Control de parámetros máquina

As previously discussed, abnormally phosphorylated NFs accumulate early in DNs found in preclinical AD (Dickson et al. 1999; Woodhouse et al. 2009). Neurofilament triplet proteins belong to the type IV intermediate cytoskeletal filament family, and are predominantly expressed in a subpopulation of neurons deriving from pyramidal cells of cortical layers 2-6 that give rise to corticocortical connections (Vickers and Costa 1992; Hof et al. 1995; van der Gucht et al. 2007; Paulussen et al. 2011). The NF ‘triplet’ refers to the three genetically and structurally interrelated subunits [68 kDa (NF-L), 160 kDa (NF-M) and 200 kDa (NF-H)] that co-express and co-polymerize to form intermediate filaments in this subset of neurons (Vickers and Costa 1992; Lee and Cleveland 1996; Kirkcaldie et al. 2002; Lariviere and Julien 2003). In the rat neocortex, NF-immunopositive pyramidal neurons account for approximately 10-13% of all neurons (Kirkcaldie et al. 2002), whereas 20-30% of neurons in human temporal cortex are NF-immunopositive (Hof et al. 1990). The effect of AD on NF-containing neuronal populations has been studied as a possible insight to the basis of neuronal vulnerability. In macaques, the

30 vast majority (45-90%) of long association pathways interconnecting the frontal, temporal and parietal neocortex are NF-immunopositive, while short corticocortical, callosal and limbic pathways are characterized by lower numbers of such neurons (4- 35%) (Hof et al. 1995).

Numerous studies employing the SMI32 antibody (recognizes dephosphorylated epitopes of the NF-M and NF-H subunits but does not cross-react with paired helical filaments, PHF-tau), indicate that a subset of layer 2, 3 and 5 pyramidal neurons containing NFs may be particularly susceptible to neurofibrillary pathology (Lewis et al. 1987; Morrison et al. 1987; Hof et al. 1990; Hof and Morrison 1990; Hof et al. 1995; Mann et al. 1996). NF triplet-containing neurons in superior frontal and inferior medial temporal association cortices (Hof et al. 1990), primary and secondary visual cortex (Hof and Morrison 1990) and hippocampal and entorhinal regions (Vickers et al. 1992, 1994) show a high degree of vulnerability for NFT formation and degeneration. Essentially, this vulnerable neuronal population is the myelinated subset of pyramidal neurons that give rise to specific corticocortical projections in the mammalian brain (Hof et al. 1995). Interestingly, cortical neurons that lack the NF-triplet, such as some inhibitory interneurons, do not develop NFTs and show a much lower susceptibility to degeneration in AD (Hof et al. 1991, 1993; Sampson et al. 1997; discussed below). Conversely, subpopulations of inhibitory neurons which do express the NF triplet are more susceptible to AD pathology and develop NFTs (Sampson et al. 1997). These findings provide evidence that a neuron’s content of the NF-triplet correlates to its specific vulnerability to AD-like changes, and that these dephosphorylated NFs form focal accumulations that correspond to initial NFT formation that is distinct from tau-labelled NFTs (Vickers et al. 1992, 1994).

31 Studies have also demonstrated that specific NF-related changes occur in AD, such as abnormal, age-related phosphorylation of NF triplets in neuronal somata (labelled with SMI312 which recognizes phosphorylated epitopes of the NF-M and NF-H subunits; Masliah et al. 1993). Indeed, NFs found near Aβ plaques harbour multiple novel phosphorylation sites that are not normally expressed (Liao et al. 2004). This is accompanied by widespread increases in hyperphosphorylated forms of all three NF subunits in both AD brains (Wang et al. 2001) and mouse AD model tissue (Yang et al. 2009). Expression patterns of neurofilaments are also perturbed: there are reductions in NF-L mRNA levels (McLachlan et al. 1988) and immunoreactivity (Nakamura et al. 1997). There are also significant increases in the levels of NF-M and NF-H (Brettschneider et al. 2006), as well as NF-L (Pijnenburg et al. 2007) in the cerebrospinal fluid of AD patients, likely indicating increased axonal degeneration. Interestingly this is accompanied by increased intrathecal production of autoantibodies to NFs, suggesting immune responses may also be involved (Bartos et al. 2012).

It has also been demonstrated in AD cases that NFs are found within a subset of plaque-associated dystrophic neurites (DNs) distinct from those labelled with thioflavine S and tau (Masliah et al. 1993; Nakamura et al. 1997; Dickson et al. 1999). A subset of cortical pyramidal neurons in the supragranular layer of human cortex are α-internexin positive while NF-triplet negative, defining a distinct subpopulation (Dickson et al. 2005). Also a type IV intermediate filament, α- internexin is closely related to and associated with the NF-triplet (Yuan et al. 2006). Its expression is highest during development, decreasing in favour of NF-triplet expression with maturity (Leriviere and Julien 2003). α-Internexin immunopositive neurons become involved relatively later in AD-related changes and degeneration

32 than those that are NF-triplet positive (Dickson et al. 2005), suggesting a hierarchical neuronal vulnerability that depends on the type of intermediate filament protein expressed. For example, DNs immunopositive for triplet NFs appear early (Masliah et al. 1993; Nakamura et al. 1997; Dickson et al. 1999) while those that are uniquely

α-internexin-immunoreactive become involved much later on (Dickson et al. 2005; Woodhouse et al. 2009).

Cumulatively, these findings suggest that NFs correlate with susceptibility to AD in the select subpopulation of neurons that express them, due to a predisposition to developing cellular changes that result in neurofibrillary pathology in AD. Recent advances in functional brain imaging techniques have also revealed certain brain networks (‘hubs’) that are particularly susceptible to AD pathology even in presymptomatic stages (Buckner et al. 2009; Brier et al. 2012). It would be interesting to determine if regional vulnerability in AD is conferred by network- intrinsic properties (i.e ‘top-down’), the cell-autonomous features of their constituents (i.e. ‘bottom-up’), or a combination of the two.