My latest Mind and Matter column in the Wall
Street Journal is on junk DNA and on the messed up genome of the
The usually placid world of molecular biology has been riven
with two fierce disputes recently. Although apparently separate,
the two conflagrations are converging.
The first row concerns the phrase “junk DNA.” Coined in
1972 by the geneticist Susumu Ohno, it is an attempt to explain
why vast stretches of animal genomes, far more in some species than
in others, seem to serve no purpose. Genes of all kinds and their
control sequences make up maybe 9% of the human genome at the very
most. The rest may be nonfunctional “junk,” mainly there because it
is good at getting itself duplicated. Yet the phrase has always
caused a surprising amount of offense. Reports of the discrediting
of junk-DNA theory have been frequent.
Why does it matter? Partly because scientists want to know if
they are right to focus on part of the genome, ignoring the rest,
but mainly because the issue tests an evolutionary theory about how
DNA sequences can proliferate even if they do not benefit the
Late last year, a huge team of scientists running a consortium
called Encode published an analysis of the human genome that they
said showed some kind of activity in 80% of the genome. They later
conceded that perhaps 20% is actually functional, yet insisted the
phrase “junk DNA” could now be “totally expunged from the
According to Dan Graur of the University of
Houston and his colleagues, even this is a wild overestimate—not
least because it uses a “causal role” definition of function that
is all wrong, as if you were to describe among the heart’s
functions adding 10.5 ounces to the weight of the body, along with
pumping blood. After a few exchanges, the Encode team leader Ewan
Birney conceded that in hindsight, the team
overstated its conclusions. But he added that whatever the
interpretation, the Encode data are sound.
Are they? Here’s where the junk-DNA row meets the other
conflagration in molecular biology. All the Encode data were
derived from cancer-cell lines. To describe human cancer cells as
having the human genome looks increasingly unwise. Most cancer
cells have extra chromosomes, fragmented and rearranged DNA and
unusual patterns of gene activity.
As if to illustrate the point, last month a consortium of
scientists based in Heidelberg, Germany, analyzed and sequenced the genome of one type
of HeLa cell, an immortal laboratory cell line widely used since
1952. They described a genome that looks like a bomb has
gone off in it. There are three copies of most chromosomes, yet
only one copy of many genes. Hefty chunks have been reshuffled to
other chromosomes, and some chromosomes have suffered from
“chromothripsis,” which one of the scientists describes as being
“blown apart and stuck back together in a random order.” A person
with this genome could never be born.
Yet—and here’s the source of the controversy—the HeLa cell line
was derived from the tumor that killed a poor, black tobacco farmer
named Henrietta Lacks in Baltimore in 1951. As Rebecca Skloot has
documented in a remarkable best seller (“The
Immortal Life of Henrietta Lacks”), the medical community never got
her consent and treated her family with tactless disrespect for
years—until Ms. Skloot’s book began to make a difference.
Not enough of a difference, apparently. The German team did not seek the
consent of the Lacks family before publishing the HeLa sequence,
claiming it revealed nothing specific about Ms. Lacks’s own genome.
“Your claim is so wrong that I don’t know where to start,” replied
one geneticist. The sequence has since been unpublished.
So here’s the paradox: A cancer genome like HeLa may not be
sufficiently representative of human genomes to resolve the junk
DNA question, but may still give away private information about the
human being from whom it derived.