Alzheimer's disease (AD) and type-2 diabetes (T2D) are two fatal human diseases and have been linked to the aberrant aggregation of two distinct peptides, amyloid-beta (A beta) and human islet amyloid polypeptide (hIAPP), respectively. These two peptide aggregates, even with distal deposition sites (brain and pancreas), act as mutual beneficiaries. We here unveiled the crosstalk in a self-consistent fashion using atomistic simulations by comparing the kinetics and thermodynamics of self- and cross-aggregations of A beta(42) and hIAPP and their modulations by preformed fibrillar templates. Templates (specifically hIAPP) generally accelerate aggregation, alter the relative order of aggregation rates (cross-aggregation > A beta self-aggregation > hIAPP self-aggregation for nontemplated and hIAPP self-aggregation > cross-aggregation > A beta self-aggregation for templated), and flip the mutual impact (hIAPP aggravates A beta aggregation in nontemplated and the reverse in templated). Higher instances of breaking larger aggregates and longer residence times of smaller aggregates decelerate aggregation, whereas interpeptide electrostatics (universal) and hydrogen bonds (templated) assist it. However, the equilibrium aggregability pattern contradicts kinetic rank-ordering, as A beta displays a higher aggregability than hIAPP, templates increase aggregability for both peptides, and A beta's self-aggregability supersedes cross-aggregability, which further surpasses hIAPP's self-aggregability. The equilibrium ensembles encompass polymorphic, nonfibrillar oligomers having substantially reduced alpha-helicity and slight beta-propensity, with both parallel and antiparallel interpeptide orientations, primarily stabilized by electrostatics. A higher equilibrium aggregability means a greater helix-breaking capacity, a bias toward parallel orientation, and a lesser structural polymorphism. Water expulsion from peptide surroundings and distortion of water tetrahedrality prove that aggregation follows the liquid-liquid phase separation (LLPS) model.

Foreign