John Collinge's mice are doing nicely. For the first 500 days of their lives, their coats have remained shiny, their noses pink and their appetites healthy. And that's good news, because Collinge's rodents carry a secret that has important implications. Each creature has been genetically altered so that it is now susceptible to the human brain ailment, Creutzfeldt-Jakob disease (CJD).
In addition, each has been inoculated with a heavy dose of brain matter taken from cows affected by the condition's cattle equivalent, bovine spongiform encephalopathy -- BSE, or mad cow disease. The aim of this experiment is simple. Mice without human genes succumb to BSE within 200 days. The fact that "humanized" mice don't yet have CJD symptoms shows there must be a "reasonable species barrier" between people and cattle. This barrier may limit mad cow disease's spread to the British population.
The findings of Collinge have been eagerly anticipated, for his understanding of prions, the protein-like entitites responsible for both CJD and BSE, is unsurpassed in Britain and his experiments at Imperial College, London, are rated as being of crucial importance. His research should reveal whether BSE will spread to only a few humans or to devastating numbers.
But in spite of the comforting implications of his experiments' initial results, Collinge's message has a startingly gloomy content. "For all I know, all 50 of my mice could come down with CJD next week," he says. "However, they may all die of old age in about another 200 days. That would have to be good news -- of a sort."
Such a result would not prove that humans are unaffected by a bovine prion disease. It would simply mean that the species barrier between cow and human is sufficiently strong to prevent that transmission happening within the average lifetime of a mouse. It may still occur over a longer period.
This is almost certainly the case, adds Collinge. "The new cases of CJD in humans have a startlingly uniform pathology: early age of onset, psychiatric disturbances and a relatively long period before death occurs, about 14 months. This indicates a single cause -- BSE -- though we cannot yet say whether it will lead to 20 or 100,000 deaths a year."
Then there's the question of who succumbs and who survives. Collinge, a key member of the British government's BSE task force, says he doubts that the new victims' youth was related to their predeliction for hamburgers. "I think young people are just going to turn out to be more susceptible in general."
However, his research has revealed that about half the British population has a genetic mix of brain prions that should enable them to resist BSE. "Of course, that leaves out half the population who do not have the right brain prion mix, and they will be the ones who provide the pool of victims."
The most encouraging part of Collinge's message is that, far from being helpless in the face of an inexorable killer, science is well equipped to fight back. He says the knowledge gained through trying to understand the unusual biology of prion diseases is paying off.
"That knowledge means we can now think realistically about selecting drugs that could counter the effects of CJD. There is certainly hope in that part of the story."
MICE that are part-human may soon provide the answer to the most pressing medical questions of the day. Have the 14 cases of new variant Creutzfeldt-Jakob disease been caused by mad cow disease, and, if so, what are the risks of a human epidemic?
Within weeks, Prof John Collinge's team at St Mary's Hospital Medical School, London, is likely to have some of the first answers, thanks to experiments on mice that contain human brain proteins. Breeding this "humanised" colony, the result of years of genetic engineering, was possible thanks to backing from the Wellcome Trust, the Biotechnology and Biological Sciences Research Council and the millionaire brothers David and Frederick Barclay.
Spongiform disease is thought to strike when prion protein flips into a rogue form. Another human protein called p53, could have a self-replicating alter ego linked to many cancers. " Once in that form," said Prof Collinge, "it has the ability to recruit more of the normal protein, setting up a chain reaction so that all the protein in the brain ends up in a distorted form." As self-replicating protein runs wild, it forms toxic deposits called plaques. Meanwhile, the brain is drained of the normal version of the prion protein it relies on.
All that stands between an epidemic and the discovery that 440,000 infected cattle went into the food chain before the offal ban is what scientists call the "species barrier". The obvious way to measure its height - to expose people to BSE - is unethical. A second approach is to sit back and watch the pattern of CJD cases that unfolds, correlating it with exposure to BSE. That could take years. Prof Collinge's method is to humanise mice so they are as susceptible to CJD as humans, expose them to BSE, and see how they fare.
When normal mice are injected with BSE, the first fall victim to disease after about 350 days and the last after 700 days, with an average of 400 days. If an extract is taken from any of these mice and used to infect others, the latter group succumbs more quickly and simultaneously, after 140 days. The species barrier has been breached. "Once you have triggered prion production in an animal, it makes its own infectious prions which are much more dangerous to other animals of that species," said Prof Collinge.
Comparing the incubation period of the first group of mice with that of the second reveals a difference of 250 days. This is the height of the barrier between mice and cows. To measure the barrier between humans and cows, Prof Collinge bred mice containing human prions. Unlike ordinary mice, they are very susceptible to CJD. "They all get sick at around 200 days," he said.
'Even if cross-infection is highly inefficient, such that we could inject thousands of mice and never see any get sick, you could still get transmission to significant numbers of people' The humanised mice even make human prions. "Now we can see whether the cow prions can persuade human protein to make the human infectious prion agent." Prof Collinge challenged his humanised mice with BSE. "If they get sick at all, it must mean that BSE can convert the human protein," he said.
The experiment has been running for more than 500 days, showing it is no pushover for the BSE agent to convert human prions to the infectious form. The worst-case scenario would be for all the rodents to succumb in the next few days. Then the species barrier would be 300 days - the 500 days this experiment has run, less the 200 days it takes humanised mice to succumb.
The barrier between humans and cows would then be at least as good as that between mice and cows, if not a little better. That, alas, is cold comfort. "Nine grams of bovine brain fed to a mouse will kill it," said Prof Collinge. "If the species barrier between cattle and humans were no better than between cattle and mice, we could be in serious trouble."
Even the most favourable scenario - the mice live to a great old age of around 700 days - is unsettling. "We only inoculated 50 mice, when tens of millions of people have been exposed to BSE," he said. "Even if cross-infection is highly inefficient, such that we could inject thousands of mice and never see any get sick, you could still get transmission to significant numbers of people."
Another experiment on humanised mice could prove as significant. Some 200 days ago, they were exposed to the new variant of CJD that triggered the crisis. This experiment has already run for the time it takes for other forms of CJD to strike and should conclude more quickly than another experiment in Edinburgh, where normal mice are used to compare strains. If Prof Collinge's humanised mice succumb in a similar way to BSE, "we may be able to draw a direct connection".
Collinge says his research has revealed that about half the British population has a genetic mix of brain prions that should enable them to resist BSE. "Of course, that leaves out half the population who do not have the right brain prion mix, and they will be the ones who provide the pool of victims."
I disagree with the conclusions of the above paragraph, which I take as saying met/val or val/val or some such at codon 129 won't be in "the pool of victims." Collinge was just trying here to get a little of the bad news in the open while still offering some hope, so as not to throw the populace into a panic.
Instead, I think we are simply seeing progressive CJD (early age of onset, rapid decline) at the start of the epidemic in the most susceptible genetic background. Other populations would be looking at later onset, different symptoms, and slower progression of the disease. For example, they could look like cases of sporadic CJD rather than v-CJD.
As I read the middle-ground scenario in Table 30 in Dealler's "Lethal Legacy", 142,000,000 individual meals were eaten in which 10,000 or more lethal infectious units (IU)were present in that serving. As I read Table 29, again the middle-ground scenario, 18,000,000 people would have been exposed to a cumulative dose of 1,000,000 IUs or more. I also see in Table 5 significant numbers of under-2 calves showing BSE whereas Anderson et al quote an experiment with cows receiving 330 grams intra-cerebrally first showing symptoms at 40 months.
I would expect codon 129 to only modulate some parameters of the illness, not be all or none:
For example, D178N M129 has a mean age of onset of 49 years, mean duration of 15 months, and presents in decreasing frequency dementia, ataxia, myoclonus, spongiform cerebral cortex, ...
whereas D178N V129 has a mean age of onset of 45 years, mean duration of 22 months, and presents in decreasing frequency insomnia, dysautonomia, ataxia, myoclonus, and thalamic nuclei atrophy.
Sporadic and iatrogenic CJD are similarly enhanced, but by no means exclusively so, I recall in homozygotes over heterozygotes.
Could anti-sense oligonucleotides be used to catch prion protein mRNAs?
Yes, this is a good therapeutic proposal that could easily be tested in mice. For all I know, the experiment is under way as we speak. Promising results have been published with other diseases. There is no technical problem is producing anti-sense oligonucleotides in quantity.
A variation on this is to permanently knock out _one_ copy, more or less, of the prion gene (ie on one chromosome).
On the hopeful side, anti-prion mRNA, meaning mRNA complementary to the normally transcribed strand, would be expected to accomplish down-regulation of prion mRNA promoter and lower prion production itself and so abnormal prion. Gene dosage experiments, such as the 7x experiments of Prusiner, suggest that higher prion levels lead to exaggerated susceptibility and more rapid course of disease. Transfection under the control of a friendly promoter might produce a steady flow in situ of the desired oligonucleotides. Prion turnover is fairly rapid so buildup might be slowed. Even a total absence of prion takes some years to bring symptoms, so knocking back to half the steady-state level might be the best of both worlds.
On the less hopeful side, Collinge noted that prion mRNA is produced constitutively in adults; however, this is not true in neurodevelopment, as Prusiner discusses in the AnnRevMicrob 48 1994. In other words, cells have the potential to respond by some feedback mechanism to curtailment of prion production by trying harder (ie, more transcription starts, more translations per mRNA, etc), maybe resulting in the same end levels of production, especially over time. Then there is the matter of the blood/brain barrier, establishing the safety of any intervention, and early identification of people who might benefit.
Remember that while the prion gene is present in only one copy per chromosome, some other gene of unknown function is already producing mRNA rather complementary to the prion anti-sense strand. Your prion anti-sense oligonucleotides might interfer with this gene by binding to its anti-sense strand, with unkown effects.
In Stephen Dealler's 'Lethal Legacy" in tables 13-13 are plotted the log of infectious units per gram of tissue versus percent of incubation time to overt symptoms, the result was roughly log-linear.
No cow data. Mouse and sheep spleen were at 100% final infectivity [800,000,000 and 10,000 IU, resp.] and steady (flat) 30% of the time into the incubation period for displaying overt symptoms.
Mouse brain: infectious titre increased steadily beginning at 40% of the incubation period but peaked and flattened at 75% at 10,000,000,000 lethal infectious units injected to the same species [IU] per gram.
Sheep brain: increased steadily with flatter slope with the earliest data point shown 65% of incubation period, so from 10,000 IU per gram to 120,000. Mink, goat, hamsters data is given as well.
We don't eat mouse brain or spleen in the US but I hope they are disposing of these animals properly (incinerating under controlled conditions). One asymptomatic mouse has enough units to kill every mammal on the planet.
I would recomment to anyone purchase of the Dealler's book 'Lethal Legacy' for 6 pounds paperback: there are 32 tables in the back. Bloomsbury, ISBN 0 7475 2904 3.