This essay was inspired by a discussion on alt.polyamory, and was originally posted there. The libraries referenced are at the University of Texas, where I got my biology degree. I'm not there anymore, so I can't go look up the specific references again, but if you have any questions about anything relating to this essay (or, heck, just random biology questions), do please e-mail me at I like explaining things, honest I do.

Cloning Extinct Species
by Theresa Mecklenborg

All right, so I went to the libraries yesterday (life science and geology -- and I need to spend more time hanging out in the geology library, yes I do) to see what's been done lately on this whole topic. So that was an educational few hours, the fruits of which I shall now share.

The first problem with cloning an extinct organism such as a dinosaur or passenger pigeon is acquiring a complete, undegraded genome. The trouble is that DNA's not very stable, and degrades pretty quickly via a number of interesting mechanisms. For recently-extinct things (passenger pigeons, quaggas, things we've got preserved tissue samples of), it doesn't look like that's a huge problem, but nobody's tried it yet. As far as I could tell from just a quick scan (and I would love to hear from anybody who knows of an actual attempt / firm plans for an actual attempt), the closest anybody's come so far is some vets who are contemplating attempting to clone a certain subspecies of Pyrenean mountain goat, the last known member of which was killed "when its head was crushed under a falling tree" (1). Fortunately, they'd taken ear tissue samples the previous year. That report's from the beginning of 2000, and I don't know if they decided to go ahead with it or not. I'm fairly sure there haven't been any cloning attempts at all done with birds, but again I'd love to hear differently.

So, getting DNA from an older sample (i.e. not deliberately preserved). First, a word about the terminology: when a molecular biologist says "cloned", he or she means only that some piece of DNA was replicated, not an entire organism. So when you see reports saying such-and-such a gene was cloned, that's what that means -- we made copies of this particular gene, and only this particular gene(*). This is usually accomplished via a technique called PCR (polymerase chain reaction), which involves taking short pieces of man-made DNA (primers) and using them as a starting point to copy a piece of unknown DNA. (If anybody's interested, I will be glad to explain in more detail.)

Lots of genes from various fossils, including human ones, have been cloned. However, there are a number of problems plaguing even this much information retrieval. (And remember, you need lots and lots and lots of genes, in the right order, pretty much undamaged, to make an organism. You can copy much smaller pieces of much more badly damaged DNA.) There are three main issues: preservation, contamination, and amplification problems.

DNA is, as I mentioned, pretty fragile. Once the cell is dead, it no longer has the machinery that normally protects it from damage and repairs any damage that may occur. The major causes of damage to ancient DNA are microbial damage, oxidative damage, base modification, disintegration of strands into short fragments, and crosslinking with itself or with proteins (3,4). Although estimates of best-case DNA survival vary, there seems to be general agreement that "no meaningful genetic information should be preserved for longer that 10^4 - 10^5 years" (4). Even this length is only possible in a favorable environment -- if you dug up the body of a roadkilled squirrel from your backyard, you'd be hard pressed to get anything really useful out of it. Archaeological (that is, recent) specimens come up with DNA "in tiny amounts, severely damaged and often contaminated by microbial or fungal DNA"(2). So far, it looks like the best samples have come from specimens frozen in permafrost, although swiftly desiccated mummies might be another possible source. If you're looking for dinosaur DNA, though, you're not going to have much luck because the climate's changed too much since then. It looks like the oldest DNA specimen that's been found so far was from a wooly mammoth that lived about 50,000 years ago (4, but I really should go find the original reference).

DNA detected from older samples (and some has been) has so far been found upon closer observation to be more likely the result of contamination from humans, bacteria, or other samples in the laboratory used to amplify the DNA. This is because those nice fresh pieces are in much better shape and present in *much* greater concentrations than any original DNA that may have survived.

This leads us into the last, and most pernicious, problem with recovering ancient DNA. PCR, while absolutely great at making lots and lots of copies of a single short stretch of purified DNA, doesn't work so well on a mixture of lots of different pieces of badly damaged DNA present at very low concentrations. First, you have to use primers (short pieces of DNA you provide) to get it going, so you can only copy pieces of DNA that fall between stretches you already know. Damage to the DNA can halt or retard the PCR process. It can also cause errors in the cloned DNA. Also, for some peculiar reason it's apparently really hard to get PCR to copy a heterozygous allele (that is, a gene which you have two different versions of). I didn't realize this, but it seems that the PCR attempts to combine bits of DNA that are sufficiently similar, so it does weird things when presented with two versions of a gene (3). This, plus the much greater concentration of mitochondrial DNA (mtDNA), and the fact that mtDNA tends to be pretty similar from organism to organism so you can use known primers, means that it's a lot easier to copy mtDNA than nuclear DNA. Some work has actually been done tracing human evolution/dispersal using mtDNA from fossils, since it's relatively easy to produce and only present in one version per cell (**). Unfortunately, what you need to clone an entire organism (as opposed to some little stretch of DNA), or even to attempt to study individual genes unique to that organism, is precisely the part that's hardest to copy.

Even if you did somehow manage to find a completely preserved genome (which is possible, though unlikely), you'd still have problems. You'd have to get an appropriate sort of host egg and host mother to put it in (easy for passenger pigeons or that subspecies of goat, a bit trickier for mammoths [elephant] or quaggas [probably a horse], pretty damn tough for something like a velociraptor [ostrich, maybe?]), and because of all that developmental stuff I kept harping on in my first post, you might or might not end up with what you wanted. But it'd still be amazingly cool, and it's certainly worth a shot. The limiting factor at this point really looks like it's finding sufficiently well-preserved nuclear DNA. The technical issues don't really look that overwhelming, but there's not a lot we can do if there just isn't the starting material to work with.

(1) Anonymous. Feb. 2000. Nature Biotechnology 18(2). 133
(2) Audic, Stéphane and Eliane Béraud-Colomb. 1997. Nature Biotechnology 15(9). 855-858
(3) Kelman, Lori M. and Zvi Kelman. March 15, 1999. Journal of Vertebrate Paleontology 19(1). 8-20 [This one is great, by the way. It explains things on a pretty non-technical level, since it's aimed at paleontologists rather than molecular biologists. Laymen, practically! :) ]
(4) Poinar, Hendrik N. and Artur B. Stankiewicz. July 20, 1999. Proceedings of the National Academy of Sciences of the United States of America 96(15). 8426-8431

(*) Er, mostly. You're not necessarily copying a piece that's broken exactly on the gene boundaries.
(**) Barring weirdness. This sort of research also gets done with genes on the Y chromosome, again because there's only one version per cell.

Go back to the biology index . Go back to the main index .