The Instapundint has written a TCS Daily column on the biowarfare column and links to a bunch of other biowarfare warnings such as Paul Boutin's "Biowar for Dummies" and a Technology Review piece. The latter has a rather scary 2-stage proposal which I think is a bit far-fetched, although doable:
Popov then described a Soviet strategy for hiding deadly viral genes inside some milder bacterium's genome, so that medical treatment of a victim's initial symptoms from one microbe would trigger a second microbe's growth. "The first symptom could be plague, and a victim's fever would get treated with something as simple as tetracycline. That tetracycline would itself be the factor inducing expression of a second set of genes, which could be a whole virus or a combination of viral genes."
However there are, I fear, a lot of other far simpler 2-stage bioweapons that could be created.
Firstly a disclaimer: I am merely (like Glenn) an informed amateur in the biotech field but I have a reasonable understanding of where we are in terms of the slicing and dicing of genese and particularly the genes of bacteria and similar unicellular organisms as I used to work for a company that did this commercially. It is entirely possible that I am wrong and that even if right that there are some wrinkles that make my schemes impractical for actual terrorists, but given that you will need a couple of good biochemist PhDs to do this stuff that is not much of a defence because such people can probably figure out where I'm wrong in the details. However I'm publishing this not to show how clever I am but because there ought to be some thought given for ways to stop this kind of thing. Essentialy a biotech arms-race.
Today biochemists routinely modify the DNA of bacteria such as e.coli to make it express a gene to create something useful such as a particular enzyme. Quite often the enzyme is one that originates in an entirely different bacterial species such as one living in a hot sulphur spring (e.coli is a bacterium of the human gut - an environment that is very different from a hot sulphur spring even if you have heartburn). Without being too specific it ought to be possible to produce GM e.coli that produce almost any arbitrary enzyme and (potentially) chemical. Much the same could be done to other bacterial species that are well known such as those that cause common bacterial infections.
One of the problems that pharmaceutical companies face is that the human body is quite good at keeping nasty stuff out. The cells and beceria flora in the gut do an excllent job of stopping many larger molecules (such as complex enzymes or other proteins) from entering the rest of the body - the reason why BSE is so scary is that the gut seems to have a problem breaking down the prion that is its cause into something simpler - and there is another stronger barrier between the brain and the rest of the body. However neither of these protections nor any of the others are perfect - if they were we wouldn't catch all these bacterial diseases, let alone the viral ones - and indeed it would seem that the fact that a number of bacterial species are good at either living in humans (e.coli and friends) or invading (MRSA and co) would make such speices ideal for the first stage of a biowarfare attack.
Let's go with the e.coli version. If you want to kill someone then a really good way would be "food poisoning" by means of e.coli. The GM e.coli would have been podified to express a gene that produces an enzyme (a nitrilase perhaps) that creates hydrogen cynaide from some fairly common food product such as potatoes. Thus after this bacterium has established itself in your gut it will divide and divide and each bacterium will be producing lots of cyanide enzyme. Then at some point you will get a die off of a significnt number of these e.coli and their cell walls will crack - a cunning bioscientist might be able to make this happen after a certain time such as the 1000th cell division, others might just decide to wait for nature to act by chance. At this point anyone who has some cooked potatoes in their stomach will get a stomach full of cyanide and die.
How would you distribute this bioweapon? Pick something that is unikely to be cooked or frozen (tuna for sushi? salad?) and spray a load of GM e.coli over it. Do this at a wholesale market or at the kitchen of a large catered event - think of all those rubber chicken fund-raisers beloved of politiicans - and you'll get quite a nice body count. More to the point do it "al Qaeda" style simulateneously to multiple food markets and hotels and you'll get a very large number of deaths before someone figures out what the attack vector is and even then potentially many many more before someone figures out how to sterilise food before serving it and without cooking it. This latter problem is one that explains Montezumas revenge and all those other food poisoning incidents. Cooking food really is a good idea if you want to kill lethal bacteria but it does ruin the taste of all sorts of food.
Finding other equally good attack vectors is left as an exercise for the reader.
The important point here is that while the required steps (the isolation, cloning and overexpression of genes that produce a particular enzyme and moving them from one bacterial species to another) are specialist tasks each isn't that difficult. OK so there is a good deal of relativeness in that statement: a good group of scientists and lab-techs will move from identification of the critcal gene toits successful cloning in e.coli in 6-8 weeks while a bad group can spend 3 years on the same task. However a three year project could easily be handled by some terrorist setting up a fake startup or subborned university research project.