Reversing the Damage

Despite these early disappointments, most researchers believe that rational, effective treatments for HD are coming. When they do, will they help patients who have already developed symptoms, or are such patients too far along in the disease process to be rescued? A remarkable series of experiments in mice has shown there is reason for optimism. Huntington’s disease may be reversible.

These experiments, performed by Rene Hen, Ai Yamamoto, and colleagues at Columbia University, employed one of the newest and most remarkable tools in the molecular biology toolbox, the conditional promoter. This promoter, which is placed at the start of the huntingtin gene and controls its expression, can be turned on or off simply by administering or withholding an antibiotic in the mouse’s water. With this tool, Hen’s group has asked a series of fundamental questions about the short- and long-term effects of mutant huntingtin. In 2000, they announced that mice that had expressed mutant huntingtin since birth, and that had already begun to develop inclusions and lose their motor coordination, got better when the gene was turned off: They became more coordinated, and their inclusions gradually cleared up. In further experiments since, they have shown that learning deficits—a surrogate in the afflicted mice for cognitive decline in humans— are also reversible, once the gene is shut off.

The exciting conclusion from this work is that the progressive deterioration of Huntington’s disease appears to feed on a continual influx of new mutant protein, and without it, dysfunctional but still living brain cells are able to repair the damage and return to normal function. If the same thing holds true for humans with Huntington’s disease, the implications of this work are enormous, and enormously hopeful. If we can just shut off the gene, the brain might not only stop getting worse; it might be able to get better.

But that is a huge “if.” How can the gene be shut off in humans, who after all are not born with highly engineered conditional promoters attached to their Huntington’s disease genes? While gene therapy may hold promise for loss-offunction conditions like cystic fibrosis, in which a protein is missing and can be supplemented, the toxic gain-of-function in Huntington’s disease presents a harder problem. How do you silence a gene? This remains the major stumbling block for any gene-based HD therapy.

While this proof of principle is reassuring, it is a very long way from cell culture to human therapies, and hundreds of treatments have stumbled along that path. In this case, delivery (getting the siRNA into the target cells) is the main challenge, but effectiveness, toxicity, and side effects are also concerns that will have to be met by siRNA or any other therapy.

To Huntington’s disease sufferers and their families, the progress toward therapeutics must seem slow indeed, especially after the elation of unexpectedly sudden success in the search for the gene. Still, most researchers are confident that rational treatments, based on a deep understanding of the pathogenic process, will soon be developed. While the field of Huntington’s research is not yet mature in Allan Tobin’s formulation, it may be in its adolescence, in which profound changes seem to occur almost daily. The pace of discovery in the last decade justifies the reasonable hope that George Huntington’s disease will not long remain “one of the incurables.”