Will Dinosaurs Ever Walk Again?

You can always get them back !

 -- The Wicked Witch, "Chronicles of Narnia".

We've all seen or read Jurassic Park. And many of us have witnessed the seemingly inevitable conflict which has been waged between scientists in its wake. On the one hand, we have scientists predominantly in the UK who say it is impossible to revive dinosaurs because the degenerated state of their DNA is beyond all help, while on the other, we have a group of scientists in the USA who actually claim to have a live culture of  T.Rex body cells.

So what is the real situation regarding the question of dinosaur revival?

Well, it is possible to bring back dinosaurs but there are also problems involved in this which most of us would, for better or for worse, regard as formidable.

If you want to make a dinosaur, there are three main problems. The first is the sheer dilapidated condition of any DNA which might remain after such a vast length of time, and its rebuilding. The second is the actual creation of a dinosaur from a test-tube of liquid DNA, and the third is the fact that you probably won't know whether your DNA really is dinosaur DNA until you've already spent all your money, done all your work, and reached the final point when the egg is supposed to hatch out. Let's look at these situations one by one.

First, the primary* accepted scenario whereby any dinosaur DNA might become available to the scientific community is the one outlined in the Jurassic Park novel. This involves the DNA originally being ingested into a parasitic insect, which sucked the blood of a live dinosaur before then becoming trapped in tree sap which eventually fossilized to become amber. About one per cent of all amber finds contain an insect. These are readily available, and very cheap; they can be picked up online for just a few quid (however most of the stuff on the market is Tertiary amber which dates from after the dinosaur era). It is of course possible, and actually likely if you accumulate enough amber, that there will be present in these insects the DNA of real dinosaurs.

However, the likely condition of this DNA is entirely another matter. The trouble is that after death the DNA in a creature begins to disintegrate. What happens is that cosmic rays, the particulate radiation from stars which explode as supernovae at the end of their lives, provide a constant background radiation flux which shoots pieces out of complex molecules like DNA -- rather like firing a hunting rifle at a ping-pong ball model of a molecule. Of course, living bodies repair the damage, their healthy immune systems spotting the harm and quickly fixing it -- but in dead specimens the damage can only accumulate until the DNA begins to look very sorry indeed -- literally shot to pieces. Further, the action of running water can also remove DNA by literally washing it out.

However, Saxon and Roman remains in the order of 2000 years old can still have their DNA in a pretty fair condition, but unfortunately the record for the oldest more or less intact DNA yet extracted from animal remains doesn't go back much further than that about 17,000 years, which hardly compares with the 66 million years which have passed since the last of the dinosaurs, and so any DNA from the dinosaur era is bound to be in very bad condition indeed. In fact, no strands of dinosaur DNA longer than 100 bases have ever been found. The expected length of a complete strand on the other hand would be in the order of 1 billion bases. Therefore, we have to face the basic fact that any dinosaur DNA which me might extract from an insect is bound to be in need of some very considerable repair.

However the good news is that we can repair it. Fortunately the DNA molecule is made up from only 4 different bases, called nucleotides. These are called adenine, cytosene, guanine and thymine.  Let's call them A, C, G and T. Now DNA is a "double" molecule, with two long base sequences laying side by side and bonded together laterally (the whole thing is also wound into a double helix shape, but that doesn't concern us here). Now it so happens that there is a hard and fast rule relating to the replication and bonding of DNA strands. A will only bond with T, and G will only bond with C. This is called base pairing. Here is a typical strand of DNA:

DNA strand. Notice how cytosene (C) only pairs with guanine (G) and adenine (A) only pairs with thymine (T). 

Therefore, if we have a DNA fragment extracted from our insect like this, (as read by a PCR Primer machine, a machine which can read the nucleotide sequencing of fragments of DNA and also replicate them millions of times)

ACCTGGCATGCCAATGCTAACT

a second fragment like this,

CCAGTGTACGATAGATACTAGA

and a third like this,

ATTGAGGTCA

then a computer will at once tell us that the second strand used to fit onto the end of the first one, because the third is a bonding strand which connects to the join between the two. Like this.

The computer notices that there is a fit between the end of fragment 1 and the start of fragment 2, as fragment 3 fits across the gap according to the rules of base-pairing. If all 3 fragments are long enough, this fit will not be coincidental, meaning that fragments 1 and 2 originally fitted together. Further, we can now recreate the entire sequence on either end of fragment 3 exactly as it must have been -- again from the rules of base-pairing.  In this way, even if some fragments are missing, we can accurately build up the entire genome again -- irrespective of how badly damaged the DNA was in the first place.

But so astronomical is the number of possible random combinations of A,C,G and T that nowhere else along the entire genome (complete DNA strand) is the above bonding strand sequence (if it is long enough) likely to be repeated. Therefore we can be sure that these two fragments do indeed fit together. We have just identified, and repaired, a break.

In this way, we can eventually read every single fragment of DNA which we extract from our insect and fit them all back together. But suppose the computer tells us that some of the fragments and their laterally paired fragments are both missing? Then, we will have to use guesswork, which will be largely based on comparing those parts of the genome with the same places in the DNA of modern reptiles and, in particular, birds. This could of course be a bit dodgy, but thankfully it is thought that the vast majority of the DNA nucleotides in a genome are not actually used for anything anyway, but are instead just redundant genes from the past which are no longer used. The similarities between the DNA of various reptiles and of birds are however so great, that any guesses we have to make will probably be pretty safe--but even so, I still don't like the thought of having to guess at all!

Also, we could recreate the chromosomes by assessing the overall sequences; for example we know that the base signature on the end of each gene is ATT and therefore we can identify the individual genes in what would otherwise be a rambling overall sequence, and we know how many chromosomes modern reptiles and birds possess and how many genes there are in each.

Eventually, then, we could arrive at a complete set of DNA chromosomes extracted from our insect. There is thankfully little chance that any DNA ingested by the insect from other creatures, or any of the insect's own DNA, will have mixed in with our recreated DNA, because none of these alien base-sequences will fit alongside our recreated DNA. The chances against their doing so by accident will be remote. But we must remember, though, that we still have no proof that the DNA we extracted from inside this insect, and which we are putting together, was from a dinosaur in the first place!

OK, so we have managed to resurrect ourselves a completed sample of ancient DNA. We have it in chromosome form too, and therefore we have the overall DNA packaged in a form pretty similar to how it would have been in nature. But to have this stuff swishing about in a test tube is one thing-- how on earth do we now take this liquid and turn it into a dinosaur? That is our next problem.

Well, we use the principle of taking an ostrich egg and replacing the ostrich DNA in the nucleus of the egg with our dinosaur (if it is dinosaur) DNA. Then, administering to the egg a small electric shock to give it a "kick", we place the egg back under the mother ostrich. At this point, it is pretty tempting to say, "I'd love to see the ostrich's face when the egg hatches out!". ! But in actual fact, it won't. Hatch out, that is.

This is because there is in the "white" of the ostrich egg another type of DNA called mitochondrial DNA. This DNA has a function to do with telling the nuclear DNA (the DNA in the nucleus of the egg) where to go and what to do, in an architectural sense, to actually build the animal. But the problem is that our egg only has ostrich mitochondrial DNA, and we never had any dinosaur mitochondrial DNA to begin with. Our overall problem is very like the old "chicken and egg" situation -- you need the one to create the other. But we started with neither. No dinosaur and no dinosaur egg. Thankfully, however, there is a way round this.

Start an ostrich farm, and take the first generation of ostrich eggs which are forthcoming. Now, using genetic engineering, replace various random strands of their ostrich DNA with similar strands from the dinosaur DNA. When we do this, probably not many of the eggs will hatch, and most that do may hatch animals which will die or be infertile. But a few may hatch healthy chicks which will have inherited a few of the characteristics of the ancient animal. We then repeat this process through successive ostrich generations, adding further, different ancient DNA fragments at each stage, until the ostrich becomes decidedly dino-ostrich looking.  Now the object of this is not to create a dinosaur by the gradual replacement of the ostrich DNA with dinosaur DNA; it is to create ourselves an egg whose mitochondrial DNA will be compatible with our original dinosaur DNA. 

Then, when we believe we have reached this level of compatibility, we take an egg from the latest generation of dino-ostriches, extract all the dino-ostrich DNA from the nucleus of the egg and discard it. And then, replacing this with the entire, original, recreated dinosaur genome, the egg should hatch out whatever creature it was, which owned the original, ancient DNA.

But only then will we discover whether this original DNA was indeed from a dinosaur (although some of the monstrosities created on the ostrich farm might already have given us a pretty shrewd idea).

This third problem, of not knowing until the very end of the process whether or not we even have a dinosaur, is of course a very serious practical-- and economic--  problem. In the actual project to rebuild the original DNA from the fragments recovered from the insect, we shall have to use thousands of PCR primer machines to read the base-sequences of the hundreds of millions of fragments, and then reproduce each fragment many millions of times to create tangible samples of it. Each of these millions of samples will then be held in a test tube. So we can envisage a building the size of an aircraft hangar whose floor is covered with an ocean of test tubes, each numbered and coded. Over these there will have to sweep a series of robotic injector/extractor machines, because the next stage will of course be to mix together the fragments adjacent to each other in the genome and then mix with these the bonding patch fragments too. So we will have computer-controlled robotic injector/extractors rushing over an ocean of test tubes removing from here and injecting into there, as the computer orchestrates the vast resynthesis of the entire genome.

But....can you imagine anybody, let alone the taxpayer, paying for all this when we don't even know that it was a dinosaur that the ancient insect bit in the first place? And that, in a nutshell, tells you from a purely realistic viewpoint just how far in the future this project is, because it will certainly not be paid for by taxpayers money and no individual company is likely to be big enough.*

Of course, as time and science progress, then PCR primer machines will become very much faster, cheaper and vastly more efficient, until eventually a stage will be reached when some suitably enthusiastic group will at last venture the cost. Fortunately, these PCR primer machines are presently developing very fast. But even allowing for their future capability projections, I feel it would still not be realistic to expect this project to begin much before 2050. Then of course there will be a great many project failures, and after the first successes we must further add many years for the animals to grow up (not catered for in Jurassic Park), before we finally find ourselves confronting the sight of live, full-sized dinosaurs.

In the scientific, logical, and commonsense order of things, therefore, I think it probable that Mammoths, (and even Neanderthals), whose DNA is in far better shape due to their much more recent age, will walk the earth again before we finally feast our eyes upon the likes of Tyrannosaurus and Diplodocus, perhaps towards the last quarter of this century.

--- Michael Alan Marshall

* There is however another way of obtaining Dinosaur DNA -- and using this method we will actually know from the start that our DNA is from a dinosaur, and we will even know the particular species of dinosaur. This is by examining thousands of well-preserved fossil samples of a particular dinosaurian species with an electron microscope and using an imaging computer to identify the individual DNA nucleotides by interpreting the shadows in the image. Fortunately there are only 4 different nucleotides in the entire genome, and so the computer could become quite good at this. The overall task does however seem rather daunting. If it is achieved, though, then the next stage will still have to be to use thousands of PCR primer machines to synthesise the genome from many fragments (the fragments will be created by the PCR primer machines from small DNA sequences provided by the computer which breaks up the overall DNA genome sequence in its memory in order to create small sequences which the PCR primers can handle), and then mixing these with robotic injector/extractors as before to build up the actual genome. This method hardly constitutes a short cut -- but at least we will always know that the DNA we have is that of a specific species of dinosaur, and this will almost certainly be a determining factor in any decision to stump up the cash to fund the project in the first place.

Desiderata Curiosa