This is an interesting paper (from George I think). There are a stack of caveats about it, but the core findings appear to hold water.
They are dealing with cancer cell lines (ie cancer cells which are immortal and live in tissue culture) which have lost complex I of the electron transport chain. These cells have elevated levels of NADH per unit NAD+, ie their NAD+/NADH ratio is low. Obviously NADH is high because there is no complex I to oxidise it back down to NAD+. Any NAD+ which gets converted to NADH simply stays there as NADH.
Glycolysis continues to generate NADH and the TCA generates more NADH because that's all you can do with acetyl CoA (pax complex II derived FADH2), barring the scenario we looked at in a previous post.
You can manipulate the NAD+/NADH ratio.
It doesn't seem to matter whether you manipulate the absolute NADH levels down or the absolute NAD+ levels upwards and it doesn't seem to matter how you manipulate either level. Having excess NADH combined with relatively depleted NAD+ makes these cancers very much more aggressive in terms of metastasis. It's fascinating to see terms like Ki67 bandied around as aggression markers, we are just starting to use Ki67 clinically to try to assess to seriousness of that almost invariable "Grade 2" score assigned to the vast majority of mast cell tumours which get as far as histopathology. Are they a good Grade 2 or a bad Grade 2? Scoring cancer aggression is not the easiest thing to do and Ki67 looks interesting for those of us who have to manage mast cell tumours with a scalpel and/or a tyrosine kinase inhibitor. Anyway, back to the paper:
The paper brings up Ndi1, a fascinating little enzyme stolen from yeasts and engineered in to cancer cells. Ndi1 is a rather small, relatively simple enzyme which inserts itself in to the inner surface of the inner mitochondrial membrane and happily converts mitochondrial matrix NADH to NAD+ in a process linked to reducing the CoQ couple, but it pumps no protons in the process. It bypasses the broken complex I completely, skipping electrons straight from NADH to CoQ while effectively lowering NADH and raising NAD+. And it tames the cancer's metastatic behaviour.
Compare this Ndi1 route in to the ETC to the FADH2 route in from fat metabolism, which uses electron-transferring-flavoprotein dehydrogenase, and also reduces the CoQ without pumping any protons.
The two produce rather similar effects. Ndi1 is essentially making NADH behave like an FADH2 based input to the CoQ couple. Ndi1 may be more effective because it is purely focused on relieving the excess NADH and could, theoretically, completely normalise the NAD+/NADH ratio, whereas running metabolism on fat will only bias electron supply to FADH2 without stopping some NADH generation.
Both mice and cell cultures are being run on glucose in this paper. If all that matters is the NADH to NAD+ ratio, what might happen on a diet which intrinsically generates less NADH? An interesting but un-asked question.
Perhaps you might not need that Ndi1 enzyme inserted in to your inner mitochondrial membrane?
The very simple converse approach, similar to giving B3 as used by Hoffer, was to add an NAD+ precursor to the drinking water. I can see that this might well be effective in raising NAD+ but the thought of long term megadosing on any nutrient to achieve this effect is beyond what I might personally want to do. I might change my mind if I had cancer.
I do think that is very interesting in its own right, but it leads on to a host of other questions. In fact there are too many questions for a simple thread. It's particularly interesting to think about what an excess of NADH to NAD+ signifies to a cell. This strikes me as a core decision making signal about a cell's future. How this ties in with superoxide production from complex I, or the lack of it when complex I is dysfunctional, is also related. Couple that with the fact that high levels of NADH providing substrates (pyruvate, malate or glutamate) massively increase superoxide generation when FADH2 generating substrates (succinate) are simultaneously provided to isolated mitochondria from normal tissues. The whole area looks to have lots of potential for working out what is going on at the primary switching point of the electron transport chain.
That looks OK but life is never quite that simple....
It's very interesting to note that there are now several publications from this group which are pretty convincing that many breast cancers, in vivo, have hyperactive mitochondria and perform oxphos to a very high level to support their aggressive growth pattern. Including both increased expression of the genes for complex I components and the enhanced ability (in mitochondria from fresh frozen human surgical breast cancer biopsies) to process large amounts of NADH to NAD+. Certainly compared to both non neoplastic surrounding epithelial cells and most especially compared to the surrounding stromal cells (mostly fibroblasts).
The results from the two groups are both very convincing and utterly incompatible.
To square the circle you get some help from this paper which suggest that at least one of the cell lines used in Santidrian's study are, in fact, melanoma derived cells, not breast cancer cells at all. In terms of aggressive phenotype most clinicians might be significantly more concerned about melanoma vs breast adenocarcinoma. I would anyway.
You may have to be married to a pathologist to realise quite how difficult it is to differentiate highly malignant cell types as both their genes and their appearance seem to eventually degenerate or evolve in to some sort of identical "cancer cell".
That seems to be the point being made by this chap.
So whether MDA-MB-435 cells are melanoma cells or breast cancer cells, they are degenerate/evolved enough to be indistinguishable from either family of cancer of origin and may well be behaviourally indistinguishable too.
The logical explanation is that MDA-MB-435 cells represent a more advanced cancer development than the still rather "normal" cells in routine surgical patients, most of the cells from which will be running on oxphos fueled by manipulating their surround fibroblasts (which will be running on glycolysis, with shut down mitochondria) to donate large amounts of TCA substrates to their controlling cancer cells
Aside: You still have a massive fuel source for cancer cells based on glycolysis here, it's just moved from the cancer cells themselves to the surrounding fibroblasts. But the down side is that while you might starve the fibroblasts by running on low glucose and high FFAs (they have lost their mitochondria) you still have active cancer cells willing to use ketones, FFAs or lactate through their very active mitochondrial electron transport chains... End depressing aside.
I really like this idea of degrees of degeneration as a possible explanation for the irreconcilable results. Again, you might need to be married to a pathologist to appreciate the phenomenal importance of fibroblasts in coming to a histopathological diagnosis about many cancer families. Fibroblasts are VERY important. Metabolic coupling may well be why.
Not sure whether to go on to mitochondria in Parkinsons or have a break and look at this whole concept of metabolic coupling between cancer cells and fibroblasts. The two subjects are vaguely convergent, eventually.