Thursday, May 20, 2010

An introduction to epigenetics


You very likely remember from high school biology the story of Gregor Mendel, the monk who in the mid 1800s observed that the inheritance of traits among pea plants in his garden was not random but seemed to follow certain patterns or laws. Despite his discovery, however, he knew nothing about the mechanism by which his pea plants transferred this information, their genetic code. We now know it is as the molecule DNA, deoxyribonucleic acid.

(Your other memory from biology class likely is this famous black and white picture of Francis Crick and James Watson standing next to a model of DNA, representing their discovery of the structure of DNA in 1953.)

Photo courtesy of UCSB

Science has rapidly progressed and within the last ten years the Human Genome Project has unraveled the full sequence or code of human DNA; yet, despite knowing the full sequence of DNA there is a lot in humans we still can't explain, predict, etc. We haven't licked human disease. Why is that? What more to the puzzle could there be?

Let me address one aspect of this puzzle: epigenetics.

How is it that in your body the exact same DNA, or code, is present in your liver cells as in your brain cells and yet the two cell types are very different? Or that identical twins share the exact same DNA and yet are different individuals, even if subtly so? Clearly, the answer to these questions is not simply DNA.

The answer instead partly lies in how the DNA is read or expressed in cells. One mechanism: regions of the DNA are made more or less accessible to being read. Two people allowed to read different sections of an encyclopedia will walk away with different knowledge.

For instance, a cell may attach or associate certain chemical groups or molecules (such as methyl groups) to certain regions of DNA and, thus, alter the structure or reading of those regions but in no way change the DNA sequence itself. These modifications are not part of the DNA (or genome) itself, but 'outside' or 'above' the DNA, and make up what is called the epigenome (from the Greek meaning literally above the genome or DNA).

In this way the epigenome, despite not being the actual code or DNA sequence, plays an important role in cells. Knowing the DNA sequence is not sufficient. What is also important is how the DNA is expressed.

Two prominent examples of the epigenome's effect on the human body include X-chromosome inactivation and genomic imprinting.

X-chromosome inactivation

Quick review: In general the cells of our body have two copies of DNA, one copy inherited from our mother and the other from our father. In order to fit into our cells, the DNA is packed up and organized as chromosomes, and human DNA consists of 23 pairs of chromosomes (again, one pair from each parent). The sex chromosomes (denoted as either X or Y) are the chromosomes that determine our sex, whether we'll be male or female. A person born with two X chromosomes is born as a female.

For years scientists have known that in any given cell of a woman, only one of the two X chromosomes (either from her mother or father) is generally expressed while the other is inactivated (not expressed). This inactivation (X-chromosome inactivation) appears to occur randomly, such that in one cell the mother's X chromosome may be inactivated and in another cell the father's X chromosome. It occurs as a result of changes to the epigenome, which allow only one X chromosome in each cell to be read despite both X chromosomes being present. Once set this inactivation and epigenetic changes are passed along as the cell divides.

Genomic imprinting

Another exception to the idea that we are essentially the expression of both our mother's and father's DNA, because each cell contains a copy of DNA from each parent, is genomic imprinting. For some locations on the DNA, who we are is determined not by both the mother's and father's DNA but by either the mother's or father's DNA. The DNA from one parent is silenced, not read or expressed due to genomic imprinting, parent-specific epigenetic changes of that DNA location.

Angelman syndrome and Prader-Willi syndrome are two examples in which genomic imprinting manifests in humans. Prader-Willi syndrome has captured some attention as many patients suffer from truly insatiable appetites that require their families to lock the refrigerators and remove all other food from the house.

Imprinting may play a role in other more complex disorders, including bipolar disorder.

Epigenome and the environment

What scientists are trying to understand now is to what extent the epigenome plays a greater role in humans.

The epigenome in our bodies is not permanent and, in contrast to our DNA sequence, is more malleable to change. One theory posits that this ability of the epigenome to change developed as a mechanism to assist us in adapting to our environment. For instance, if a child were born into a period of drought and starvation, there would be a benefit if the child were adapted to such an environment. Evolution, which occurs over many years and generations, however, would be of little use to this child.

Animal studies and human epidemiological (population) studies suggest that the environment during our time in the womb and shortly after birth affect our risk of developing various diseases later in life, leading some scientists to question whether the epigenome is to blame. Does the environment alter our epigenome and thus affect the long term expression of various parts of our DNA with repercussions later in life? To return to our example, changes made early on to adapt to starvation could be detrimental if the child is later exposed to plentiful food.

A well-known experiment in mice involves a location on the DNA in mice (called the Agouti gene) that plays a role in the color of their coat. In certain mice who have a variation of the Agouti gene, Avy, the color of the coat is related to the degree of epigenetic changes (methyl groups) that have been added in variable fashion during pregnancy, allowing mice born with identical DNA to have different coat colors, from yellow to brown. The extent of added methyl groups affects the expression of the DNA. Mice with few methyl groups added to this portion of DNA have yellow coats (and appear predisposed to obesity) while those with numerous methyl groups have brown coats.

What scientists discovered is that if they fed the mothers vitamins and other supplements considered “methyl-donors” (thought to promote the addition of methyl groups) during pregnancy, that they increased the number of brown mice born. This finding was true as well when the mothers were fed a phytoestrogen (a component of some plant foods that act like estrogen in parts of the body) called genistein, which at levels seen with high soy diets is also known to increase the amount of methyl groups added to DNA.

Epigenome and cancer

The epigenome may also have a role in cancer. There are certain parts of the DNA in every cell that promote growth (oncogenes) and that inhibit growth (tumor suppressor genes). In a well regulated cell there is a proper balance between these forces, but a tip in the balance of expression or lack of expression of certain genes can produce unregulated growth in a cell, or cancer.

Epigenome and behavior?

One mouse study suggests the potential inheritance of certain behaviors via the epigenome. In mice the mother's nurturing of her offspring (maternal instincts such as licking and grooming), is learned early after birth. (Mice that have been nurtured also appear as adults less fearful and better able to respond to stress.) Studies suggest that this nurturing occurs and is passed along due to epigenetic changes in the offspring during nurturing; the offspring grow up to be mothers and nurture their own offspring and continue the cycle.

In conclusion

The possibilities of epigenetics are eye-opening although a lot of questions still remain. To what degree is epigenetics, beyond a few examples, actually relevant in humans? If to a significant degree, then what in our environment affects the epigenome? Can we take advantage of this knowledge to prevent or treat disease? Eat certain foods or take certain vitamins? Or would such treatments be too indiscriminate and while addressing one disease possibly promote another?

For those interested in more detail:

Jirtle, Randy L and Michael K Skinner. "Environmental epigenomics and disease susceptibility," Nature Reviews Genetics, Vol 8 (April 2007), 253-262.

Tuesday, January 5, 2010

Arsenic, our friend? Arsenic as chemotherapy.

Recently, in the process of caring for a patient with leukemia, I was reminded by the oncologist of arsenic as a treatment option for certain leukemias. That's right, arsenic...

ar·se·nic (√§rs-nk)

  1. Symbol As, A highly poisonous metallic element having three allotropic forms, yellow, black, and gray, of which the brittle, crystalline gray is the most common. Arsenic and its compounds are used in insecticides, weed killers, solid-state doping agents, and various alloys.

Chemotherapy in today's age is to a large degree still poison - potentially effective poison in extending patients lives but poison nevertheless. Therefore, to be receiving arsenic, in the eyes of you and your oncologist the benefits outweigh known toxicity and risks. However, the idea of arsenic as a chemotherapy can be a tough one to swallow.

As far as medicine has come over the years (and it has become quite advanced in certain regards), medicine still has a long way further to advance – an idea that can be forgotten by the lay population. Our expectations in a society now with ubiquitous iPhones and Wifi internet on airplanes may not carry over well to medicine. Certain successes in medicine, such as pacemakers or heart transplants, may have helped create this impression. The reporting of medical news by the lay press often reinforces it, feeding unrealistic expectations by its tendency or bias towards portraying miraculous success stories, early technological breakthroughs, or early, intriguing but unconfirmed experimental results.

And then there is arsenic as an addition to standard chemotherapy or a second or third line agent in the treatment of various leukemias.

Arsenic in our environment is found in both natural and industrial sources but as it can be poisonous, particularly with chronic exposure (arsenic poisoning is a serious health problem in various part of the world), our exposure is regulated by the government. For example, the Maximum Contaminant Level (MCL) of arsenic in drinking water is currently set by the EPA at 10 micrograms per liter of water.

On a given day we may ingest up to a few hundred micrograms from our water or food. The amount of arsenic given with each daily infusion of chemotherapy to the average male (~70 kg in weight) might equal 10.5 milligrams (mg) – on the order of 100 times more than our daily intake.

Which is not to say I wouldn't take arsenic if it were indicated or recommended by my doctor given my medical condition. I merely think people might appreciate this perspective on medicine.

A similar example in cardiology and primary care

A similar example would be Warfarin - initially marketed as rodent poison but used in medicine since the 1950s and remains today the most prescribed anticoagulant in the United States with increasing numbers of prescriptions written to treat various blood clotting disorders, such as atrial fibrillation (a type of irregular heart beat), pulmonary emboli (clots in the lungs), and deep vein thrombosis (clots in the legs).

Can you think of other examples?

Thursday, October 22, 2009

Cancer screening

Finally... For those who haven't yet read this article in the NY Times, worthwhile reading.

Cancer Society, in Shift, Has Concerns on Screenings
Published: October 20, 2009