Tucker Goodrich emailed me the link to this lovely paper from the 1930s
and included it in a blog postFat and Weight Gain (a Note to Peter) and the Essentiality of Linoleic Acid
TLDR, grossly linoleic acid (or the arachidonic acid derived from it) deficient mice are small, emaciated and lose a ton of water through their skin. Think superoxide.
It's an interesting situation, with a number of possible explanations and the sort of data you might expect from 1930, good for its time but we're not going to get the massive detail which people struggle to interpret nowadays.
I started off alongside the authors and caloric loss due to the evaporation of water from both the "leaky" skin and possibly from "leaky" lungs. If you have ever tried to keep a mouse warm under anaesthesia while it breathes anhydrous cold gas and is clipped and water/alcohol prepared for surgery you will sympathise. If not, you can take my word for it that without serious attention to heat conservation, their temperature drops like a stone.
The mice lose about 10ml of water per day through their skin. With a latent heat of vaporisation of 0.541kcal/g this means the 10ml of water steals about 5.41kcal per day of energy as lost heat, just to evaporate the leaked water. That's about an extra gram of carbohydrate or protein they would need.
But the mice were ad lib fed. They may have been pretty sick from LA deficiency but they managed to eat exactly as much as the healthy mice did each day. I find it implausible that they were so skinny because they were simply too ill to manage an extra gram of food. There is absolutely no doubt that they were very sick, but I still don't find this plausible for the emaciation.
A simple extrapolation from high PUFA diets facilitating fat loss in to adipocytes to the converse, that very low PUFA diets fail to allow fat storage is nice but seems very unlikely because the curative dose of linoleic acid seems too small to represent a supply of calories for bulk oxidation.
My core question relates to the failure to store lipid in adipocytes, with a spin off question in to the poor health of the fat free rats. This is all guesswork and speculation.
I have to accept the possibility that it might be as simply as that you cannot build a functional cell membrane or mitochondrial membrane without some bendy/fluid molecules. Given an inadequate fluidity to the mitochondrial inner membrane I can see that molecules such as CoQH2 or reduced cytochrome C, which move within/over a fluid inner mitochondrial membrane might lose electrons prematurely to molecular oxygen if they cannot reach their docking site in the next complex down the ETC. That a continuous, pathological loss of electrons from carriers to molecular oxygen, to give inappropriate superoxide and H2O2 generation, would lead to unrestrained insulin resistance and failure to grow, especially failure of adipocytes to grow, seems to be a plausible explanation.
As Burr and Burr comment, the kidney is also a very energy dependent and might well fail given a failure of insulin signalling due to excessive H2O2 generation. While the rats would lack ROS damage to membrane lipids they would not lack ROS damage to cellular proteins or DNA. The cells, if insulin signalling cannot occur due to excess H2O2 generation, would also be calorie starved as well as ROS damaged.
Quite why the tails fell of of the rats is beyond even my speculation ability.
So. The failure of adipocytes to grow suggests severe insulin resistance within those adipocytes. More generalised application of this idea would lead to explaining failure of the whole rat to grow.
Who knows what actually happened?
I've long carried the idea in my head that PUFA depletion might be protective against radiation induced injury. I think I picked this up early in my low carb journey, probably from a Ray Peat follower. But that will have been over 15 years ago and I am unable to locate where this idea came from. So I went looking on Pubmed and opened a whole can of worms. Some of which are interesting.
The first is that rats fed a completely fat free diet are at increased risk of dying from radiation injury. That was this one:
THE EFFECT OF FAT LEVEL OF THE DIET ON GENERAL NUTRITION
XIII. THE EFFECT OF INCREASING DOSAGES OF X-IRRADIATION
ON THE PROTECTIVE ACTION OF FAT ON RADIATION INJURY
"More recently, it has been observed that the survival time of male rats, as judged by the intervals at which an LD25, an LD50, or an LD75 were reached, or by the average length of survival, was progressively improved when ethyl linoleate was given in doses of 10, 50, or 100mg daily (Cheng et al., '54)."
The same lab has produced at least four studies to support this finding.
Giving a rat 10mg of a source of linoleic acid will have no effect on substrate oxidation. But it might well allow the development of an effective electron transport chain which is less likely to release a random excess of electrons to molecular oxygen.
Of more interest is this study. It's not clear (to me) whether the diets contained 10, 20 or 30% of cotton seed oil by calories or by weight. Low PUFA diets (butter) were ineffective in small amounts but effective at 20% or 30%:
Deleterious effects of high fat diets on survival time of X-irradiated mice
"At levels of 2% or 10% of the diet cottonseed oil and margarine fat increased survival time over that on the fat-free ration. When these fats were fed at higher levels ( ie. ~ 20% or 30% of the diet), however, survival time was decreased below that obtained at the lower levels of supplementation."
My take home is that rats do, absolutely, need a few milligrams of linoleic acid. Notice that the amounts used in all of the studies are in the same ball park as found by Burr and Burr to prevent their fat deficiency syndrome. I think it is an interesting speculation that LA is particularly need to manufacture an effective membrane for the ETC, especially when electrons are going to be knocked around by x-ray irradiation over and above background radiation conditions. As the dietary dose increases then the deleterious effect of the PUFA eventually predominate, certainly in the radiation injury models.