Humanin research: what the published studies show

Discovery and Selectivity

A 2001 study identified humanin by functional expression screening of a complementary DNA library, and the transfected complementary DNA was transcribed to the polypeptide and secreted into the medium.^[1]

The polypeptide abolished death of neuronal cells caused by several familial Alzheimer's disease genes, and death caused by amyloid beta.^[1] It had no effect on death caused by Q79 or by superoxide dismutase-1 mutants.^[1] That null is why the result reads as selective rather than general.^[1]The rescue action depended on primary structure.^[1] This is in vitro work using native humanin.^[1]

The BAX Interaction, Two Accounts

A 2003 study reported that native humanin, a 24-residue peptide, interacts with BAX and prevents it translocating from the cytosol to mitochondrial membranes.^[2] Humanin peptides also blocked BAX association with isolated mitochondria and suppressed cytochrome c release in vitro.^[2]

The same paper noted that the mitochondrial genome contains an identical open reading frame, and that the mitochondrial version of the peptide also binds and suppresses BAX.^[2] The authors speculate that humanin arose from mitochondria and later transferred to the nuclear genome, and that word is theirs.^[2]

A 2019 study from an unconnected group used recombinant protein with light scattering, circular dichroism and fluorescence spectroscopy, and reports that humanin and BAX form fibers together in vitro.^[5] BAX undergoes structural rearrangement and incorporates into the fiber.^[5] Humanin mutations known to alter anti-apoptotic activity also altered the morphology of those fibers.^[5]

These two papers describe the same inhibition by two different routes, one blocking translocation and one trapping BAX in an aggregate.^[2] Reference 5 calls its account "a potential mechanism" and the first report of it, and this page reports both without choosing between them.^[5]

Receptor Identification

A 2009 study found gp130, the common subunit of the interleukin-6 receptor family, essential for humanin-induced neuroprotection.^[3] Overexpressing CNTF receptor alpha or the interleukin-27 receptor subunit WSX-1 increased humanin binding to neuronal cells, and overexpressing any other tested gp130-related subunit did not.^[3]

Knocking down endogenous CNTF receptor alpha or WSX-1 by small interfering RNA reduced binding, and loss of function in either nullified responsiveness to humanin-mediated protection.^[3] In reconstituted binding assays humanin, but not a control peptide, induced hetero-oligomerisation of the three subunits.^[3]All of this is in vitro.^[3] It used neuronal cells, reconstituted systems and native humanin throughout.^[3]

Insulin Action in Rats

A 2009 study infused native humanin into rat cerebral ventricles and measured insulin action by clamp.^[4] Overall insulin sensitivity improved, the effect was associated with activation of hypothalamic STAT-3 signaling, and co-inhibiting that signaling negated it.^[4]

Intravenous infusions of potent humanin derivatives, which the abstract does not name, reproduced the insulin-sensitising effect, and inhibiting hypothalamic STAT-3 canceled that analogue's effect on liver.^[4] There was no direct effect of humanin on primary hepatocytes.^[4] That null is what makes the hepatic action centrally mediated.^[4]

A single treatment with a highly potent analogue lowered blood glucose in Zucker diabetic fatty rats.^[4] The analogue is not identified in the abstract, so that result belongs to an unnamed molecule rather than to the compound sold here.^[4]

Detectable humanin fell with age in rodent hypothalamus, skeletal muscle and cortex, and circulating levels fell with age in mice and in humans.^[4]

State of the Evidence

Four of the five sources are in vitro and reference 4 is the only in vivo work, in rats.^[1] Nothing here is a clinical trial.^[3]

References 1, 2, 3 and the central arm of reference 4 used native humanin, the molecule sold here.^[1] The peripheral and glucose-lowering results in reference 4 used analogues the abstract does not name.^[4] That names the wrong molecule.^[5]

Independence is unusually good.^[5] The discovery and receptor papers share a first author, but the BAX work comes from two unconnected groups and the metabolic work from a third.^[2]