Here is a little more from this paper:
High selenium impairs hepatic insulin sensitivity through opposite regulation of ROS
This is insulin signalling under massively supra-physiological insulin exposure in cell culture:
This is, obviously, their best gel, that's the one you publish. The insulin resistance (fainter P-Akt band) when insulin and Se are both used compared to insulin w/o Se exposure does appear to be there. At physiological levels of insulin this differential seems likely to be maintained.
This implies blunting of insulin signalling, which allows more FFA oxidation, which generates greater levels of ROS than would occur under continued insulin action. These ROS would be physiological on a ketogenic diet or under extended fasting but are not so in cells under culture using 11mmol glucose (which is what I think is in the medium they used, they don't actually say) plus whatever insulin is present in 10% FBS. So we have this:
Control is from cells under RPMI 1640 alone, traditionally 11mmol/l glucose. Excess selenium blunts insulin signalling so allows FFA release from intracellular triglyceride stores, so increases ROS (in the same way as metformin does but w/o the suppression of gluconeogenesis intrinsic to metformin's action). Adding rotenone, as you would expect, blocks RET so blocks ROS generation. CCCP uncouples respiration, drops delta psi so blocks RET/ROS. Etomoxir blocks access of FFAs to mitochondria so blocks input at mtETFdh, so blocks RET/ROS. MitoQ powerfully targets all mitochondrial ROS so over-rides the FFA oxidation ROS generation effect. Chromium picolinate restores insulin signalling by repleting the Cr depletion induced by Se. MSA is an inhibitor of glutathione peroxidase, so it eliminates the effects of excess GPX. And SS, sodium salicylate, appears to block intracellular lipolysis in hepatocytes, so suppresses fatty acid supply to mitochondria, much as insulin or etomoxir would.
All a very plausible narrative.
Except for oligomycin. What does anyone expect the blockade of ATP synthase to do to ROS generation, throughout the electron transport chain? It is going to increase delta psi, reduce all of the redox complexes and generate a ton of RET and ROS through complex I and probably at ton at complex III too. It is specifically used to generate ROS in many other studies, example here:
The specificity of neuroprotection by antioxidants
I'm not very comfortable with oligomycin as a suppressor of FFA oxidation induced ROS. It is another, rather serious, blight on the paper. It certainly should have been discussed.
I would usually ignore the whole paper except Tom Naughton gave us all the heads up on a recent report of a chap taking what might have been a hefty dose of green tea extract who went in to liver failure. Obviously most folks just excrete antioxidants like GTE with little harm done. I just wonder if he got unlucky or took a huge dose while walking round with the sort of liver full of lipid so beloved of Public Health England. Losing the protection of insulin's inhibition of lipolysis simply dumped a ton of unregulated intra-hepatocyte FFAs from lipid droplets on to his mitochondria, which then popped their clogs.
Who knows? It's another nice narrative. I just wish I wasn't so suspicious of the selenium paper...
Both reports also play rather too well to my biases against antioxidants, but that's how it is...