杏吧原创

A POPULAR type of DNA fingerprinting used by conservation biologists may give
false results with some species of animals, say researchers in Germany.

There are two sources of DNA in animals: the nucleus and the energy-producing
powerhouses of the cell called the mitochondria. Mitochondrial fingerprinting is
normally used when the genetic material in a sample is degraded, for while there
are only two copies of each nuclear gene in a cell, there are often hundreds of
copies of mitochondrial DNA (mtDNA).

In these cases, a technique called polymerase chain reaction, or PCR, is used
to amplify undamaged strands of mtDNA so that they can be analysed. PCR analysis
of mtDNA is increasingly used by field biologists to track the relationships or
numbers of animals in the wild. It is also used by forensic experts to track
criminals.

But Alex Greenwood and Svante P盲盲bo of the University of Munich
came across a potential flaw in the fingerprinting technique while studying
woolly mammoths. The team wanted to use tissue from hundreds of mammoth
carcasses found in permafrost to work out the relationships among the animals
and compare them to their distant cousins, modern elephants.

To start with, they took hairs from an Asian elephant (Elephas
maximus) at a nearby zoo and amplified a fragment of mtDNA. But when they
tried to sequence the products of the PCR, they found they were scrambled. 鈥淪o
we suspected we were dealing with a mixture,鈥� says P盲盲bo. The
suspicion was confirmed when they purified and sequenced individual DNA strands
from the PCR. They found four different classes of sequence.

When the researchers looked at elephant blood, however, they found only one
of the four sequence classes found in the hair. They also discovered that it was
the only active sequence of mtDNA, meaning that the other three found in the
hair were fragments of mtDNA that had migrated to the nucleus early in
evolution. This suggests that elephant hair has few mitochondria, giving a high
chance that PCR would amplify one of these 鈥減seudogenes鈥�, rather than the true
sequence. P盲盲bo points out that if no blood had been available for
analysis鈥攁s is often the case in forensics or molecular ecology鈥攊t
may have been impossible to sort out the true sequence.

Mitchell Holland, a DNA identification expert at the Armed Forces Institute
of Pathology in Rockville, Maryland, says the finding highlights the great care
that must be taken in the analysis of mtDNA data
(see 鈥淕hosts in the machines鈥�).
鈥淵ou can鈥檛 just get a PCR result and walk away from it,鈥� he says. But he
adds that nuclear pseudogenes are not a problem in human hair analysis. He says
that researchers can recognise the sequences of mtDNA that have migrated to the
nucleus, and that they rarely see them in their PCR reactions.

P盲盲bo agrees that 鈥渃ontaminating鈥� nuclear mtDNA sequences are
probably only an issue for particular tissues in particular animals. 鈥淢ost
sequences out there are probably fine,鈥� he says. 鈥淭his is just a warning to
people that you have to be aware of this possibility and be really careful.鈥� But
his plans to map familial relationships in mammoths seem to have been scuppered.
PCR amplification of mtDNA from the hair of one of these ancient beasts reveals
a variety of mitochondrial sequences as confusing as that in E. maximus.