Humanin is a tiny protein that can protect brain cells, but its shape changes a lot depending on the surrounding liquid. In water it clumps together, while in a more oily environment it stays single and forms a helix. When it meets a model of brain cell membranes, it sits on the surface, takes on a sheet‑like shape, and makes the membrane a bit more fluid without actually going inside.

Humanin (HN) is a recently identified neuroprotective peptide able to inhibit neurotoxicity induced by various insults which can be related to Alzheimer disease (AD) as well as to cell death induced by other stimuli. Previous CD and NMR studies demonstrated that HN adopts an unordered conformation in water, a alpha-helix conformation in 30% TFE, and a beta-sheet structure in PBS. Furthermore, other studies clearly indicated HN as a secreted peptide, able to prevent neuronal cell death caused by amyloid beta (Abeta) derivatives. Although Abeta was found to interact with neuronal membranes, currently there is not experimental evidence unveiling HN interaction with membranes. In this paper a spin labeling technique coupled with electron paramagnetic resonance (EPR) and circular dichroism (CD) has been used to study the structure and dynamics of HN in solution and for the first time in the presence of model cerebral cortex membranes (CCM). We have demonstrated that HN has a great tendency to aggregate even at low concentrations in water solutions at different ionic strengths and monomerizes in the TFE apolar environment. We also showed that HN slightly perturbs model CCM at the surface assuming a clear beta-sheet conformation. In addition, HN increases the fluidity of the bilayer core without penetrating into the membrane.