We study equilibrium statistical mechanics of Fermion lattice systems which require a different treatment compared with spin lattice systems due to the non-commutativity of local algebras for disjoint regions. Our major result is the equivalence of the KMS condition and the variational principle with a minimal assumption for the dynamics and without any explicit assumption on the potential. Its proof applies to spin lattice systems as well, yielding a vast improvement over known results. All formulations are in terms of a C*-dynamical systems for the Fermion (CAR) algebra [Formula: see text] with all or a part of the following assumptions: (I) The interaction is even, namely, the dynamics αt commutes with the even-oddness automorphism Θ. (Automatically satisfied when (IV) is assumed.) (II) The domain of the generator δα of αt contains the set [Formula: see text] of all strictly local elements of [Formula: see text]. (III) The set [Formula: see text] is the core of δα. (IV) The dynamics αt commutes with lattice translation automorphism group τ of [Formula: see text]. A major technical tool is the conditional expectation from [Formula: see text] onto its C*-subalgebras [Formula: see text] for any subset I of the lattice, which induces a system of commuting squares. This technique overcomes the lack of tensor product structures for Fermion systems and even simplifies many known arguments for spin lattice systems. In particular, this tool is used for obtaining the isomorphism between the real vector space of all *-derivations with their domain [Formula: see text], commuting with Θ, and that of all Θ-even standard potentials which satisfy a specific norm convergence condition for the one point interaction energy. This makes it possible to associate a unique standard potential to every dynamics satisfying (I) and (II). The convergence condition for the potential is a consequence of its definition in terms of the *-derivation and not an additional assumption. If translation invariance is imposed on *-derivations and potentials, then the isomorphism is kept and the space of translation covariant standard potentials becomes a separable Banach space with respect to the norm of the one point interaction energy. This is a crucial basis for an application of convex analysis to the equivalence proof in the major result. Everything goes in parallel for spin lattice systems without the evenness assumption (I).

The problem of existence and uniqueness of a state of a joint system with given restrictions to subsystems is studied for a Fermion system, where a novel feature is non-commutativity between algebras of subsystems.For an arbitrary (finite or infinite) number of given subsystems, a product state extension is shown to exist if and only if all states of subsystems except at most one are even (with respect to the Fermion number). If the states of all subsystems are pure, then the same condition is shown to be necessary and sufficient for the existence of any joint extension. If the condition holds, the unique product state extension is the only joint extension.For a pair of subsystems, with one of the given subsystem states pure, a necessary and sufficient condition for the existence of a joint extension and the form of all joint extensions (unique for almost all cases) are given. For a pair of subsystems with non-pure subsystem states, some classes of examples of joint extensions are given where non-uniqueness of joint extensions prevails.

Highlights d Nanopores of Drosophila olfactory sensillum are modified cuticular envelope d Endocytic membrane structures are associated with the site of nanopore formation d Gore-tex/Osiris23 endosomal protein is required for nanopore formation d gore-tex/Osiris23 mutants showed greatly reduced olfactory response

The distribution of axonal spheroids was examined in the central nervous system of gracile axonal dystrophy (GAD) mutant mice. Only few spheroids are observed in the gracile nucleus of the medulla in normal mice throughout the period examined, while they are first noted in GAD mice as early as 40 days after birth. The incidence of spheroids shifts from the gracile nucleus to the gracile fasciculus of the spinal cord with the progress of disease, suggesting that the degenerating axonal terminals of the dorsal ganglion cells back from the distal presynaptic parts in the gracile nucleus, along the tract of the gracile fasciculus, toward the cell bodies in the dorsal root ganglion. This phenomenon indicates that the distribution of spheroids is age dependent and reflects a dying-back process in degenerating axons. In addition to the gracile nucleus and the gracile fasciculus, which is one of the main ascending tracts of primary sensory neurons, it was noted that the other primary sensory neurons joined with some of the second-order neurons at the dorsal horn and neurons at all levels of the dorsal nucleus (Clarke's column) are also severely affected in this mutant. The incidence of the dystrophic axons are further extended to the spinocerebellar tract and to particular parts of the white matter of the cerebellum, such as the inferior cerebellar peduncle and the lobules of I-III and VIII in the vermis. These results indicate that this mutant mouse is a potential animal model for human degenerative disease of the nervous system, such as neuroaxonal dystrophy and the spinocerebellar ataxia.

The oxidative stability of lipids from salmon roe and herring roe was compared with those of commercial fish oils originated from sardine and tuna. Both fish roe lipids contained high amounts of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). Total EPA and DHA was more than 35% of total fatty acids in both roe lipids. On the basis of oxygen consumption, fish roe lipids showed the higher oxidative stability than both fish oils. This tendency in oxidative stabilities was also confirmed by the determination of propanal formation during oxidation. Analyses of lipid compositions suggested that the higher oxidative stabilities of fish roe lipids would be mainly due to the presence of phospholipids in them. Dietary effects of salmon and herring roe lipids were also determined. Little increase in total cholesterol level was observed in plasma lipids in rats fed salmon and herring roe lipids, although cholesterol content in fish roe lipids were 6.3% and 9.7% of total lipids for salmon roe and herring roe.