I can tell that no plant geneticists reviewed this paper, certainly no maize geneticists.
The authors claim to use a 'near-isogenic' line for a control.
> The varieties of maize used in this study were DKC 2678 Roundup-tolerant NK603 (Monsanto Corp., USA), and its nearest isogenic non-transgenic control DKC 2675
Yet no mention of how 'near' these lines are is made, nor is the seed provenance discussed. This is, frankly, absurd, as the Monsanto transgenic line began development decades ago. Presuming that the other inbred had been established at that point, there is a minimum of 20 years that these inbreds have been independently maintained.
A proper experiment would have introgressed the EPSPS transgene into a known inbred. That is the only way you can ensure that you have a consistent genetic background, and is standard practice in the maize community. This is, almost without exception, reported in a manuscript. It is that important.
To contextualize this, I've compared control groups of a particular inbred (e.g. B73, Mo17, A619) and seen substantial differences in RNAseq data, simply because it was seed stock from different labs. These are lines that have been shared within the past 10 years. The B73 inbred (from which the maize genome was/is sequenced) is notorious for being defined as 'a barrel of seed in Mike Freeling's lab'. Honesty, the sheer diversity of maize, along with wind pollination, complicated field conditions, and undergraduate help =P, make all maize inbreds a moving target.
This is not a small detail of the paper; all the results rely on these lines actually being 'near-isogenic'. Any small differences between them can easily account for the observed differences. Hell, even after 3 or 4 rounds of introgression, you can often visually see differences between lines that aren't actually due to the trait in question; it's just background noise.
Whatever your views on transgenic technology, I would not base any decision on this work.
Note for comments-first readers: this shows differences in corn metabolism, not in humans or other animals that have consumed the corn. However they reference studies where rats had worse health after eating GMO corn vs normal corn.
In a sense this probably isn't terribly surprising; GM corn is known to have reduced yield due to disruption from the genetic modifications. The interesting part is that the composition of the resulting corn differs as a result of this in ways that weren't expected or previously known.
The research compared an original strain to a modified strain, to the modified strain with roundup. It is unsurprising that the modified strain, that is producing an extra protein in high amounts will be under a heavier metabolic load than the same strain that is not producing that extra protein. It seems the metabolic load showed up in a few particular places that haven't obviously been investigated before, and now we can more carefully probe those differences now that the team did a thorough screen to find them.
However, what the research did not do is compare the metabolic load in the modified strain to anything else. They did not compare it to an average corn strain, an 'heirloom' corn strain, a corn that is under cold/heat/viral or any other kind of environmental stress or bounty. Nor did they compare the corn metabolite levels to other foods. I suspect the difference in metabolic load between a strain +/- a particular resistance protein is likely less than the difference in metabolic load between various corn strains, much less the difference between fertilized/shaded/cold-conditioned corn, much less the original grass strain from which human-edible corn came from, much less from another food crop. The difference appears in the ratio of particular metabolites in the corn. And at the end of the day the difference in amounts of small molecules upon adding this metabolic stress is likely completely overwhelmed by the literal amount of material (ex. if you don't want that much of the metabolite, use 0.01% less corn. And if the corn is processed at all, this kind of difference becomes moot very quickly.)
The result is not surprising and is framed in a particularly click/news-bait worthy way. The nice part of the research is that it did a lot of hard work to find the variables that now can be looked for in those above cross-wise comparisons without having to do the major hard work of screening an entire transcriptome for each comparison. But it does not in any way provide evidence for any danger to human health at all.
Some math:
The modified strain's largest difference in metabolites was a 28x increase in cadaverine (btw, it's pretty entertaining how many nitrogen-containing compounds have fantastic names because of their odor). According to [1] the difference in the amount of cadaverine between dried and canned corn is 5.5x (with the low cadaverine in dried corn mentioned as nutritionally deficient). And the total amount of cadavarine in fresh corn is ~ 10mg/100g which is ~ 0.01% by mass. Most of the other large changes in metabolites were changes of less than an order of magnitude.
Yes, the proper comparison would be an additional transgenic strain expressing a suitably 'inert' protein to thus differentiate between the metabolic demands of high levels of protein expression vs. some particular effect of the heterologous EPSPS. Ideally, you'd use the same protein with a deletion of an active domain. It is common use multiple deletion lines to identify the causative domain; this sort of mechanistic work is essential for understanding what is actually going on.
Moreover, the use of 'near-isogenic' lines is very troubling, as they are rarely that similar. You can do e.g. RNAseq on 'identical' inbreds from different labs and get substantial differences. Seed is passed around so much, mis-labled in the field or out, etc. They certainly do not have an identical inbreed from something that was developed decades ago, not even close. Calling them near-isogenic is, well, let's say 'not that honest'. Very important to note that they have no way to control for this difference in genetic background.
Also noteworthy; Michael Antoniou has co-authored papers with e.g. Seralini, notorious for a widely-discredited paper on RoundUp https://en.wikipedia.org/wiki/S%C3%A9ralini_affair That certainly does not discredit this work, but it is concerning when an investigator agrees to co-author a manuscript of such poor quality.
Finally, having done a lot of work on '-omics' in plants, I've yet to see an experiment that doesn't turn up some differences. Their list of GO terms is not remarkable, and changes of '30 fold' are not nearly as dramatic as they appear. This is well within the range of what I'd expect to see between different similar inbreds, which these lines certainly are.
Let me be more specific: the question was "how will the GM industry will react"
‘substantially equivalent’ is a proxy for food safety and it's also a tricky standard. When interpreting this study's results the public and policy makers will look to academics:
So do you have evidence it is unsafe or just are you simply against any organization that uses lobbyists? (Teachers unions, ACLU, Climate and Gay rights organizations, etc)
I didn't say anything about lobbyists. I do have an issue with supposedly unbiased experts being on a private payroll without full disclosure all-around. Would you be okay with a NY Times reporter getting funding for a story from Monsanto (without disclosure)?
If I'm understanding you correctly, I think you might be blurring the lines here. A NYT reporter is a journalist, not an expert. They may use experts as sources for their stories. I agree that it would be a serious breach of ethics for a journalist to take money from Monsanto. It's not the same as an expert receiving funding from industry. That has it's own ethics issues, which I think are the ones you're getting at.
The authors claim to use a 'near-isogenic' line for a control.
> The varieties of maize used in this study were DKC 2678 Roundup-tolerant NK603 (Monsanto Corp., USA), and its nearest isogenic non-transgenic control DKC 2675
Yet no mention of how 'near' these lines are is made, nor is the seed provenance discussed. This is, frankly, absurd, as the Monsanto transgenic line began development decades ago. Presuming that the other inbred had been established at that point, there is a minimum of 20 years that these inbreds have been independently maintained.
A proper experiment would have introgressed the EPSPS transgene into a known inbred. That is the only way you can ensure that you have a consistent genetic background, and is standard practice in the maize community. This is, almost without exception, reported in a manuscript. It is that important.
To contextualize this, I've compared control groups of a particular inbred (e.g. B73, Mo17, A619) and seen substantial differences in RNAseq data, simply because it was seed stock from different labs. These are lines that have been shared within the past 10 years. The B73 inbred (from which the maize genome was/is sequenced) is notorious for being defined as 'a barrel of seed in Mike Freeling's lab'. Honesty, the sheer diversity of maize, along with wind pollination, complicated field conditions, and undergraduate help =P, make all maize inbreds a moving target.
This is not a small detail of the paper; all the results rely on these lines actually being 'near-isogenic'. Any small differences between them can easily account for the observed differences. Hell, even after 3 or 4 rounds of introgression, you can often visually see differences between lines that aren't actually due to the trait in question; it's just background noise.
Whatever your views on transgenic technology, I would not base any decision on this work.