From fuel to signal

In a 2014 review Newman and Verdin argued that ketone bodies, BHB above all, are not only fuels but signalling metabolites: they bind receptors, inhibit enzymes and serve as raw material for protein modifications, and so affect gene expression and inflammation (Newman and Verdin 2014). This chapter presents three influential findings and their later revisions, and then uses the Hill equation to connect concentration with effect.

Histone deacetylases

The DNA in the nucleus is wound around histone proteins. When lysines on histone tails are acetylated, chromatin is usually looser and the genes concerned are more easily expressed; histone deacetylases (HDACs) remove those acetyl groups.

The inflammasome

This finding has been widely cited, and often simplified to “a ketogenic diet is anti-inflammatory”. Note that the evidence comes mainly from cells and mouse models; in humans the effects of fasting or a ketogenic diet on inflammatory markers are inconsistent, and “inhibiting NLRP3” is not always a good thing, since inflammation is also needed to fight infection.

The receptor

BHB is also an endogenous ligand of the cell-surface receptor HCAR2 (also known as GPR109A, the niacin receptor). Taggart and colleagues reported a half-maximal effective concentration of about 0.7 mmol/L for activation by BHB, within the range blood reaches during fasting (Taggart et al. 2005). In fat cells, activation of this receptor inhibits lipolysis, forming a negative feedback: as ketones rise, the release of fatty acids is restrained, limiting further ketone production.

Concentration decides the effect

The half-maximal concentrations reported in these studies differ widely, while blood BHB can range within a day from 0.1 to several mmol/L. Only by putting the two together can we judge when a given mechanism might operate.

# Concentration-response with the Hill equation: fraction of effect = C^n / (EC50^n + C^n).
# Half-maximal concentrations reported in the literature (single studies, in vitro):
#   HCAR2 (GPR109A) receptor activation by BHB: EC50 about 0.7 mM (Taggart et al. 2005)
#   inhibition of immunoprecipitated HDACs by BHB in vitro (Shimazu et al. 2013):
#       HDAC1 IC50 5.3 mM, HDAC3 2.4 mM, HDAC4 (class IIa) 4.5 mM
# Hill coefficient n = 1 is assumed throughout (no cooperativity information is used).
TARGETS = {"HCAR2 activation": 0.7, "HDAC3 inhibition": 2.4, "HDAC4 inhibition": 4.5, "HDAC1 inhibition": 5.3}
LEVELS = [("after an overnight fast", 0.1), ("around 24 h of fasting", 0.4), ("nutritional ketosis, typical", 1.0),
          ("nutritional ketosis, high", 3.0), ("prolonged fasting, days", 5.0)]

def hill(c, ec50, n=1.0):
    return c ** n / (ec50 ** n + c ** n)

print("blood BHB (mM)                    | " + " | ".join(f"{t:16s}" for t in TARGETS))
for label, c in LEVELS:
    print(f"{label:28s} {c:4.1f} | " + " | ".join(f"{hill(c, e):16.0%}" for e in TARGETS.values()))
# Two cautions the numbers make concrete:
# 1. in vitro enzyme assays need not reflect inhibition inside cells and tissues;
# 2. a steep Hill coefficient would change the picture: with n = 2, HDAC3 inhibition at 1 mM is
print(f"   n = 1: {hill(1.0, 2.4):.0%};  n = 2: {hill(1.0, 2.4, 2):.0%}  (same IC50, 1 mM BHB)")
# blood BHB (mM)                    | HCAR2 activation | HDAC3 inhibition | HDAC4 inhibition | HDAC1 inhibition
# after an overnight fast       0.1 |              13% |               4% |               2% |               2%
# around 24 h of fasting        0.4 |              36% |              14% |               8% |               7%
# nutritional ketosis, typical  1.0 |              59% |              29% |              18% |              16%
# nutritional ketosis, high     3.0 |              81% |              56% |              40% |              36%
# prolonged fasting, days       5.0 |              88% |              68% |              53% |              49%
#    n = 1: 29%;  n = 2: 15%  (same IC50, 1 mM BHB)

The table puts several popular claims back into numbers. After an overnight fast BHB is only about 0.1 mmol/L: the HDACs are hardly affected and the receptor is only about one-tenth activated. In typical nutritional ketosis (about 1 mmol/L) the HCAR2 receptor is about 60 % activated, while HDAC inhibition is only 16–29 %, and that is calculated from in vitro enzyme data; if inhibition inside cells is much weaker than in vitro, as Chriett and colleagues found, the real inhibition will be smaller still. Only in prolonged fasting, with BHB around 5 mmol/L, does HDAC inhibition reach one-half to seven-tenths. The last line of the program shows too that with the same half-maximal concentration, a Hill coefficient of 2 instead of 1 cuts the inhibition at 1 mmol/L from 29 % to 15 %: our quantitative grasp of these mechanisms is still rough.

Summary

BHB does act beyond being a fuel: it is a receptor ligand, it inhibits histone deacetylases in the test tube, it inhibits the NLRP3 inflammasome in cells and mice, and it is the source of a new histone modification. But each action has its own concentration range; on present data, at the blood ketone levels most people reach, receptor actions are more likely to matter than HDAC inhibition, and the cellular evidence for HDAC inhibition itself has been questioned. The next chapter turns to another much-discussed topic: fasting and autophagy.