Observations that caloric restriction, i.e. a reduction in total energy intake from what would usually be consumed (called ad libitum intake), retards tumor growth in experimental animals have been published as early as 1909 by C. Moreschi . In recent years, calorie restriction has been more systematically investigated, usually as a 20-40% proportional reduction of energy intake from ad libitum feeding. It has been found that it not only retards tumor growth but could also sensitize tumor cells to pro-oxidative therapies such as ionizing radiation and chemotherapeutics. The big question is through which mechanism(s)? In contrast to restriction of specific nutrients, caloric restriction restricts several nutrients at the same time. Usually, carbohydrates constitute the major proportion of the usual diet, so that an absolute reduction of energy intake, when keeping this proportion constant, results in carbohydrates being restricted the most. Take a 2000kcal diet: If 50% energy are from carbs (250g), 20% from protein (100g) and 30% from fat (66.7g), a reduction to 1600 kcal (-20%) would be equivalent to a reduction of carbs by 50g, but protein by 20g and fat by only 13.3g (assuming 4kcal/g for carbs, 9kcal/g for fat, 4kcal/g for protein). As me and my colleagues have repeatedly pointed out, the main effects of calorie restriction are thus most probably due to the overall restriction of carbohydrates [2–4]. As the seminal work of Klein and Wolfe has shown , fat is the most neutral and carbohydrate the most disruptive macronutrient in humans with respect to their interference with the response to fasting, and hence calorie restriction in general. This is the rationale of using very low carbohydrate, high-fat ketogenic diets as fasting-mimicking diets for prevention or treatment of various diseases or general health promotion.
Now a new study nicely shows that it’s indeed the reduction of carbohydrates that is responsible for the effects of calorie restriction in a mouse model of non-Hodgkin lymphoma . This cancer type is frequently found to have a genetic defect in the gene encoding the “induced myeloid leukemia cell differentiation protein” Mcl-1, resulting in its overexpression. Mcl-1 belongs to the family of Bcl-2 proteins which are inhibitors of apoptosis, the cell death mechanism that should ideally be activated if a normal cell turns towards the malignant direction. Their counterparts are the so-called BH3 proteins (e.g. Bim, Puma,Noxa, Bad, etc.) which are pro-apoptotic. Bcl-2 and the BH3-proteins are able to bind to and inhibit each other. For this reason, BH3 mimetics have been developed with the aim of binding members of the Bcl-2 family in the hope of inducing apoptosis in cells overexpressing Bcl-2 proteins such as Mcl-1. The problem, however, is that so far BH3-mimetics have shown only low affinity for Mcl-1, so that aggressive forms of lymphoma with a high expression of Mcl-1 are hardly susceptible to them (a brand new study in Nature, however, just reports encouraging findings of a Mcl-1 specific BH3 mimetic ).
In 2013, the working group of Jean-Ehrland Ricci in Nice, France, has already shown that 25% calorie restriction in mice is able to reduce Mcl-1 expression and sensitize lymphoma cells to BH3-mimetics . In their new paper , the authors additionally introduced diets restricted by 25% in either carbohydrates or protein (by weight), but with the same energy content as the control diet. Their work shows that the carb-restricted diet, but not the protein-restricted diet, induced very similar effects as the calorie restricted diet they evaluated before. In particular, 25% carb restriction, but not protein restriction, led to a 50% reduction in Mcl-1 expression and sensitized experimental non-Hodkin lymphomas to cell death through treatment with a BH-3 mimetic. Noteworthy, carb restriction alone did not improve survival of the mice, but combined with the BH3-mimentic it doubled the maximal survival time from 40 to 79 days. This is another example of an experiment in which dietary manipulation alone was not sufficient to prolong survival, but supported the action of a targeted therapy.
What is striking is that the low-carb diet still yielded 54% energy from carbohydrates (compared to 70.9% in the control diet and 73.6% in the low-protein diet). In the real world, I would still consider such a diet as high carb. Thus, the non-Hodgkin lymphomas tested in these mice must have been extremely sensitive to the amount of carbohydrate in the diet. Nevertheless, this study provides further evidence that it’s the carbs (or the sugar once they are digested) that mostly interfere with cancer treatment, much more than protein and far way more than fat. Yes, there are examples in which protein restriction or restriction of certain amino acids resulted in tumor growth retardation, but I would predict that (besides a few exceptions) in these cases an alternative carb restriction would have performed even better. Another very recent study in rats  also indicated that even high amounts of the amino acid leucine would not promote tumor growth (despite leucine being a major stimulator of the cell growth-promoting mTOR pathway) but probably benefit the tumor host, e.g. by stimulating skeletal muscle anabolism. Exactly for the latter reason restricting protein is hardly an option for treating cancer patients. Therefore carb restriction – besides the fact that it’s the best strategy to accommodate for the altered metabolism of many tumor patients – should be the first consideration − period.
 Moreschi C. Beziehungen zwischen Ernahrung und Tumorwachstum. Zeitschrift Für Immunitätsforsch 1909;2:651–75.
 Klement RJ. Mimicking caloric restriction: what about macronutrient manipulation? A response to Meynet and Ricci. Trends Mol Med 2014;20:471–2.
 Klement RJ, Fink MK. Dietary and pharmacological modification of the insulin/IGF-1 system: exploiting the full repertoire against cancer. Oncogenesis 2016;5:e193.
 Fine EJ, Champ CE, Feinman RD, Márquez S, Klement RJ. An Evolutionary and Mechanistic Perspective on Dietary Carbohydrate Restriction in Cancer Prevention. J Evo Health 2016;1:15.
 Klein S, Wolfe RR. Carbohydrate restriction regulates the adaptive response to fasting. Am J Physiol 1992;262:E631–6.
 Rubio-Patiño C, Bossowski JP, Villa E, Mondragón L, Zunino B, Proics E, et al. Low carbohydrate diet prevents Mcl-1-mediated resistance to BH3-mimetics. Oncotarget 2016 [Epub ahead of print]
 Kotschy A, Szlavik Z, Murray J, Davidson J, Maragno AL, Le Toumelin-Braizat G, et al. The MCL1 inhibitor S63845 is tolerable and effective in diverse cancer models. Nature 2016;538:477–82.
 Meynet O, Zunino B, Happo L, Pradelli LA, Chiche J, Jacquin MA, et al. Caloric restriction modulates Mcl-1 expression and sensitizes lymphomas to BH3 mimetic in mice. Blood 2013;122:2402–11.
 Viana LR, Canevarolo R, Luiz ACP, Soares RF, Lubaczeuski C, de Mattos Zeri AC, et al. Leucine-rich diet alters the 1H-NMR based metabolomic profile without changing the Walker-256 tumour mass in rats. BMC Cancer 2016;16:764.