Transformation of singly dispersed protein molecules into amyloid fibrils is a complex multistage process involving coupled phase- and conformational transitions. Formation of such fibrils is, in principle, accessible to various proteins, as well as synthetic peptides [[1], [2], [3], [4]] and is often thermodynamically favored under the physiological conditions (vis à vis the native state). This, in turn, underlies the biomedical importance of amyloidogenesis of various proteins implicated in the etiology of several degenerative maladies including Alzheimer's disease, Parkinson's disease, and type II diabetes mellitus [[5], [6], [7]]. Cellular damage occurring in the course of these diseases may be caused by certain on-pathway intermediates rather than mature fibrils which themselves could be non-toxic. In fact, diverse organisms from bacteria to mammals take advantage of the stability of amyloid fibrils ([[8], [9], [10], [11]]) to accomplish various tasks, hence the term ‘functional amyloid’ was coined [[12], [13], [14], [15]]. While similar thermodynamic forces drive folding and misfolding (and the self-assembly of amyloid fibrils), the energy landscapes envisaged for these two phenomena are distinct with multiple minima (corresponding to amyloid structural variants, or polymorphs (strains)) separated by significant barriers expected for the latter [16,17]. This, coupled to the auto-catalytic character of the fibril's elongation process, subjects proteins undergoing amyloidogenesis to the kinetic control [18]. In a canonical elongation scenario, a pre-existing structural amyloid variant (seed) propagates (imprints) its structural phenotype in the recruited protein monomers. This holds true even under physicochemical conditions favoring formation of an alternative variant. Such a conformational memory effect, a specific to amyloidogenic proteins type of heteroepitaxial growth, is a typical manifestation of self-propagating polymorphism of amyloid fibrils – nowadays a well-recognized phenomenon [6,9,[19], [20], [21], [22], [23], [24]]. Structural differences between amyloid strains obtained from a single protein may vary from tiny [6,25] to extreme [26]. In either case, the resulting properties of fibril may be essential in defining the biological impact (as is the case of prions [27]). At the stage of de novo nucleation of amyloid fibril, selection of a particular structural variant is often influenced by physicochemical conditions – e.g. temperature [28], presence of particular salts [29], or co-solvents [[30], [31], [32]]. The latter follows from the fact that protein aggregation is accompanied by significant changes in hydration [[32], [33], [34], [35], [36], [37]]. We have shown earlier [33] that insulin aggregation is accompanied by negative heat capacity changes pointing to substantial reduction in the number of water–protein contacts upon fibrillization. From this perspective folding and aggregation could be conceptualized as two competing transitions minimizing the amount of frustrated water in protein's hydrational layers. Presence of organic co-solvents modulating protein-water interactions would therefore affect the overall energetics of the aggregation process. Preference for de novo formation of a particular amyloid strain may also be caused by small changes in the primary structure. We have shown earlier that fibrils formed upon aggregation of recombinant LysB31-ArgB32 human insulin analog ([KR]) pass their characteristic infrared features to daughter fibrils upon cross-seeding to bovine insulin (and vice versa) [38]. In this study, we are showing that acetone-induced perturbation of aggregating bovine insulin leads to the same IR features as in the [KR] amyloid phenotype. A comparative analysis of the stability of bovine insulin fibrils formed in the presence [BI-ace] and absence [BI] of acetone reveals that the [KR]-like infrared characteristics depends on the hydration state of mature fibrils.