A team led by researchers at the Johns Hopkins School of Medicine has found that several inherited traits in the mouse genome break the rules Gregor Mendel laid down in the 19th century about how genes work. This is not because the genes mutate but because of chemical modifications of the genome that can jump across generations, change depending on whether an animal is male or female, and even copy themselves from one copy of a gene to another.
Their findings were reported recently in Nature Genetics.
While the study was in mice, the implications go far beyond — as the same mechanisms may also shape human health, and be able to explain conditions that have long resisted genetic analysis.
‘On’ and ‘off’ genes
Every cell in the body carries two kinds of genetic information. One is the sequence — the order of the four chemical bases, A, T, G, and C, that spells out which proteins a gene produces. This is the genome that geneticists have spent decades mapping. The other is about whether a small chemical compound — called the methyl group (–CH3) — is attached to the DNA at specific locations.
When cytosines are methylated, the gene is usually silenced. When the group is removed, the gene can be expressed again. This process of switching genes ‘on’ and ‘off’ without changing the underlying sequence is called epigenetic modification.
Epigenetic modifications are not permanent. They can even differ between tissues — e.g. liver cells and muscle cells methylate differently. Many of them are also reversed when the body makes fresh sperm or eggs.
The new study has found this rule to be violated.
An illustration of cytosine and methylated cytosine. | Photo Credit: Mariuswalter (CC BY-SA)
Vastly underestimated
The researchers used a technology called nanopore sequencing. A single strand of DNA is threaded through a microscopic hole — the pore. As each base passes through, it produces a distinct blip in the electrical current flowing across the pore. Methylated cytosines produce a different blip from unmethylated ones.
The team sequenced DNA from the liver and muscle tissues of mice belonging to two inbred strains. Their genomes differed at millions of positions.
Of the 7,600 or so locations on the genomes where the methylation patterns differed, around 93% followed Mendel’s rules. That is, methylation had passed reliably from parent to offspring. But in the remaining 7% or so, the researchers identified more than 500 instances of non-Mendelian epigenetic inheritance.
To illustrate, in 305 genome regions, all in the liver, the methylation depended on whether the mouse was male or female. Mendel’s rules don’t allow this. In 304 of these regions, females’ genomes were more methylated than males’ genomes.
While sex-specific methylation in the liver has been reported before in both mice and humans, the new study found that the scale has been vastly underestimated.
A new first
The researchers also identified at least five new genes displaying genomic imprinting — where the level of methylation depends on whether the gene was inherited from the mother or the father.
The most striking finding had to do with a phenomenon called paramutation. Here, the methylation status of one copy of a gene — say, the one inherited from the father — is also transferred to the other copy, inherited from the mother. The ‘rewritten’ mother’s copy then persists through the next generation.
Again, scientists had observed paramutation in plants and engineered mice before. This is the first time it has been reported to be naturally occurring in a mammal’s genome.
In the study, the paramutated gene was Capn11, which contains the recipe for a protein that is active in the testes during meiosis. When this protein is not available in sufficient amounts, the effects may include infertility and azoospermia (having no sperm in the semen).
The team also identified two likely instances of paramutation related to the gene Vps37c and a downstream stretch of the genome. Both these areas are associated with intracisternal A particles — the genetic remains of ancient viruses that became embedded in the mouse genome.
According to the study’s authors, these particles may be central to the paramutation process since they have been known to resist the body’s mechanisms to remove epigenetic modifications when producing sperm or eggs.
Genome-wide association studies
The study also offers a new framework to explain heritable traits that researchers have struggled to study using genetics. For instance, a condition called hypertrichosis pinnae auris leads to the growth of coarse, dark hair on the outer ear and is prevalent among South Asian men. It is transmitted from fathers to sons but never to daughters.
According to the study’s results, if a father carries a paramutagenic version of the gene that causes this condition and the mother carries a ‘normal’ version, then in a son the paramutagenic copy would rewrite the maternal copy, leaving the son to have hairy ears. He could then pass the paramutagenic variant of the gene to his own children. But if the paramutagenic variant is demethylated in females — as the new findings suggest can happen — a daughter’s own copies of the gene would be unmethylated and her ears would be unaffected.
Even more broadly, the findings imply that genome-wide association studies — an increasingly popular kind of study that looks for the genetic variants linked to diseases — may be missing a significant fraction of heritable variation because that variation is not in the sequence. Instead, the study’s authors have argued, the work opens the door to a new kind of analysis: allele-specific epigenome-wide association studies, which trace methylation patterns rather than sequence variants, to find parts of the genome associated with diseases that conventional methods miss.
Human genetics has long assumed that if you sequence the genome, you have read the instruction manual. It may be that the manual has a second volume — and we have barely opened it today.
D.P. Kasbekar is a retired scientist.
Leave a Reply