The Extensive Nitration of Neurofilament Light Chain in the Hippocampus Is Associated with the Cognitive Impairment Induced by Amyloid β in Mice

Increased nitration of proteins, a surrogate marker of wide-spread oxidative damage in brains affected by the amyloid β peptide (Aβ), is evidently correlated with the severity of cognitive dysfunction in humans as well as animals (Smith et al., 1997; Lim at al., 2001; Perry et al., 2002; Kim et al., 2003; Andersen, 2004; Bastianetto and Quirion, 2004; Walsh and Selkoe, 2004).

We have previously reported the contribution of tyrosine nitration to Aβ-induced, oxidative damage-mediated cognitive dysfunction in mice (Alkam et al., 2007, 2008). A mouse monoclonal anti-nitrotyrosine antibody in Western blot analysis identified the tyrosine-nitrated hippocampal proteins at approximately 70 kDa as a single band with which the severity of cognitive impairments in mice was well associated (Alkam et al., 2007). In this study, we aimed to identify the nitrated proteins in the single band for the specification of the contribution of the extensive nitration of tyrosine to the cognitive impairment. To produce strong and stable nitrative damage, we applied Aβ_25–35, a toxic Aβ fragment that is detected in the human brain (Pike et al., 1995; Kubo et al., 2002). The tyrosine-nitrated proteins were examined by using liquid chromatography/mass spectrometry/mass spectrometry (LC-MS/MS) and immunodetection. Intense nitration of the neurofilament light chain (NFL) was observed. The intensive nitration was associated with serine hyperphosphorylation and reduced interaction of NFL with nuclear distribution element-like (NUDEL), a protein essential for the stable assembly of neurofilaments (NFs). The results provided further support for the conception that extensive nitration of tyrosine in proteins underlies one of the key mechanisms contributing to the cognitive pathology induced by Aβ.

Materials and Methods

Animals. Male ICR mice (Nihon SLC Co., Shizuoka, Japan) were used. The animals were housed in a controlled environment (23 ± 1°C, 50 ± 5% humidity) and allowed access to food and water ad libitum. The room lights were kept on between 8:00 AM and 8:00 PM. All experiments were performed in accordance with the Guidelines for Animal Experiments of Nagoya University Graduate School of Medicine. The procedures involving animals and their care conformed to the Guidelines for Proper Conduct of Animal Experiments (Science Council of Japan, 2006).

Treatment and Experimental Design. Aβ_25–35 (Bachem, Bubendorf, Switzerland) was dissolved in sterile double-distilled water to a stock concentration of 1 mg/ml and stored at -20°C before use. The dissolved Aβ_25–35 was incubated for aggregation at 37°C for 4 days. The distilled water was incubated at the same conditions and used as the vehicle. Aβ_1–40 (Bachem) was dissolved to a stock concentration of 1.0 mg/ml in 35% acetonitrile/0.1% trifluoroacetic acid and stored at -20°C before use. The solution of peroxynitrite (ONOO^-; 144 mM) (Millipore, Billerica, MA) was stored at -80°C before use. Incubated Aβ_25–35 (3 μg/3 μl), incubated distilled water (3 μl), Aβ_1–40 (5 μg/5 μl), and ONOO^- (144 mM/1 μl) were administered by i.c.v. injection as described previously (Maurice et al., 1996; Alkam et al., 2007, 2008). In brief, a microsyringe with a 28-gauge stainless steel needle 3.0-mm long was used for all experiments. Mice were anesthetized lightly with ether, and the needle was inserted unilaterally 1 mm to the right of the midline point equidistant from each eye, at an equal distance between the eyes and the ears and perpendicular to the plane of the skull. A single shot of the indicated volume of agents was delivered gradually within 3 s. Mice exhibited normal behavior within 1 min after the injection. The injection placement or needle track was visible and was verified at the time of dissection. Neither insertion of the needle nor the volume of injection had a significant influence on survival, behavioral responses, or cognitive functions. Uric acid (UA) (Wako Pure Chemicals, Osaka, Japan) was prepared as a suspension in saline. Immediately after the single injection of Aβ_25–35 or Aβ_1–40, mice were given UA (100 mg/kg i.p.) daily for 6 consecutive days. The schedule of administration of peptides and drugs as well as biochemical, histochemical, and behavioral investigations is shown in Fig. 1.

Novel Object Recognition Task. This task, based on the spontaneous tendency of rodents to explore a novel object more often than a familiar one, was performed on days 3 to 5 after the i.c.v. injection of Aβ_1–40, Aβ_25–35, or peroxynitrite (day 0) as described previously (Alkam et al., 2007). A plastic chamber (35 × 35 × 35 cm) was used in low light conditions during the light phase of the light/dark cycle. The general procedure consisted of three different phases: 1) a habituation phase, 2) an acquisition phase, and 3) a retention phase. On the 1st day (habituation phase), mice were individually subjected to a single familiarization session of 10 min, during which time they were introduced into the empty arena to become familiar with the apparatus. On the 2nd day (acquisition phase), the animals were subjected to a single 10-min session, during which time two floor-fixed objects (A and B) were placed in a symmetric position from the center of the arena, 15 cm from each other and 8 cm from the nearest wall. The two objects, made of the same wooden material with a similar color and smell, were different in shape but identical in size. Mice were allowed to explore the objects in the open field. A preference index for each mouse was expressed as a ratio of the amount of time spent exploring object A (TA × 100)/(TA + TB), where TA and TB are the time spent exploring object A and object B, respectively. On the 3rd day (retention phase), mice were allowed to explore the open field in the presence of two objects: the familiar object A and a novel object C in different shapes but in similar color and size (A and C). A recognition index, calculated for each mouse, was expressed as the ratio (TC × 100)/(TA + TC), where TA and TC are the time spent during the retention phase on object A and object C, respectively. The time spent exploring the object (nose pointing toward the object at a distance ≤ 1 cm) was recorded by hand.

Sample Preparation. Animals were decapitated, and the hippocampi were removed on an ice-cold glass plate and stored at -80°C. Hippocampal protein extracts were obtained by homogenization in diverse ice-cold lysis buffers that included the radioimmunoprecipitation assay (RIPA) buffer, phosphate-buffered saline (PBS) buffer, Triton X-100 buffer, and 6 M urea buffer. The RIPA lysis buffer contained 20 mM trizma hydrochloride, pH 7.6, 150 mM sodium chloride, 2 mM EDTA·2Na, 50 mM sodium fluoride, 1 mM sodium vanadate, 1% Nonidet P-40, 1% sodium deoxycholate, 0.1% SDS, 1 mg/ml pepstatin, 1 mg/ml aprotinin, and 1 mg/ml leupeptin. The PBS lysis buffer, pH 7.4, contained 135 mM sodium chloride, 3.2 mM disodium hydrogen phosphate 12-water, 1.3 mM potassium chloride, and 0.5 mM potassium dihydrogen phosphate. The Triton X-100 buffer contained 10 mM trizma hydrochloride at pH 7.5, 150 mM sodium chloride, 1 mM EDTA at pH 8.0, and 1% Triton X-100. The 6 M urea lysis buffer contained 10 mM trizma base at pH 8.1, 6 M urea, and 1 mM dithiothreitol. All of these lysis buffers, with the exception of the RIPA buffer, were supplemented with complete protease inhibitor cocktail tablets (Roche Applied Science, Mannheim, Germany). Homogenates were centrifuged at 13000g for 20 min to obtain the desired supernatant of the extracts. The centrifuged pellets were washed twice with the previous buffer before being solubilized. The washing procedure consisted of complete dispersion of the pellets by vortexing and incubation in ice for 30 min followed by centrifugation at 13000g for 20 min. The unassembled NFL and NUDEL proteins were obtained within the soluble proteins in Triton X-100 buffer (Nguyen et al., 2004), and the insoluble protein pellets that include the assembled NFL and NUDEL proteins were then solubilized in 6 M urea lysis buffer (Crow et al., 1997). The cytoplasmic water-soluble proteins were obtained in PBS lysis buffer (Aoyama and Kitajima, 1999), and the insoluble pellets were then solubilized in Triton X-100 buffer. The concentrations of PBS-soluble and ureasoluble proteins were determined with a Bio-Rad protein assay reagent kit (Bio-Rad, Hercules, CA). The concentrations of the Triton X-100-soluble proteins were determined with a BCA protein assay reagent kit (Pierce, Rockford, IL).

Western Blot Analysis. Equal amounts (20 μg) of protein sample were resolved bya4to20% gradient or 7% SDS-polyacrylamide gel electrophoresis (PAGE). The proteins were then transferred electrophoretically to a polyvinylidene difluoride (PVDF) membrane (Millipore). Membranes were incubated in 3% skim milk or 3% bovine serum albumin (for phosphor-protein) in phosphate-buffered saline containing 0.05% (v/v) Tween 20 for 2 h at room temperature. Anti-nitrotyrosine mouse monoclonal 1A6 antibody (catalog number 05-233; Millipore), anti-NFL mouse antibody (Sigma-Aldrich, St. Louis, MO), anti-heat shock protein 70 (HSP70) polyclonal antibody (Assay Designs, Ann Arbor, MI), anti-dihydropyrimidinase-like 2 (DRP-2) mouse antibody (IBL, Takasaki, Japan), anti-NUDEL rabbit antibody (Abcam Inc., Cambridge, MA), anti-phosphoserine rabbit antibody (Zymed Laboratories, South San Francisco, CA), anti-β-actin goat antibody (Santa Cruz Biotechnology Inc., Santa Crutz, CA), and anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mouse antibody (Imgenex, San Diego, CA) were used. To confirm the specificity of the detected single band of tyrosine-nitrated proteins, the reduction of nitrotyrosine to aminotyrosine was performed. In brief, the membrane was treated with 10 mM sodium dithionite (SD) in 50 mM pyridine-acetate buffer, pH 5.0, for 1 h at room temperature. After the reaction, the membrane was rinsed with distilled water and then equilibrated with washing buffer and blocked for 1 h at room temperature in blocking solution before standard procedures of Western blotting were followed. The absent band in the SD-treated membrane compared with the routine-treated control membrane was regarded to be a genuine for nitrated proteins. To confirm the specificity of the detected band for phosphoserine, the anti-phosphoserine inhibitor (the inhibitor) that contains phosphoserine was used to block the specific interaction of anti-phosphoserine primary antibodies with serine-phosphorylated proteins in the membrane. In brief, the anti-phosphoserine primary antibody and the inhibitor at a final concentration of 20 mM were mixed into a bovine serum albumin-containing blocking buffer and preincubated for 10 min for the ample binding of the antibodies with the phosphoserines (to cover up all of the specific anti-phosphoserine antibodies) before the application to the membrane. Incubation of the antibody-inhibitor mixture with the membrane was carried out for 1 h at room temperature. After the incubation, standard procedures were followed for blot washing and incubation with a secondary antibody. The absence of the bands in the membrane after the antibody-inhibitor treatment compared with the membrane subjected to routine treatment was regarded as genuine proof of serine phosphorylation. The intensity of each protein band on the film was analyzed with the Atto Densitograph 4.1 system (Atto, Tokyo, Japan) and was corrected with the corresponding β-actin or GAPDH level. The results were expressed as a percentage of that in the naive group.

Liquid Chromatography/Mass Spectrometry/Mass Spectrometry. Protein bands in the SDS-PAGE were stained with Coomassie Brilliant Blue (CBB) (Fluka, Buchs, Switzerland). The band of interest was excised from the gel. The gel piece was digested in trypsin solution at 35°C for 20 h for analysis by LC/MS/MS (Aproscience Lifescience Institute, Tokushima, Japan).

Immunoprecipitation. Hippocampal homogenates for Western blottings were used for immunoprecipitation. The antibodies against the proteins of interest were incubated overnight with 50 μl of protein A-Sepharose beads (GE Healthcare, Little Chalfont, Buckinghamshire, UK). To obtain tyrosine-nitrated proteins, anti-nitrotyrosine agarose-conjugated mouse antibody (Millipore) was used. The bead-antibody complexes were incubated overnight with 500 μg of precleared proteins in the corresponding buffers, with the exception that urea lysis buffer does not include dithiothreitol. Immunocomplexes were collected by centrifugation at 13000g for 1 min at 4°C and then washed three times with ice-cold PBS. Immunoprecipitated samples were recovered by resuspending in 2× sample loading buffer, immediately fractionated by reducing in 7% SDS-PAGE, and analyzed by Western blotting with the corresponding antibodies.

Histology. Each mouse was anesthetized with diethyl ether and quickly intracardially perfused with physiological saline followed by 4% paraformaldehyde in 100 mM PBS, pH 7.4. The brains were quickly removed, postfixed for 24 h in the same fixative solution, and cryoprotected in a graded 10 to 40% sucrose solution in 100 mM PBS. Coronal sections were cut 20-μm thick using a cryostat (Leica, Wetzlar, Germany) and stained with 0.1% cresyl violet reagent (Wako Pure Chemicals) according to standard procedures. The sections were mounted in fluorescent medium (Dako North America, Inc., Carpinteria, CA), and images of CA1, CA3, and the granular layer of the dentate gyrus of the hippocampus were taken using a Carl Zeiss Axioskop phase-contrast microscope with a cooled CCD camera system (SenSys; Photometrics Ltd., Tucson, AZ). The Nissl-positive neuronal cells were counted using Image J software (version 1.38; National Institutes of Health, Bethesda, MD). The total cell count in per millimeter square was averaged from four sections per animal (n = 4) according to previous reports (Nabeshima et al., 1991; Nitta et al., 1997).

Statistical Analysis. The results are expressed as the mean ± S.E. Statistical significance was determined with a one-way analysis of variance followed by the Bonferroni multiple comparisons test. p < 0.05 was taken as a significant level of difference.

Results

The Tyrosine Nitration of Proteins Induced by Aβ_25–35 in the Hippocampus of Mice. Anti-nitrotyrosine mouse antibody detected only a single band of hippocampal proteins at approximately 70 kDa for tyrosine nitration, which induced a potent nitrating agent after the i.c.v. injection of Aβ_1–40, Aβ_25–35, and peroxynitrite (ONOO^-) (Fig. 2A). Aβ peptides induced extensive nitration of proteins in the hippocampus and impairment of recognition memory, both of which were prevented by UA, a potent scavenger of ONOO^-. ONOO^- induced marked tyrosine nitration of proteins in the hippocampus and impairment of recognition memory (Fig. 2, B and C). The intensity of the nitration was inversely associated with the recognition memory in mice (Fig. 2D). The authenticity of nitration was confirmed by the reduction of nitrotyrosine to aminotyrosine with SD in the membrane and by detecting the nitrotyrosine using the same antibody. The absence of this band after SD treatment was regarded as a genuine band for proteins with tyrosine nitration (Fig. 2, E and F). Proteins in SDS-PAGE were stained with CBB, and the 70-kDA protein band was excised for identification (Fig. 2G). The proteins in the excised gel were in-gel-trypsin-digested and subjected to LC/MS/MS, and several proteins were successfully identified (Table 1).

The Identification of the Tyrosine-Nitrated Proteins and the Level of Nitration. The nitration of the identified proteins was examined by applying the immunoprecipitation method. For peptide match scores, HSP70, DRP-2, and NFL were favored for the further study. Because the antibody that was used to detect the nitrated proteins in Western blot analysis could not be used for immunoprecipitation, a specially designed agarose-conjugated mouse anti-nitrotyrosine monoclonal antibody was used. We applied Aβ_25–35 for the rest of the study, considering its property to produce stronger and stable oxidative damage (Pike et al., 1995) as evidenced in Fig. 2. Immunoprecipitated nitrated-proteins were fractionated by SDS-PAGE and blotted with the antibodies raised against the proteins of interest (Fig. 3A). Intensive nitration was observed for NFL in the Aβ_25–35 group compared with the naive or vehicle group (Fig. 3, A and B). No differences were observed in the nitration of HSP70 and DRP-2 proteins among the three groups (Fig. 3, A, C, and D). The increased nitration of NFL was inversely associated with recognition memory in mice that received Aβ_25–35 injections (Fig. 3E).

Association between Extensive Nitration of NFL and Serine Hyperphosphorylation. Hyperphosphorylation of the serine residues of NFL could lead to disruption of the subtle regulation of the NF network (Hisanaga et al., 1990; Nixon and Shea, 1992). After being nitrated in vitro, NFL is not able to form the NF assembly (Crow et al., 1997). The question of whether extensive nitration of NFL influences serine phosphorylation of the protein stimulated our interest. We immunoprecipitated NFL and blotted against nitrotyrosine and phosphoserine. Equal amounts of NFL protein were immunoprecipitated in each group (Fig. 4, A and B). The intensity of the tyrosine nitration and serine phosphorylation of NFL was greater in the Aβ_25–35 group than in the naive or vehicle group (Fig. 4, A, C, and D). The authenticity of the phosphoserine band was confirmed as indicated under Materials and Methods. Treatment with UA prevented the Aβ_25–35-induced intensive tyrosine nitration and serine hyperphosphorylation of NFL (Fig. 4, A, C, and D), indicating a positive association between the extensive nitration of NFL and the serine hyperphosphorylation (Fig. 4E).

Association between Extensive Nitration of NFL and Its Reduced Interaction with NUDEL. To examine whether the extensive nitration of NFL practically influences its interaction with partner proteins, we focused on the free, unassembled NFL that could be differentiated from the assembled NFL. The majority of the newly synthesized unassembled NF proteins, including NFL, are Triton X-100-soluble before being incorporated into the NF assembly, which is Triton X-100-insoluble (Black et al., 1986). NFL constitutes the core of the NF network, and without NFL, no filaments are formed (Zhu et al., 1997). Without binding directly with NUDEL, the Triton X-100-soluble NFL can barely lead the assembly of a stable NF network, regardless of its own abundance (Nguyen et al., 2004). We probed equal amounts of NFL immunocomplexes with antibodies raised against the nitrotyrosine and NUDEL (Fig. 5, A and B). Less NUDEL was coimmunoprecipitated in the Aβ_25–35 group that bears extensively nitrated NFL (Fig. 5, A–D). The protein expression of NUDEL did not differ among the groups (Fig. 5E). UA prevented the Aβ_25–35-induced increase of NFL nitration as well as the reduced coimmunoprecipitation of NUDEL (Fig. 5, A, C, and D). The extensive nitration of NFL was associated with its reduced interaction with NUDEL (Fig. 5F). These results suggested that the intensive nitration of NFL could disturb the normal function of the protein.

Association between Extensive Nitration of NFL and the Reduced Content of NUDEL in the Cytoskeleton Fraction. A majority of NF proteins, after their synthesis in the cytoplasm, are rapidly converted to a Triton X-100-insoluble filamentous network and move down the axon using the transport machinery (Nixon and Shea, 1992). After direct and specific binding with NFL, NUDEL facilitates the assembly of a stable NF network and remains bound to the assembled filaments (Nguyen et al., 2004). Thus, the level of interaction between NFL and NUDEL in cytoplasm (Triton X-100-soluble fraction) should be reflected by their protein levels in the axonal cytoskeleton (Triton X-100-insoluble fraction). The Triton X-100-insoluble fractions from the previous step (Fig. 5) were washed twice with Triton X-100 lysis buffer before being solubilized in urea lysis buffer. Western blot analysis revealed that the level of NUDEL protein was reduced in the Aβ_25–35 group compared with the naive and vehicle groups, whereas the treatment with UA prevented the reduction (Fig. 6, A and D). This was consistent with the reduced interaction between NFL and NUDEL in the Aβ_25–35 group (Fig. 5, A and D). However, the level of NFL in the Aβ_25–35 group was surprisingly not different from that in the naive and vehicle groups (Fig. 6, A and B). Considering the increase of the intensity of the protein nitration in the Aβ_25–35 group (Fig. 6, A and C), we examined the nitration of NFL by immunoprecipitation. Intense nitration for the NFL protein in the Aβ_25–35 group was observed (Fig. 6E). Applying the multiplicative inverse (in which the inverse or reciprocal of “n” is “1/n”), a mathematical method that is useful in medical science (Silberberg, 1990), the reciprocal level of the extensively nitrated NFL in the Triton X-100-insoluble fraction was estimated (Fig. 6F). The reciprocal level of extensively nitrated NFL in the Aβ_25–35 group paralleled with that of NUDEL in the same group (Fig. 6, D and F), signifying a negative effect of the extensive nitration of NFL on NUDEL-dependent NF assembly. The increased nitration of tyrosine could modify protein function by altering the three-dimensional conformation and hydrophobicity (Dalle-Donne et al., 2005; Reynolds et al., 2007). It was therefore assumed that the overnitrated, free NFL would become less Triton X-100 soluble and, as a result, would be detected in the Triton X-100-insoluble fraction along with the assembled NF proteins. It is hardly practical to separate the unassembled extensively nitrated NFL from the assembled NFL in the Triton X-100-insoluble fraction. The majority of the cytoplasmic water-soluble proteins could be separated from the Triton X-100-soluble protein pools by using PBS lysis buffer in the first step (Aoyama and Kitajima, 1999). After the separation of the PBS-soluble and Triton X-100-soluble proteins as described under Materials and Methods, we examined the amount of NFL protein in these two different fractions. The majority of NFL protein in all groups was found in the PBS-soluble cytoplasmic fraction as indicated by GADPH, a cytoplasmic marker (Fig. 7A). The levels of NFL protein in both the PBS-soluble and Triton X-100-soluble fractions were increased in the Aβ_25–35 group (Fig. 7, A–C). It is interesting to note that the increase of NFL in both fractions was prevented by the treatment with of UA, a potent scavenger of ONOO^-, suggesting that the Aβ_25–35-induced ONOO^- may increase the protein synthesis of NFL before extensively nitrating the protein (Fig. 7, A–C). The Triton X-100-soluble NFL that became insoluble in PBS in the Aβ_25–35 group was extensively nitrated (Fig. 7D), and the intensity of nitration was associated with the level of the PBS-insoluble, Triton X-100-soluble NFL (Fig. 7E). These results revealed new possibilities for Triton X-100-insolubile NFL in association with extensive nitration.

The Cell Numbers in the Hippocampus of Mice with the Impairment of Memory Induced by Aβ_25–35. On day 5 after the i.c.v. injection of Aβ_25–35, cell numbers in CA1, CA3, and the granular layer of the dentate gyrus of the hippocampal formation were examined using cresyl violet staining. The quantification of the stained cells revealed no cell loss induced by Aβ_25–35 (Table 2). These results were consistent with reports that at a dose of 3 to 5 μg, Aβ_25–35 could induce memory impairment but not cell loss within a time session of 1 month after its injection in mice (Maurice et al., 1996; Tohda et al., 2003). These results suggest that cell loss was not involved in the impairment of memory induced by Aβ_25–35 in mice.

Discussion

Neuronal oxidative damage has long been hypothesized as a critical mechanism of cellular dysfunction in neurodegenerative ailments (Perry et al., 2002). Reports showing that antioxidants delay or reduce progressive cognitive decline in both animal models and humans have emphasized the direct contribution of oxidative damage to cognitive pathology (Sano et al., 1997; Yamada et al., 1999; Lim et al., 2001). Oxidative damage is generally manifested by the increase of lipid peroxidation, DNA oxidation, protein oxidation, and peroxynitrite-mediated tyrosine nitration of proteins. The increased nitration of tyrosine could irreversibly disrupt the function of proteins (Koppal et al., 1999), and it might play a key pathogenic role in the progression of cognitive impairment (Smith et al., 1997; Keller, 2006). Until now, various proteins with tyrosine nitration have been reported in association with neurodegeneration and cognitive decline (Strong et al., 1998; Castegna et al., 2003; Tran et al., 2003; Sacksteder et al., 2006; Sultana et al., 2006). The diversity of nitrated proteins in these reports seems to depend on the species of the sources of samples (Sacksteder et al., 2006; Sultana et al., 2006), the proteomic detections on various conditions (Castegna et al., 2003; Sultana et al., 2006), immunodetections by means of different anti-nitrotyrosine antibodies with the diverse recognition property for nitrotyrosine (Strong et al., 1998; Tran et al., 2003), or the biological selectivity of tyrosine nitration (Ischiropoulos, 2003; Sacksteder et al., 2006). Dissimilar reports about the nitrated proteins in the brains of humans with Alzheimer's disease (AD) (Castegna et al., 2003; Sultana et al., 2006) emphasize the importance of the sources of protein, even in the same species or under the same conditions of detection during the identification process, while illustrating the diversity of nitration due to the dissimilar expression of proteins during the different stages of the disease.

In the present study, we looked for further evidence for the pathogenic role of protein nitration as one of the key contributors to the decline of cognitive function induced by Aβ. Using LC-MS/MS and immunodetection, we identified the hippocampal proteins with nitrated tyrosine residues after the i.c.v. injection of Aβ_25–35 in mice. Preferentially, in respect with currently examined proteins, intense nitration of NFL was observed, demonstrating a good correlation with the severity of cognitive impairment induced by Aβ_25–35.

NFL, one of the three subunits of NF proteins, is the indispensable core of the NF assembly (Zhu et al., 1997). Studies have reported that NFL is selectively nitrated compared with the majority of other proteins present in brain homogenates, and they suggested that newly synthesized free NFL is particularly susceptible to peroxynitrite-mediated nitration (Crow et al., 1997; Strong et al., 1998). The extensively nitrated NFL inhibits the assembly of unmodified NF subunits (Crow et al., 1997). On the other hand, the extensive serine phosphorylation of NFL could sufficiently block NF assembly (Nixon and Shea, 1992; Gibb et al., 1996). Therefore, we have evaluated the effect of tyrosine nitration on the phosphorylation of NFL at serine residues in general. The increased tyrosine nitration of NFL was associated with its serine hyperphosphorylation. Prevention of the extensive nitration of NFL by UA, a scavenger of ONOO^- that nitrates proteins, restrained the serine phosphorylation of NFL at a normal level. The results indicated that the increased nitration of NFL could give rise to its serine hyperphosphorylation.

NFL requires direct binding with NUDEL, whereas NUDEL can not directly bind with other subunits of NF proteins, to initiate the assembly of NF (Nguyen et al., 2004). After the assembly of the NF network, NUDEL remains bound to the assembled Triton X-100-insoluble neurofilaments and may promote, in conjunction with molecular motors, the axonal transport of the neurofilaments (Nguyen et al., 2004). Thus, the level of interaction between NFL and NUDEL in the Triton X-100-soluble cytoplasmic fraction could be reflected by their protein levels in the Triton X-100-insoluble cytoskeletal fraction. In the current study, the increased nitration of Triton X-100-soluble NFL proteins in the Aβ_25–35 group was associated with its decreased interaction with NUDEL. In the Triton X-100-insoluble fraction, the protein level of NUDEL was reduced in the Aβ_25–35 group, and the reduction was prevented by treatment with UA. In the same fraction, the protein level of NFL surprisingly did not differ among groups, whereas the intensity of the nitration of NFL was strong in Aβ_25–35 group. Estimation by the multiplicative inverse approach indicated that the reduced level of nonextensively nitrated NFL in the Aβ_25–35 group parallels with that of NUDEL. These results required an explanation for the detection of the extensively nitrated NFL in the Triton X-100-insoluble cytoskeletal fraction, because the assembled NFL is nitration-resistant and the intensely nitrated NFL can not participate in the NF assembly (Crow et al., 1997). The alteration of the solubility of the overnitrated NFL might be involved in the detection of the extensively nitrated NFL in the Triton X-100-insoluble cytoskeletal fraction in the Aβ_25–35 group. Interpretation of the emergence of the intensely nitrated NFL in PBS-insoluble, but Triton X-100-soluble, protein pools in the Aβ_25–35 group indicates that extensive nitration would render NFL protein to have poor solubility in PBS. By this rate, it is possible that a considerable level of overnitrated NFL protein in the Aβ_25–35 group would even become Triton X-100 insoluble over a period of time, and that it would be detected along with a reduced level of NUDEL-associated assembled NFL, which is also Triton X-100 insoluble. The observation of detectable levels of nitration in NFL in the RIPA-soluble, Triton X-100-soluble, and Triton X-100-insoluble fractions in the naive and vehicle groups implies that natural nitration of tyrosine, as serine phosphorylation, might exist as a physiological property of NFL and might not be detrimental to the function of the protein, whereas extensive nitration is detrimental. The nitration-susceptible tyrosine residues of NFL are identified particularly as tyrosine 17 in the head region and tyrosines 138, 177, and 265 in the α-helical coil regions of the rod domain of the protein (Crow et al., 1997). It needs to be determined which tyrosine residue is the site for natural nitration or for extensive nitration. It has been reported that, although the exact mechanism is not clear, the newly synthesized Triton X-100-soluble NF proteins, including NFL, could separately undergo axonal transport before being incorporated into the Triton X-100-insoluble axonal cytoskeleton (Jung et al., 1998). We do not know whether the NFL proteins with natural nitration undergo axonal transport after the NF assembly or undergo axonal transport before being incorporated into the Triton X-100-insoluble axonal cytoskeleton.

The observation of no cell loss in CA1, CA3, and the granular layer of the dentate gyrus of the hippocampus in mice that received Aβ_25–35 injections favored the contribution of extensive nitration of NFL to the impairment of memory. A recent study demonstrated that rapidly formed fresh amyloid plaques cause axonal and dendritic structural changes within a minimum of 5 days after the “birth of the plaques” (Meyer-Luehmann et al., 2008). Given the time windows of Aβ neurotoxicity, Aβ_25–35 may require longer time to cause cell loss in our mouse model of cognitive impairment.

The disrupted interaction between NFL and NUDEL is regarded as the most important factor for the destabilization of the NF assembly that leads to the axonal dysfunction, which is an early event in the cognitive pathology of AD (Nguyen et al., 2004; Stokin et al., 2005). Therefore, our results suggest that disrupted interaction between NUDEL and NFL with extensive nitration could be one of the major factors that associated with the cognitive dysfunction induced by Aβ in mice (Fig. 8). However, further studies are required to investigate whether the extensive nitration of NFL and the impaired interaction with NUDEL induced by Aβ are associated with the disruption of axonal transport.