Life's Greatest Secret (39 page)

Although Temin’s more extreme claims were misplaced, his discovery of reverse transcriptase was significant because it showed how viruses could cause cancer by altering the DNA of the cells they infect and by deregulating genes, leading to uncontrolled growth. The enzyme also went on to play an important role in the development of molecular genetics and of our ability to genetically modify organisms by introducing new sequences into DNA – it is used by scientists to make complementary DNA (cDNA) from mature mRNA. In 1975, Temin and Baltimore, along with Temin’s PhD supervisor, Renato Delbucco, won the Nobel Prize in Physiology or Medicine for their work on how cancer viruses affect our genes.

*

In his 1970 clarification of exactly what he meant by the central dogma, Crick highlighted three kinds of information transfer that he postulated would never occur: protein → protein, protein → DNA and protein → RNA. However, even as he made such a clear prediction, Crick was cautious, underlining our ignorance and the fragility of the evidence upon which he based his slightly revised ‘dogma’:

our knowledge of molecular biology, even in one cell – let alone for all organisms in nature – is still far too incomplete to allow us to assert dogmatically that it is correct.

And in the very next sentence he highlighted a potential exception:

There is, for example, the problem of the chemical nature of the agent of the disease scrapie.

Scrapie is a neurodegenerative disease affecting sheep and goats that has been known for hundreds of years. In 1970, its cause was mysterious – the disease-causing agent was known to be resistant to heat, formalin, ultraviolet radiation and ionising radiation (all of which destroy nucleic acids and inactivate viruses) and it left no sign of infection in the animal’s immune system. This curious set of facts led some scientists to argue that scrapie was in fact a genetic disorder rather than an infectious disease. Others daringly suggested that the scrapie infectious agent was a protein – this was what lay behind Crick’s remark in 1970. At this time, all known infectious agents were based on nucleic acids and were either organisms or viruses. A protein-based infectious agent would be a truly radical discovery, and this suggestion was therefore treated with some scepticism.
⁷

In 1982, Stanley Prusiner’s group discovered that scrapie could be detected by the presence of a protein that was also a potential infectious agent – they called it the prion protein, and it seemed to act by altering the shape of non-infectious proteins that were otherwise identical to the prion.
⁸
This was utterly novel, both because it suggested that a protein could transmit a disease and because it implied that the prion might breach the central dogma by allowing the transmission of information from protein → protein. In the 1940s, Mirsky had suggested that there might be minute levels of protein contamination in Avery’s purified DNA; in the 1980s, some of Prusiner’s critics argued that there must be small amounts of nucleic acid in the apparently pure prion protein extracts. The prejudices that prevented some scientists from accepting Avery’s discovery that DNA is the hereditary material reappeared in the case of this infectious agent that was apparently not based on nucleic acids.
⁹

Interest in scrapie grew in the late 1980s and the 1990s with the horrific outbreak of variant Creutzfeldt–Jakob Disease (vCJD) in humans and its equivalent in animals, ‘mad cow’ disease (bovine spongiform encephalopathy, or BSE). These diseases infected millions of cattle and caused the deaths of hundreds of people, most of them teenagers and young adults. Both BSE and vCJD showed similarities to scrapie, and again the evidence suggested that an infectious protein was involved. It was eventually shown that the same prion protein causes all three diseases. Although it is still unclear how the BSE outbreak began, it is possible that cows initially got the disease from scrapie-infected sheep, the remains of which were fed to cows as meat and bonemeal. Whatever the original source of BSE, people caught the disease by eating diseased sections of the bovine nervous system that had been included in processed meat such as burgers.

It is now accepted that the aberrant prion protein alters the conformation of normal prion proteins, thereby producing the brain pathology in sheep, goats, cows and humans.
¹⁰
In 1997, Prusiner was awarded the Nobel Prize in Physiology or Medicine for his discovery, but despite the widespread acceptance of the prion hypothesis, a few scientists continue to argue that virus-like particles are involved in scrapie and similar diseases.
¹¹
Although it is known that yeast prion transmission involves only proteins, there remains the slim possibility that unknown nucleic acid-based cofactors may be involved in mammalian prion diseases.
¹²

In 1982, Prusiner suggested that the prion codes directly for the synthesis of another prion, not merely for its shape. This would have completely destroyed one fundamental point of the central dogma, that protein does not code for protein. Prusiner was wrong. Prion proteins are produced by the action of the prion gene, encoded in DNA and transcribed into RNA and then translated into a chain of amino acids – the normal prion protein plays a role in producing myelin, which protects nerves.
¹³
In both the benign and the pathogenic forms of the prion, the amino acid sequence remains the same, so there is no transfer of information as defined by the central dogma, which referred solely to the sequence, not the structure. Although it can be argued that three-dimensional conformation is a form of information – indeed, Crick accepted as much – the change induced by the prion protein is probably more similar to the action of a crystal growing by assembling identical copies of itself than it is to that of a DNA molecule, which can produce a correspondence with a sequence in a different kind of molecule.
¹⁴
Despite the highly unusual and pathological conditions that produce prion disease, the central dogma remains fundamentally intact.
¹⁵

*

It was 1977, and I was an undergraduate at Sheffield University listening to a lecture by Professor Kevin Connolly, a world expert on child development and behaviour genetics with whom I eventually studied for my PhD. Kevin was describing the effects of social deprivation, and he highlighted a 1967 study that showed that if a female rat pup was removed from its mother for only three minutes a day, her sons and daughters and even
their offspring
would show pathological changes in their activity and weight, even if they and their parents had been reared under normal conditions. For my excited 20-year-old brain, this study had two implications. First, it suggested that the effects of social deprivation in humans might continue to echo down the generations, even if people were subsequently provided with an excellent environment. Second, and more fundamentally, as the title of the paper put it, the effect involved a ‘nongenetic transmission of information’.
¹⁶
Not all hereditary information is made of DNA, the result suggested. Almost speechless with surprise, I went up to Kevin after the lecture to confirm that I had understood correctly. I had. I went away, thinking hard about what this effect might mean, and above all how it might work.

Nearly forty years on, I am still interested in the non-genetic transmission of information down the generations and the underlying mechanisms – one of my PhD students, Becky Lockyer, has recently studied them in
Drosophila.
¹⁷
The existence of intergenerational transmission of environmentally induced changes is now well established, and it is known that these effects can buffer populations of organisms against rapid environmental change before new adaptations evolve.
¹⁸
They form part of the complex route from DNA to phenotype, presenting biologists with fascinating examples of plasticity – how a given DNA sequence can generate a variety of different phenotypes.

Some of these effects can be thought-provoking, like the study I heard about in 1977. For example, in 2009 Larry Feig’s group at Tufts University in Boston reported that if female rats were given an enriched environment during their adolescence, their offspring – conceived after the enrichment had ceased – showed an increased learning ability. The effect was even strong enough to overcome genetic defects in learning.
¹⁹
There is no evidence that such memory effects occur in humans, nor is it known how this particular transgenerational effect works, but it does not necessarily involve genes. In the case of deprivation, poor parenting can cause behavioural and hormonal changes in offspring that lead to those individuals’ being poor parents in turn, and so on.

This kind of phenomenon is often described as an ‘epigenetic’ effect, even where no effect on genes has been demonstrated. Strictly speaking, ‘epigenetic’ refers to any way in which the genetic code is modulated on its route from DNA sequence in a cell into an expressed character; that is, how genes are regulated.
²⁰
However, the term is increasingly being used primarily to describe rare cases in which changes in gene regulation are transmitted down the generations. Journalists, philosophers and scientists have claimed that transgenerational epigenetics radically alters our understanding of inheritance and evolution, and even marks ‘victory over the genes’ as the German magazine
Der Spiegel
put it in 2010.
²¹
The truth is somewhat less dramatic.

In their most widespread form, epigenetic effects explain how genes are turned off and on in our cells, enabling each specific cell type to appear, allowing a complex organism with various kinds of tissues to develop from a single-celled embryo – precisely the mystery highlighted by Jacob and Monod when they discovered the first example of a regulator gene. In other words, epigenetic effects, whether they are transgenerational or occur in a single organism, are examples of gene regulation.*

Epigenetic regulation often involves the activity of small RNA molecules, which are produced by genes in complex regulatory networks.
²²
One of the most widely studied forms of epigenetic control is the placing of epigenetic marks on genes, which occurs when the cell adds a methyl group (CH
₃
) onto a cytosine base of a DNA sequence. This process, known as methylation, does not alter the sequence but can result in the gene being silenced – the gene is ignored, as though the transcription machinery no longer recognises the sequence. Methylation is relatively common in plants, but it is rare in animals. Where methylation does occur in animals, it is overwhelmingly in somatic cells, which form the organism’s body, not in the germ cells that pass genes to the next generation. Methylation marks that may occur in germ cells are mostly removed during the formation of eggs and sperm; any that survive this process are generally wiped immediately after fertilisation.
²³
In a recent example, the undernourishment of a female mouse affected DNA methylation in the sperm of her sons, although this was not then transmitted to the son’s offspring.
²⁴

Epigenetic chemical marks can also be left in histone – the proteins that package DNA. It is widely assumed that changes to histones are involved in gene regulation, but the evidence is unclear, and in
Drosophila
one kind of histone can be deleted completely without any effect on gene transcription.
²⁵
For the moment, there is no evidence that histone modifications are directly passed on during ordinary cell division, never mind to the next generation. Instead, in some circumstances enzymes may re-introduce histone marks in daughter cells from which these marks have been wiped.
²⁶

Epigenetic effects are particularly important in the development of certain cancers: genes that normally silence genes that can lead to uncontrolled growth can be themselves silenced by epigenetic effects from the environment, leading to disease. In the late 1950s, Szilárd, Jacob and Monod called this effect the ‘derepression of repression’, and we now know that it can lead to some forms of cancer that, in certain rare circumstances, can be passed from one generation to another.
²⁷
For example, an increased susceptibility to a genetic form of testicular cancer in mice was found to be transmitted down several generations, suggesting that expression (or silencing) of a gene in the parent can lead to that gene being expressed (or silenced) in the offspring.
²⁸
It is this kind of transgenerational effect that captures the imagination of journalists, scientists and the general public, because it apparently contradicts the basic teachings of genetics.

One of the most widely recounted examples of apparent transgenerational epigenetic effects is the terrible ‘Dutch famine’ that took place during the winter of 1944–45. Women in the Netherlands who were pregnant at the time had smaller children; now that these children are adults, they show poor glucose tolerance and are more likely to suffer from diabetes. They also show differences in their levels of DNA methylation.
²⁹
But it is unclear whether the methylation differences are the cause of the abnormalities or are the consequence of them, and above all there is no evidence that the germ line in these people has been affected.
³⁰

Mammals often show a particular form of transgenerational epigenetic effect known as genomic imprinting. We inherit two copies of each gene, one from each parent; in some cases, either the maternal or paternal copy is silenced (or, more rarely, enhanced) by the action of epigenetic marks or imprinting, and the other parental form reappears in the offspring.
³¹
Although genomic imprinting affects only a small proportion of mammalian genes, in females it lies at the basis of the inactivation of one of their two X chromosomes that occurs shortly after fertilisation and is essential for normal development. Like other epigenetic changes, genomic imprinting effects can be reversed in the next generation – this is not a permanent change that alters the genes that are carried by a population.