Is it possible, either theoretically or even practically, to modify or engineer a pathogen to do one specific thing? Such as to kill or render infertile a more complex organism?
Example: are we years? decades? centuries? away from some lab in St. Louis genetically engineering a bacterium that could only multiply in the blood stream of the Asian carp (an invasive species wreaking havoc on the Mississippi River ecosystem)? That pathogen could be specifically engineered to kill or render the fish infertile, while having absolutely no effect on any other species that lives in and around the species’ habitat.
Is this possible even in theory?
The practical applications of this are limitless. Zebra mussels, kudzu, murder hornets… just about any invasive species could be destroyed once and for all (apart from where they’re not invasive, of course) with this technology.
Sure, you could in theory build something that specific. And it would immediately begin mutating beyond your control, losing that specificity, potentially hopping from secies to species.
More than in theory but the examples are not of the type you imagine.
Using viruses as vectors to deliver new genetic material into the cells of already grown organisms is a thing being researched and tested. It is called gene therapy. That is not really a pathogen that spreads freely and reproduces like you were imagining.
The FDA approved a genetic therapy that treats three kinds of tumors a couple days ago.
Some purposes are a lot easier to genetically engineer for than others. The easiest case is to engineer something to produce one specific protein. For instance, you can genetically engineer bacteria to produce human insulin. I’m pretty sure that this is in fact the primary source for the insulin used by diabetics, nowadays.
If your goal is to kill a specific species, like Asian carp, probably the easiest way to go about it would be to start with an existing disease that affects that organism, and try to modify it to make it more dangerous. But what makes a disease dangerous depends on a great many factors in a very complicated way, so probably the best you’d be able to do would be to generate many random variations and test them out in the lab, basically like a highly-accelerated version of natural evolution.
Chances are that some subset of the target species will be immune, or at least less susceptible, to your artificial disease. So you might knock the population of the invasive species way back, but then in a few years the immune ones will come back, meaner than before. OK, maybe not that last bit, but didn’t this happen with a rabbit disease introduced in Australia?
Before I retired I worked a lot with the weed scientists at Monsanto in St. Louis. We’re talking about chemicals, not organisms, but one thing they impressed on me was that nothing works 100% of the time. Somewhere out there are mutant members of a species that are pretty much immune to something. And they will reproduce, and then your sure method of killing everything will turn into survival of the fittest.
Yeah, this. Let’s not forget that COVID-19 was specific to something else (bats, apparently) and hopped to us probably as a consequence of a mutation that enabled the little strings to thrive in the human-body environment.
Both myxomatosis and rabbit hemorrhagic disease (calicivirus) have had this issue:
In 1997, RHD escaped a research station near Australia and killed a large number of rabbits on the Australian mainland. New Zealand farmers in the South Island defied the authorities and illegally imported and spread the disease at exactly the wrong time:
It may be possible to use a different strain of calicivirus that was released in Australia in 2017, but for maximum effectiveness, these things need to be researched, planned, and implemented well to have any hope of being effective.
Holy crap, the New Zealanders really shot themselves in the foot on that one!
O.P. it’s tough to locate a biological “target” that exists in that species (like your carp example) and only that species, and is also vulnerable enough to get close to 100% mortality rate if messed with.