[PSA] Biohacking FAQ #2: "But what if you end up making grey goo?"

[maradydd]

Over the last 24 hours I've seen a lot of concern and speculation about what happens if one of my experiments somehow "goes out of control" and turns into some kind of "grey goo" event. It seems that there's a mistaken impression that I'm just randomly mutating things (perhaps with UV stimulation) to see what comes up. This actually couldn't be further from the truth, so let me explain what I'm really doing.

How Your Genes Work can be summed up in a single sentence: "DNA makes RNA makes protein." Your genes are instructions for making several different types of RNA, and those RNA molecules assemble the proteins that your body is made of and which make your body run. Some proteins are structural, some are enzymes used to catalyze chemical reactions (such as digestion), some are used to transport other molecules around (e.g. hemoglobin, which carries oxygen around in your red blood cells) -- proteins are everywhere. So, when I think about something I'd like for a cell to do, I start looking around for relevant proteins.

In the case of "let's detect melamine", I went to MetaCyc -- a browsable database of metabolic pathways -- and looked for proteins which interact with melamine. I found one, called melamine deaminase. It's the beginning of a metabolic pathway called the melamine degradation pathway, which -- go figure -- takes melamine apart. To use this reaction in our detector, we'll need to give some species of bacteria the ability to produce melamine deaminase, which means giving it the appropriate gene. To do that, we either extract the gene from a species that already has it, or we get a lab like IDT to make it for us. Then we insert the gene into a plasmid, which is a circular DNA molecule that a bacterium can "take up" in order to gain some new function.

So, no, there is no deliberate randomness going on here -- rather, it's a concerted effort to make just one type of bacteria do just one additional thing (or, really, some sequence of additional things). The whole experimental setup is also designed so that if I screw something up, the bugs die and that's it. And, naturally, I'm doing everything I can to make sure that stray spores, phages, and other contaminants don't end up in my experiments -- heat sterilization, alcohol sterilization, flame sterilization, you name it.

Do you need to worry about these synthetic bacteria degrading you? Only if you are a whiteboard or certain species of plastic fork.

Comments

[3-2-1-letsdance.livejournal.com]

I'm not a proponent of "gray goo" hysteria, and I'm only asking out of curiosity, but is there any chance that adding a gene to a bacteria will cause the other genes to change in unexpected ways?

[maradydd.livejournal.com]

Second- and third-order effects are always difficult to predict, but this is a good question. It largely depends on what new functionality you add. If the function you're adding has some interaction with some other metabolic pathway already in the bacterium, then that certainly applies some pressure. For instance, if the function you're adding provides some simpler route for producing something the bacterium needs to survive, then the less efficient way might evolve out. (There are a lot of speciation events like this.)

Thus far, nothing I'm working on has any real likelihood of doing this, and the odds are actually quite high that the functions I plan to add will evolve out, as they're not of any particular advantage to the bacteria themselves.

[3-2-1-letsdance.livejournal.com]

That's pretty cool, and thanks for the explanation :)

Another question! :D
Is there any particular place that you put a gene? Do you stick it at the front or the end? Or somewhere in the middle next to other specific genes?

[bigby.livejournal.com]

If memory serves, you dope the plasmids in the bacteria's path and it adds in where the whatever the bugger's DNA and RNA process it to.