"Heredity" and "inheritance" were once near-synonyms. Both referred to the physical goods an heir would receive upon the death of his (rarely her) parents or other testator. Carl Zimmer begins She Has Her Mother's Laugh: The Powers, Perversions, and Potential of Heredity by teasing out the history of these words, how they have changed through time. He takes off from there, producing the most amazing book I have read so far this year.
This book is quite a tome, at 600+ pages, and only the author's skill in writing terse, yet gripping prose, kept it from being twice as long. It is more than a review of genes, genetics and their history. We get a behind-the-scenes look at many of the players in the developments that became the modern cluster of "genetic sciences." I was particularly intrigued to learn that Gregor Mendel didn't just grow peas as a hobby, and then keep a diligent diary about it. Throughout Europe there had been "in the air" a rapidly growing interest in selective breeding, and he and his abbot decided on a research program to determine ways of improving the practice.
In five parts, the book draws together nearly twenty big themes, and how each was discovered and, effectively, turned into an engineering discipline. "Genetic engineering" has gone on since the first farmers selected only the grains that were biggest and most easily grown for their seed stock, and since the aurochs was gradually transformed into domestic cattle in a similar, but slower, fashion.
I remember reading a couple of years ago that we pass on to our descendants between 50 and 100 mutations, variations from the genetic baggage we received from our parents. Zimmer brings out something I had not taken thought to calculate: Each of our germ cells—ova for females, sperm for males—has some mutations that differ from every other germ cell. They share many, but each has some unique ones that arose when the individual germ cell was produced from the stem cell. Throughout our body, the total number of mutations may be dozens of quadrillions!
"Coat" (skin) color in humans is sometimes visibly mosaicked; a friend of my son in college looked like she had been splattered with gray-brown paint. But in every multi-celled creature, every time a new cell is formed by cell division, either during early development or from a stem cell, there is an opportunity for a DNA copy error of some kind to occur. When that occurs the new cell is genetically a little different from its "sister cells". But, the production of the stem cells also produces variants, so that all the descendants of such stem cells have the new variation. I have a number of brown spots on my skin (on average, people have 25 over their bodies, some less, some more, and some much more). Each grew from a single cell with a different expression of melanin. A key point of this section: Cancer is a mosaic expression.
We are all mosaics, of numerous characteristics, nearly all of them invisible. I find this astounding. Yet it shows the balance between perfect copying and occasional errors that characterize growth and development. For the DNA in a cell to be copied perfectly requires a stupendous level of accuracy, a level not seen outside computer technology. The human genome contains about 3 billion base pairs, meaning that when the DNA is "unzipped" for copying, there are 6 billion bases to be copied.
In a business context, you may have heard of Six Sigma Methodology (6σ), which aims to produce products and services with no more than 3.2 defects per million. There is a dirty secret to 6σ: a "process deviation" of 1.5σ is allowed, so the stated defect rate is actually 4.5σ. True 6σ is actually about one defect per billion. DNA copying, to be perfect "most of the time", requires less than one defect per 6 billion, or a production-perfection level of at least 6.28σ…with no "process deviation".
The perfection of copying computer files is about at this level, but only because of parity bits added to 8-bit bytes, which are actually stored as 10 bits, and checksums for larger chunks, which allow a lot of error correction, and detection of many errors that could not be corrected, so the software doing the copying can try again. Did you know that, when you get a hard disk failure, it is only reported to you after the software has tried 50 times without success to make a copy that passed all the checksum tests and other error-detection codes? Programs like CHKDSK look for portions of the hard disk that cannot yield perfect copies and marks them so the software will no longer try to store data there. The cellular mechanisms for DNA copying include methods that are similar in philosophy, though much different in implementation.
So, more often than not, when one of your cells (or a cell in a cat, horse, tree or blade of grass) divides, the result is a perfect copy. Some percent of the time, a small copy error occurs. Some smaller percent of the time, a larger error occurs, like getting a snippet of DNA turned around or copied twice…and there are other kinds of errors. Some errors are bad enough that the new cell cannot function and so it dies.
But there is another level of "error tolerance", in the DNA-to-protein code. Each amino acid has more than one code associated with it. Some important ones have as many as six, among the 64 3-letter codes that represent the 20 amino acids plus START and STOP. So certain small errors make no difference at all in the protein produced. Such "silent mutations" accumulate, and form the "DNA clock" used to determine how closely related one species is from another, and how long ago the two species split from a common ancestor.
Once in a while a mutation that is not silent actually makes the cell or the organism function better. Such beneficial mutations also accumulate, and over time can lead to new species. Non-silent mutations that are not beneficial may lead the cell or the organism to die, and are thus instantly weeded out, or they may handicap it to some extent, making it less likely to reproduce. This combination of a steady, but low, error rate, and an environmental filter on what is and is not beneficial, is the mechanism of natural selection.
So, back to the thesis of the book, at least in its earlier parts, we find that genetics includes a certain tension between near-perfect copying and the weeding out of most, but not all, copying errors. If copying were perfect, nothing would change (this is the Creationist view). But if copying had been perfect from the first dog, the first apple tree, or the first human, every member of each "kind" would be exactly the same, without variation. You and I would look identical (unless you're female, then you and every other woman would look identical). There would be no moles or other kinds of pigmented spots on our bodies. There would be no mosaicking.
On the other hand, if copying were too sloppy, a large fraction of pregnancies would terminate early in miscarriage, a huge load of birth defects would occur, and everything would die out. Actually, with a little luck, natural selection would drive the surviving creatures in the direction of more perfect copying, until a balance, such as the present balance, were achieved.
The later parts of the book present the several stages of "genetic engineering", including stories of the triumphs and disasters along the way. We are friends of a family named Gelsinger, and a cousin of theirs was Jesse Gelsinger, the young man who died at age 19 of a huge allergic response to the virus that was being used in an experimental genetic therapy for his chronic condition. That particular kind of "gene therapy" came to an end then and there.
Now there is Cas9/CRISPR, and some allied methods (some natural) called "Gene Drives" (only one of them makes use of CRISPR). If these methods fulfill their promise, DNA will become as editable as a Microsoft Word document. Carl Zimmer is guardedly optimistic about the possibilities, but this is the source of the word "Perversions" in the book's title. The genie is already out of the bottle. Not everyone who works with these new technologies is righteous. None is wise enough to think through the implications. After all, less than a generation has passed since Jesse Gelsinger died, and he died primarily because the researchers had not expected allergies to be a problem. Well, Duh! A good friend of mine is most likely to die of taking a breath at the wrong time, in the presence of people eating peanuts. A woman we knew well some years ago came into our house after we had sprayed an insecticide in a back bedroom. She almost collapsed, and had to be half-carried out of the house to fresher air. We could visit her, but not the reverse, thereafter.
Will the new genetic tools somehow "get out" and wind up blasting the biosphere with the biological equivalent of that (so far imaginary) nemesis of nanotechnology, Gray Goo? Maybe. It cannot be ruled out! However, the hope I see, and I think that our author sees, is that there is so much variety in nature, so much variation in the 7 billion of us also, that any biological Gray Goo will not affect everyone. One of his correspondents told him that there have been gene drives unleashed in natural ways in the past, but that the biosphere has eventually deactivated them all.
I have just skated on the surface of a couple of the ideas in Her Mother's Laugh. This book contains more ideas per column-inch than I know what to do with. A tour de force.
No comments:
Post a Comment