Open Left

Oversimplifying (0.00 / 0)

You rather overstate, yourself, biologists' understanding of how all this works together.

We know a lot, but not that much, about the multiple interactions in a biological system. Nutritionists can barely even tell us what's healthy to eat - because they don't know except by observing outcomes.

Blueberries are good for you, as an example. This is well known and uncontroversial. But why? No one really knows the answer to this. If you take vitamin supplements containing all the nutritionally equivalent constituents of the blueberry that have been identified as healthful, you still can't replicate the benefits of a handful of fresh berries.

How can we claim that something absolutely isn't disturbing some property of a complex organism that we don't even understand well enough to measure?

And though you say that scientists understand the one gene, one protein model is overly simple, that there are multiple interactions, again, they don't know what all of them are. The prediction of protein folding configurations is a "conditionally intractable", np complete problem. Genes make proteins, but which ones, and how are they shaped, and what do they do? Anyone who says they know is just making stuff up.

Maybe this frankenfood is fine, though, after all. But I don't want to have to eat it until there's real safety testing done on it.

---

by: Natasha Chart @ Tue Jun 09, 2009 at 04:16:54 AM CDT

[ Parent | ]

don't really disagree (0.00 / 0)

with what you are saying here. And I'm not sure why you  think that what you wrote disagrees fundamentally with what I wrote. There are several things I will say in response, but I can't necessarily fit them into a coherent structure so I am just going to list them and apologize for the disjointed nature.

1. Let me just say up front that what we don't know about the fundamentals underlying biology is sooooooooo much more than what we do know. Thats what makes it so much fun to study!!!

2. The protein folding problem can really be divided into two questions:
a) how do you take a primary sequence and fold it into a protein from scratch using basic principles.
b) If you can compare a given sequence to those of all the proteins who's structure is known, can you use this knowledge to predict the folded structure of a protein.
These are very different questions and protein folding contests split them out into different categories. There is also a fair amount of cross disdain between the two camps (or at least there was 10 years ago when I was more closely involved in the field), which can be very entertaining at talks.

Anyway, if the question is:
Given a protein, can we predict its structure? maybe
Can we determine it's structure? most likely
Is it possible that if it's expressed in a different organism, could it be misfolded? absolutely
If it's expressed in a different organism, can we determine with a high degree of confidence that it has the same structure? yes
Can we be sure that expressing the protein isn't altering the folding (or some other aspect) of another protein? Absolutely not. BUT this is true of every genetic change you make.  The crosses that breeders do introduce millions of genetic changes, and we have no idea what most of them do (actually, most of them likely do very little, but that still leaves many, many that do something and we don't know what it is).

So its basically impossible to be sure that expressing a gene, or making a cross, or collecting seed from a plant and replanting it, or a change in the the weather during the growth cycle won't make any changes to any component of the plant.  But as you point out, we really don't know much about what changes matter for nutrition. So your left trying to do testing to figure out if the changes you have made have a detrimental impact on the environment or the people eating it.

So that leads to two questions:
Do you test every new line, or just the transgenics?
and
How much testing do you need to do to declare something safe?

For the first, I think you could reasonably state that you should test each line that is radically different from a line that people have been eating for awhile. But that would include a lot of lines that conventional breeders produce as well. One of the key to the green revolution was breeding dwarf plants, which is screwing up with hormone signalling (plant hormones, not animal). This will cause all kinds of changes throughout the plant. If you take those lines bred in Mexico and cross them into African varieties, you are going to get something drastically different.

The second gets much trickier. Clearly we should do basic tests to make sure that foods aren't immediately toxic. By that measure, transgenics clearly pass. [Sidenote: One of the papers Lotter cites looked at transgenic potatoes fed to rats. But the transgenic potatoes were much higher in protein than the controls, so were the effects due to the transgenic or the high protein? We can certainly produce high protein crops through conventional breeding.]

Longer term tests of course take much longer, are more expensive and are harder to control. And if we are concerned with the cross interactions, we need to test each new line/gene/trait in all the different backgrounds it will be put in. We practice monoculture, but it is monoculture on a local scale: the lines in Kansas are different from the lines in Nebraska etc etc.

So when you say that we need to do real safety testing, I wholeheartedly agree that we need to do some basic tests, but I get the impression (I could be wrong) that you are calling for massive, long term tests which just aren't feasible.   And give that I feel fairly comfortable with our knowledge of the main effects of the transgenics, I don't feel that they are necessary either.

But I absolutely agree that we should do basic safety testing  on new foods, I think we just disagree on what constitutes basic safety testing.  I also happen to think that we should be doing basic safety testing on vitamins and nutracuticals, which are basically unregulated as well.

---

by: ibaxter @ Tue Jun 09, 2009 at 22:05:08 PM CDT

[ Parent | ]