Faking it with selenium

 I've looked at this paper in the past, here and here and it is, to say the least, a little dubious in places

High selenium impairs hepatic insulin sensitivity through opposite regulation of ROS

and I really, really want it to be genuine because it supports this hypothesis

High selenium -> high ROS scavenging -> impaired insulin signalling   -> adipose lipolysis

A follow on from this, also plausible, is that

High adipose lipolysis -> excess FFAs to liver -> hepatic ROS from beta oxidation -> hepatic insulin resistance

which would be normal under high FFA oxidation in hepatocytes. I'm willing to accept that reducing ROS in adipocytes might increase ROS in hepatocytes, at a push. But...

These are images taken of hepatocytes from recently euthanased rats fed supplementary selenium for six weeks. This is section A from Figure 3 as a simple copy paste:

What we are comparing is the colour intensity of MitoSOX, a marker of ROS generation,  between control square and the HSe, high selenium square. These two:

Control is essentially black, ie no ROS production and HSe is red, lots of ROS production. Let's cut out the intermediate LSe row and abutt the HSe to the control to make this as obvious as possible:

Now I have another problem. The bright blue colour is an Hoechst stain developed to show DNA, ie it shows the nuclei. I would expect the same intensity of staining with Hoechst stain irrespective of exposure to selenium for 6 weeks or not. The control Hoechst stain is dull in the control while in the HSe image it is bright. Perhaps there has been a problem with the photography?

I can help out by simply increasing the exposure factor for whole of the control row using the colour adjustment in Preview software. This is what it looks like:

Which makes for an interesting comparison between the MitoSOX fields at this (???corrected???) exposure/brightness:

I defy anyone to see any difference between the control MitoSOX intensity and the high selenium exposure MitoSOX intensity. Which makes the rest of the paper, attempting to explain this non-finding, of little value.

So what do I make of this table:

showing a clear dose response to sodium selenite supplementation in both bodyweight and fat weight? For someone who thinks that markedly reducing ROS in adipocytes should cause lipolysis this is exactly what you would expect.

Bear in mind that these rats were gavaged with selenium, it wasn't in the food, there is no "palatability" issue.

I think this is real. Why? Because I want it to be and because of this paper

Acute and Subchronic Toxicity Studies on Sel-Plex, a Standardized, Registered High-selenium Yeast

which found exactly the same effect. This is food intake per week:

The results for weight gain are exactly as expected from the food intake and are given in Table 4, which is huge and turned on its side over two pages. You need a VPN and a line to a server facilitated from Kazakhstan or just take my word for it, food intake predicts final weight, this time.

This is explained away by the authors using ad hoc hypothesis number 12,352 thus:

"A reduced dietary intake was noted throughout the observation period for all treatment groups, which may be considered as a consequence of the unpalatability of the dietary mixture. This resulted in a decreased bodyweight gain in all treatment groups, particularly the females."

I guess, if you were dumb enough, you could call this "reverse Reward". No sniggering!

Combining parts of both papers leaves me deeply embedded in my own confirmation biases, which make sense, at least to me.

ROS are fundamental.

Peter