Sites of Tau Important for Aggregation Populate β-Structure and Bind to Microtubules and Polyanions
2005; Elsevier BV; Volume: 280; Issue: 26 Linguagem: Inglês
10.1074/jbc.m501565200
ISSN1083-351X
AutoresMarco D. Mukrasch, Jacek Biernat, Martin von Bergen�, Christian Griesinger, Eckhard Mandelkow�, Markus Zweckstetter,
Tópico(s)Advanced Neuroimaging Techniques and Applications
ResumoThe aggregation of the microtubule-associated tau protein and formation of "neurofibrillary tangles" is one of the hallmarks of Alzheimer disease. The mechanisms underlying the structural transition of innocuous, natively unfolded tau to neurotoxic forms and the detailed mechanisms of binding to microtubules are largely unknown. Here we report the high-resolution characterization of the repeat domain of soluble tau using multidimensional NMR spectroscopy. NMR secondary chemical shifts detect residual β-structure for 8–10 residues at the beginning of repeats R2–R4. These regions correspond to sequence motifs known to form the core of the cross-β-structure of tau-paired helical filaments. Chemical shift perturbation studies show that polyanions, which promote paired helical filament aggregation, as well as microtubules interact with tau through positive charges near the ends of the repeats and through the β-forming motifs at the beginning of repeats 2 and 3. The high degree of similarity between the binding of polyanions and microtubules supports the hypothesis that stable microtubules prevent paired helical filament formation by blocking the tau-polyanion interaction sites, which are crucial for paired helical filament formation. The aggregation of the microtubule-associated tau protein and formation of "neurofibrillary tangles" is one of the hallmarks of Alzheimer disease. The mechanisms underlying the structural transition of innocuous, natively unfolded tau to neurotoxic forms and the detailed mechanisms of binding to microtubules are largely unknown. Here we report the high-resolution characterization of the repeat domain of soluble tau using multidimensional NMR spectroscopy. NMR secondary chemical shifts detect residual β-structure for 8–10 residues at the beginning of repeats R2–R4. These regions correspond to sequence motifs known to form the core of the cross-β-structure of tau-paired helical filaments. Chemical shift perturbation studies show that polyanions, which promote paired helical filament aggregation, as well as microtubules interact with tau through positive charges near the ends of the repeats and through the β-forming motifs at the beginning of repeats 2 and 3. The high degree of similarity between the binding of polyanions and microtubules supports the hypothesis that stable microtubules prevent paired helical filament formation by blocking the tau-polyanion interaction sites, which are crucial for paired helical filament formation. Alzheimer disease is characterized by abnormal protein deposits in the brain, such as amyloid plaques or neurofibrillary tangles, formed by fibrous assemblies of the Aβ peptide (1Selkoe D.J. Schenk D. Annu. Rev. Pharmacol. Toxicol. 2003; 43: 545-584Crossref PubMed Scopus (754) Google Scholar) or of the microtubule (MT) 1The abbreviations used are: MT, microtubule; PHF, paired helical filament; MES, 4-morpholineethanesulfonic acid; Pipes, 1,4-piperazinediethanesulfonic acid; HSQC, heteronuclear single quantum correlation. -associated tau protein (2Mandelkow E.M. Mandelkow E. Trends Cell Biol. 1998; 8: 425-427Abstract Full Text Full Text PDF PubMed Scopus (445) Google Scholar). These aggregates are thought to be toxic to neurons, either by causing some toxic signaling defect or by obstructing the cell interior. Therefore, one of the top priorities in Alzheimer research is to understand the reasons for the pathological aggregation and to find methods to prevent it. Although the structural principles governing Aβ aggregation are known in some detail, little is known for the tau protein. Tau is a microtubule-associated protein that regulates MT stability, neurite outgrowth, and other MT-dependent functions. The three or four repeats in the C-terminal half of the protein and the flanking proline-rich basic domains are known to be involved in MT binding (3Gustke N. Trinczek B. Biernat J. Mandelkow E.M. Mandelkow E. Biochemistry. 1994; 33: 9511-9522Crossref PubMed Scopus (547) Google Scholar). The affinity is regulated by phosphorylation particularly at KXGS-motifs in the repeats (4Biernat J. Gustke N. Drewes G. Mandelkow E.M. Mandelkow E. Neuron. 1993; 11: 153-163Abstract Full Text PDF PubMed Scopus (662) Google Scholar). Interestingly the same phosphorylation sites have an inhibitory influence on aggregation (5Schneider A. Biernat J. von Bergen M. Mandelkow E. Mandelkow E.M. Biochemistry. 1999; 38: 3549-3558Crossref PubMed Scopus (460) Google Scholar). Unbound tau can assemble into Alzheimer-like paired helical filaments (PHFs) whose polymerization can be enhanced by oxidation of SH groups and by polyanions (e.g. heparin, poly-Glu (6Barghorn S. Mandelkow E. Biochemistry. 2002; 41: 14885-14896Crossref PubMed Scopus (286) Google Scholar)). On the other hand, tau has a hydrophilic character, is highly soluble, and belongs to the class of natively unfolded proteins with no apparent ordered secondary structure detectable by far-UV CD or Fourier-transform infrared spectroscopy (7Schweers O. Schonbrunn-Hanebeck E. Marx A. Mandelkow E. J. Biol. Chem. 1994; 269: 24290-24297Abstract Full Text PDF PubMed Google Scholar, 8von Bergen M. Friedhoff P. Biernat J. Heberle J. Mandelkow E.M. Mandelkow E. Mol. Biol. Cell. 2000; 11: 363AGoogle Scholar). Therefore, it is unclear why tau should aggregate in a specific manner and what structural principles could be responsible for this. Tau can aggregate as an intact protein, 352–441 residues in length (depending on isoform), so that all six tau isoforms are found in Alzheimer PHFs (9Buee L. Hamdane M. Delobel P. Sambo A.V. Begard S. Ghestem A. Sergeant N. Delacourte A. J. Soc. Biol. 2002; 196: 103-108Crossref PubMed Scopus (7) Google Scholar). The isoforms differ by two inserts near the N-terminal end and the presence of either four or three imperfect repeat sequences in the C-terminal half of the protein (see Fig. 1). The region comprising the repeat sequences forms the core of PHFs (10Wischik C.M. Novak M. Thogersen H.C. Edwards P.C. Runswick M.J. Jakes R. Walker J.E. Milstein C. Roth M. Klug A. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 4506-4510Crossref PubMed Scopus (821) Google Scholar) and also promotes PHF assembly in vitro (11Wille H. Drewes G. Biernat J. Mandelkow E.M. Mandelkow E. J. Cell Biol. 1992; 118: 573-584Crossref PubMed Scopus (436) Google Scholar, 12Friedhoff P. von Bergen M. Mandelkow E.M. Davies P. Mandelkow E. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 15712-15717Crossref PubMed Scopus (295) Google Scholar). For PHF aggregation two hexapeptides at the beginning of the second and third repeats (275VQI-INK280 and 306VQIVYK311) are crucial. These are able to initiate the aggregation of tau into bona fide paired helical filaments with cross-β-structure and thereby represent minimal tau-tau interaction motifs (13von Bergen M. Barghorn S. Li L. Marx A. Biernat J. Mandelkow E.M. Mandelkow E. J. Biol. Chem. 2001; 276: 48165-48174Abstract Full Text Full Text PDF PubMed Scopus (454) Google Scholar, 14Goux W.J. Kopplin L. Nguyen A.D. Leak K. Rutkofsky M. Shanmuganandam V.D. Sharma D. Inouye H. Kirschner D.A. J. Biol. Chem. 2004; 279: 26868-26875Abstract Full Text Full Text PDF PubMed Scopus (190) Google Scholar). Because tau is a highly flexible protein, it has resisted all attempts at crystallization for a high-resolution x-ray structure. X-ray analyses exist only in the form of solution scattering (confirming the unfolded nature of the protein, (7Schweers O. Schonbrunn-Hanebeck E. Marx A. Mandelkow E. J. Biol. Chem. 1994; 269: 24290-24297Abstract Full Text PDF PubMed Google Scholar)) and fiber diffraction (confirming the cross-β-structure of PHFs, (13von Bergen M. Barghorn S. Li L. Marx A. Biernat J. Mandelkow E.M. Mandelkow E. J. Biol. Chem. 2001; 276: 48165-48174Abstract Full Text Full Text PDF PubMed Scopus (454) Google Scholar)). We have now applied NMR spectroscopy as an alternative approach to structural studies. Here we report the high-resolution characterization of the repeat domain of tau, in which either all four repeats are present (K18) or repeat two has been deleted (K19), corresponding to the adult and fetal tau isoforms htau40 and htau23 (see Fig. 1). Our findings show that the repeat domain of soluble tau contains regions of residual β-structure that have the potential to serve as seeds for aggregation of tau into PHFs, and we identify residues involved in the interaction with MTs and with polyanions that promote PHF-aggregation. Expression of Recombinant Tau Constructs and Isotope Labeling— Human tau constructs were expressed in the vector pNG2 (a derivative of pET-3a, Merck-Novagen, Darmstadt) in Escherichia coli strain BL21(DE3) as described (Gustke et al. (3Gustke N. Trinczek B. Biernat J. Mandelkow E.M. Mandelkow E. Biochemistry. 1994; 33: 9511-9522Crossref PubMed Scopus (547) Google Scholar)) (see Fig. 1). K18 comprises all four repeats of the largest tau isoform (residues Gln244-Glu372 plus initial Met243), K19 is similar but lacks the second repeat, corresponding to fetal tau (residues Met243 + Gln244-Lys274, Val337-Glu372, without R2 = Val275-Ser305). The expressed proteins were purified from bacterial extracts by making use of the heat stability of the protein and by FPLC SP-Sepharose chromatography (Amersham Biosciences). To label the tau proteins with 15N and 13C stable isotopes, the E. coli culture expressing K18 protein was grown in a M9 minimal medium with 15NH4Cl (1 g liter-1) and 13C glucose (4 g liter-1) (Campro Scientific), and the E. coli culture expressing K19 protein was grown on rich growth medium based on chemolithoautotrophic bacteria labeled with 13C and 15N isotopes (Silantes). Protein samples uniformly enriched in 15N were prepared by growing E. coli bacteria in minimal medium containing 1 g liter-1 of 15NH4Cl. The cell pellets were resuspended in boiling extraction buffer (50 mm MES, 500 mm NaCl, 1 mm MgCl2, 1 mm EGTA, 5 mm dithiothreitol, pH 6.8) complemented with a protease inhibitor mixture. The cells were disrupted with a French pressure cell and subsequently boiled for 20 min. The soluble extract was isolated by centrifugation, the supernatant was dialyzed against two changes of cation exchange chromatography buffer A (20 mm MES, 50 mm NaCl, 1 mm EGTA, 1 mm MgCl2, 2 mm dithiothreitol, 0.1 mm phenylmethylsulfonyl fluoride, pH 6.8) and loaded on an FPLC SP-Sepharose column. The proteins were eluted by a linear gradient of cation exchange chromatography buffer B (20 mm MES, 1 m NaCl, 1 mm EGTA, 1 mm MgCl2, 2 mm dithiothreitol, 0.1 mm phenylmethylsulfonyl fluoride, pH 6.8). NMR samples contained 0.9–1.5 mm 15N- or 15N/13C-labeled protein in 95% H2O, 5% D2O, 50 mm phosphate buffer, pH 6.8, with 1 mm dithiothreitol. Preparation of MTs—Porcine brain tubulin was purified as described (15Mandelkow E.M. Herrmann M. Ruhl U. J. Mol. Biol. 1985; 185: 311-327Crossref PubMed Scopus (169) Google Scholar) and incubated at concentrations higher than 200 μm in MT assembly buffer (100 mm Pipes, pH 6.9, 1 mm EDTA, 1 mm MgSO4, 1 mm dithiothreitol) in the presence of 1 mm GTP at 37 °C for 5 min. After addition of 100 μm Paclitaxel (Sigma-Aldrich) the polymerization was performed for 20 min at 37 °C. Electron Microscopy—Proteins were diluted to a concentration of ∼0.1 mg/ml, placed on 600-mesh carbon-coated copper grids for 1 min, washed with two drops of H2O, and negatively stained with 2% uranyl acetate for 45 s. The specimen were examined in a Philipps 12M electron microscope at 100 kV. NMR Spectroscopy—NMR spectra were acquired at 5 °C on Bruker DRX 800, Avance 600, and DRX 600 NMR spectrometers. Aggregation did not occur under these low temperature conditions. NMR data were processed and analyzed using nmrPipe (16Delaglio F. Grzesiek S. Vuister G.W. Zhu G. Pfeifer J. Bax A. J. Biomol. NMR. 1995; 6: 277-293Crossref PubMed Scopus (11837) Google Scholar) and Sparky 3. Three-dimensional triple-resonance experiments were collected to obtain sequence-specific assignments for the backbone of K18 and K19 (supplemental Table S1). Secondary shift values were calculated as the differences between the measured Cα/C′ chemical shifts and the empirical random coil value for the appropriate amino acid type at pH 3.0 (17Schwarzinger S. Kroon G.J.A. Foss T.R. Chung J. Wright P.E. Dyson H.J. J. Am. Chem. Soc. 2001; 123: 2970-2978Crossref PubMed Scopus (519) Google Scholar). Random coil values for histidines, glutamates, and aspartates were taken from Wishart et al. (18Wishart D.S. Sykes B.D. Methods Enzymol. 1994; 239: 363-392Crossref PubMed Scopus (942) Google Scholar), as the chemical shifts of these residues are particularly sensitive to pH, and the pH in the studies by Wishart et al. (18Wishart D.S. Sykes B.D. Methods Enzymol. 1994; 239: 363-392Crossref PubMed Scopus (942) Google Scholar) (pH 5.0) is more similar to the one used here (pH 6.9). To estimate the β-sheet propensity in contiguous segments of tau, the observed Cα secondary shifts were normalized by the empirically determined secondary shift expected for that residue type in a fully β-sheet conformation (18Wishart D.S. Sykes B.D. Methods Enzymol. 1994; 239: 363-392Crossref PubMed Scopus (942) Google Scholar), summed, and normalized by the number of residues in the segment. Tau-polyanion titrations were carried out with uniformly 15N-labeled protein containing 0.2 mm (heparin titration) and 0.3 mm (poly-Glu titration) K18, and 0.12 mm (heparin titration) and 0.14 mm (poly-Glu titration) K19 at pH 6.8. Heparin (average molecular weight 3000, ∼5.8 disaccharide subunits, charge/subunit ∼2.5, ∼0.31 z/Å) and poly-Glu (average molecular weight 10,000, 7–8 Glu residues, 7–8 negative net charge, in extended conformation 0.33 z/Å) were from Sigma. Complex formation was monitored by recording two-dimensional 1H-15N HSQC spectra for increasing polyanion concentrations (in mm): 0.03, 0.06, 0.15, 0.3, 0.58, 1.37, and 2.53 (K18 with heparin), 0.008, 0.015, 0.03, 0.072, 0.142, 0.28, 0.68, 1.3, 2.4, and 4.3 (K18 with poly-Glu), 0.008, 0.015, 0.03, 0.07, 0.14, 0.28, 0.67, and 1.3 (K19 with heparin), and 0.015, 0.03, 0.074, 0.146, 0.29, 0.70, 1.35, and 2.49 (K19 with poly-Glu). For tau-MT titrations NMR samples contained 0.13 and 0.14 mm uniformly 15N-labeled K18 and K19, respectively. Complex formation was monitored at 5 and 20 °C for MT concentrations (αβ-tubulin dimers) of 26.5, 79.5, and 159.0 μm in the K18 titrations, and 2.8, 5.4, 20.0, 35.0, and 84.3 μm in the K19 titration. Normalized weighted average chemical shift differences for amide 1H and 15N chemical shifts upon polyanion/MT binding were calculated using Δav(NH) = [(ΔH2 + (ΔN/5)2)/2]1/2, where ΔH and ΔN are the differences between the free and bound chemical shifts. Backbone Resonance Assignment of K18 and K19 —NMR resonances in 1H-15N HSQC spectra of the tau repeat domain constructs K18 and K19, uniformly labeled with 15N, were recorded at 5 °C and pH 6.9. The resonances were sharp and showed only a limited dispersion of chemical shifts, reflecting a high degree of backbone mobility and unfolded nature, in agreement with CD and FTIR measurements (7Schweers O. Schonbrunn-Hanebeck E. Marx A. Mandelkow E. J. Biol. Chem. 1994; 269: 24290-24297Abstract Full Text PDF PubMed Google Scholar, 13von Bergen M. Barghorn S. Li L. Marx A. Biernat J. Mandelkow E.M. Mandelkow E. J. Biol. Chem. 2001; 276: 48165-48174Abstract Full Text Full Text PDF PubMed Scopus (454) Google Scholar). To enable a study of the structure and dynamics of tau with atomic resolution, the assignment of the NMR resonances was required. Assignment of globular proteins with a molecular mass below 15 kDa is straightforward today using multidimensional NMR techniques. Backbone assignment of K18 (130 residues) and K19 (99 residues) was complicated, however, by the limited dispersion of NMR resonances because of the absence of well defined secondary and tertiary structure elements. In addition, only five amino acid types make up more than 50% of the primary sequence, and these amino acids are arranged in sequence motifs that occur repetitively, further increasing NMR signal overlap. Therefore, an unambiguous assignment was not possible just relying on the conventional three-dimensional spectra HNCO, HN(CA)CO, HNCACB and CBCA(CO)NH (19Bax A. Grzesiek S. Acc. Chem. Res. 1993; 26: 131-138Crossref Scopus (800) Google Scholar). A high-resolution (HA)CANNH (20Zweckstetter M. Bax A. J. Am. Chem. Soc. 2001; 123: 9490-9491Crossref PubMed Scopus (63) Google Scholar) together with three-dimensional HNN and HN(C)N experiments were required (21Panchal S.C. Bhavesh N.S. Hosur R.V. J. Biomol. NMR. 2001; 20: 135-147Crossref PubMed Scopus (176) Google Scholar). With these NMR spectra the backbone resonance assignment of K18 and K19 could be achieved in an iterative fashion based on automatic assignment using the program MARS (22Jung Y.S. Zweckstetter M. J. Biomol. NMR. 2004; 30: 11-23Crossref PubMed Scopus (241) Google Scholar) and manual assignment mainly relying on the HNN experiment. The sequence motif PGGG, which is located characteristically at the C-terminal end of each of the four repeats, caused the biggest problems. Because of severe chemical shift overlap, it was only possible to assign Gly273, Gly304, and Gly335. The resonances of the remaining glycines within these motifs could be identified in the spectra but not unambiguously assigned to one of the repeats. Otherwise, all 1H, 15N, 13Cα, 13Cβ, and 13C′ chemical shifts were assigned in both K18 and K19. Soluble Tau Contains Residual β-Structure—NMR chemical shifts, in particular of Cα and C′ atoms, are very sensitive probes of secondary structure both in globular and unfolded proteins (Ref. 23Dyson H.J. Wright P.E. Nat. Struct. Biol. 1998; 5: 499-503Crossref PubMed Scopus (214) Google Scholar and Wishart et al. (18Wishart D.S. Sykes B.D. Methods Enzymol. 1994; 239: 363-392Crossref PubMed Scopus (942) Google Scholar)). These shifts show small but distinct deviations from random coil values for both K18 and K19 (Fig. 2). In the case of K18, continuous stretches (containing more than six amino acids) of negative Cα secondary chemical shifts were found for residues Lys274-Leu284, Ser305-Leu315, and Gln336-Asp345 (numbering according to the longest isoform htau40), indicative of nascent β-structure (Fig. 2A). These stretches lie at the beginning of repeats R2, R3, and R4, downstream from PGGG motifs and encompass the hexapeptide motifs VQIINK and VQIVYK known to be important for the abnormal aggregation of tau into PHFs (8von Bergen M. Friedhoff P. Biernat J. Heberle J. Mandelkow E.M. Mandelkow E. Mol. Biol. Cell. 2000; 11: 363AGoogle Scholar). In case of C′, negative secondary chemical shifts were detected for a similar range of residues, Gly273-Leu282, Gly304-Asp314, and Gly335-Asp345 (supplemental Fig. S1). It is notable that the preponderance of negative secondary chemical shifts persists even at the beginning of the less homologous repeats R1 (residues Ala246-Leu253) and R5 (Asn368-Ile371). In the case of R1 this is likely not because of a propensity to form β-structure but rather because of the occurrence of three prolines (Pro247, Pro249, Pro251). Indeed, the C′ secondary chemical shifts alternated in this region consistent with the fact that Cα atoms of residues preceding prolines show unusual chemical shifts due to the absence of an amide proton in the proline. In the case of R5 the results must be viewed with caution because of the presence of only three residues of R5 (not counting the C-terminal residue). A remarkably similar pattern is observed for K19 even though repeat R2 (exon 10) is absent (Fig. 2C), suggesting that the repeats represent independent structural units. A quantitative analysis of the averaged Cα and C′ secondary chemical shifts (Fig. 2B) indicates that β-structure-like conformations are populated ∼16, 24, and 13% of the time for residues Lys274-Asp283, Ser305-Asp314, and Gln336-Asp345, respectively. Besides the regions in the beginning of each repeat, residues 295KDNIK299 and 357LDNIT361 in repeats R2 and R4 show a stretch of negative Cα/C′ secondary shifts, indicating a β-structure propensity. Note that no β-structure propensity is present in the corresponding region of repeat 3 (325LGNIH329), possibly because of the presence of a glycine (Fig. 2B). The regions of high propensity for the β-structure are limited upstream by PGGG motifs, where the three glycines show positive Cα and C′ secondary chemical shifts in all four repeats. This indicates either a helical propensity or a turn, in agreement with the fact that GGK and GGS sequences (present in R1 or R2) are the most common motifs in classic γ-turns. Alternatively, the PGGG motif is also common in type II β-turns. The downstream ends of the regions of high β-structure propensity are most prominently identified by positive Cα/C′ secondary chemical shifts of the 10th and/or 11th residue in repeat 2–4 (Ser285, Ser316, Phe346) (Fig. 2B). Taking further into account that sequence dyads such as 285SN286 and 316SK317 (the motifs following high β-propensity regions in R2 and R3) are common in type I β-turns, the Cα and C′ secondary chemical shifts suggest a β-turn at the C-terminal end of the regions of high β-structure propensity. In conclusion, 8–10 residues at the beginning of repeats 2–4 show a residual β-structure sandwiched between regions with a high turn propensity. The most pronounced β-structure propensity is present at the N terminus of the third repeat, whereas in the second and fourth repeat it is approximately a factor of 1.5–2 lower. These β-conformations in K18 and K19 are presumably in rapid exchange with random like structures. The Cα secondary chemical pattern observed for K19 is almost identical to that in K18. This indicates that the removal of the second repeat does not affect the local β-structure propensity in the remaining part of the repeat domain of tau. Characterization of K18 and K19 and Its Polyanion Complexes—NMR signals of backbone amides constitute excellent probes of complex formation, providing maps of the interaction interfaces and binding constants (24Craik D.J. Wilce J.A. Methods Mol. Biol. 1997; 60: 195-232PubMed Google Scholar). The binding of the polyanions heparin and poly-Glu to K18 and K19 was monitored by 1H-15N HSQC spectra. The size of the chemical shift changes for increasing polyanion concentrations depends on the binding strength to the corresponding residues. At high polyanion concentrations chemical shift changes of some residues were accompanied by a decrease of the intensity of their signals, indicating chemical shift exchange intermediate on the NMR time scale (Fig. 3). In the case of heparin the residues exhibiting the largest displacements in the amide resonances of K18 were Val275-Leu284 at the beginning of the second repeat (Fig. 4A). In addition, strong chemical shift changes were observed in the environment of selected lysine and histidine residues: Leu253-Lys254, His268-Gln269, Lys298-Val300, His329-Lys331, and His362. The importance of lysine and histidine residues suggests that the binding to polyanions is electrostatically driven. Upon titration of K18 with poly-Glu, a similar pattern of chemical shift changes was observed (Fig. 4B), indicating that the overall binding behavior for the two types of polyanions is similar. In K19, repeat 2 is missing and can therefore not contribute to the interaction with polyanions. However, the same lysine and histidine residues as in K18 were strongly affected. In addition, slightly more pronounced chemical shift changes were observed for Val306-Lys311, the six residues in the third repeat that are important for tau aggregation to PHFs (Fig. 4, C and D). Concomitant with the chemical shift changes and the disappearance of some resonances at high polyanion concentrations (caused by ligand binding), the overall signal intensity in the 1H-15N HSQC spectra decreased with increasing polyanion concentration (after correction for dilution effects). The latter effect is likely due to partial aggregation, because biochemical analysis indicated that at the end of the polyanion titration 50% of K18 was polymerized, as analyzed by pelleting and SDS-PAGE. A similar behavior was observed in the K19-poly-Glu titration. To investigate the effect of aggregation further, we performed a series of 1H-15N HSQC spectra, in which the ratio of concentration between heparin and K18 (1:4) was held constant, and the sample was kept for variable times at 50 °C to induce and accelerate aggregation. After measurement of initial 1H-15N HSQCs without heparin and with a heparin:K18 ratio of 1:4 at 5 °C, the sample was heated to 50 °C and kept at this temperature for 2 h. After the 2 h the sample was cooled down to 5 °C, and an HSQC spectrum was recorded. This procedure was repeated five times, such that the protein was kept at 50 °C for 10 h in total. The addition of heparin induced the chemical shift changes that were reported in Fig. 4. The peak intensities for most residues were reduced by less than 15% upon addition of heparin with the exception of those residues showing very strong chemical shift changes. After the first 2 h at 50 °C the intensities of almost all peaks were reduced to 72% compared with the starting spectrum without heparin. There was no area in the protein that was affected particularly strongly, rather an overall loss of signal intensity was observed. At the same time, the chemical shift changes that were initially induced by addition of heparin were largely lost, i.e. the spectrum after 2 h at 50 °C with a heparin:K18 ratio of 1:4 was highly similar to the one of K18 before addition of heparin (shown in Fig. 3). After 4, 6, 8, and 10 h at 50 °C the overall signal intensity in 1H-15N HSQC spectra continuously decreased to 68, 65, 64, and 61% compared with the starting spectrum without heparin, respectively. No large chemical shift differences relative to the spectrum after 2 h at 50 °C were observed. In addition, the average transverse 15N relaxation time did not change compared with free K18 upon addition of heparin or after keeping the protein up to 10 h at 50 °C (data not shown). At the end of the temperature titration biochemical analysis indicated that ∼50% of K18 was polymerized. These data indicate that the observed chemical shift changes were because of the binding of the polyanion to the protein and not because of aggregation. When the protein was exposed to high levels of heparin for prolonged times or to high temperature at a constant heparin:K18 ratio of 1:4, part of the protein aggregated. As the chemical shift changes, which were induced by addition of heparin, were largely removed upon aggregation, heparin may become bound to the aggregates, thereby reducing the effective concentration of heparin in solution. The Binding Mode of MTs and Polyanions to Tau Is Similar—The binding of K18 and K19 to MTs was characterized using the NMR chemical shift perturbation method (24Craik D.J. Wilce J.A. Methods Mol. Biol. 1997; 60: 195-232PubMed Google Scholar) in which two-dimensional 1H-15N HSQC spectra of K18 and K19 were recorded in the presence of increasing amounts of taxol-stabilized MTs. The influence of temperature was probed by performing the NMR titrations at 5 and 20 °C. To ascertain the stability of the taxol-stabilized MTs at these temperatures, aliquots from the reactions of tau with MTs were characterized by sedimentation, SDS-PAGE, and electron microscopy (Fig. 5). After assembly at 37 °C the MTs exhibit the typical morphological features of protofilaments spaced ∼5 nm apart. During the NMR analysis the sample is incubated for more than one h at 5 °C, but this does not lead to the disintegration of MTs in the presence of taxol (Fig. 5, A and B). In the presence of tau the surface features of MTs become somewhat fuzzier but without a change in subunit arrangement (see Santarella et al., 2004 (39Santarella R.A. Skiniotis G. Goldie K.N. Tittmann P. Gross H. Mandelkow E.M. Mandelkow E. Hoenger A. J. Mol. Biol. 2004; 339: 539-553Crossref PubMed Scopus (124) Google Scholar)) (Fig. 5C). For a quantitative analysis the samples were centrifuged, and the supernatants and pellets were checked by SDS-PAGE (Fig. 5, D and E). After MT assembly with taxol and incubation at 20 or 5 °C tubulin is found mostly in the pellet fraction (Fig. 5D). When increasing amounts of MTs were added to K18, the binding resulted in increasing amounts of K18 in the pellet fraction (Fig. 5D). The same was true for K19. These experiments show that the taxol-stabilized MTs persist over the time course of the NMR measurement and that binding parameters of both K19 and K18 are in good agreement with earlier results (3Gustke N. Trinczek B. Biernat J. Mandelkow E.M. Mandelkow E. Biochemistry. 1994; 33: 9511-9522Crossref PubMed Scopus (547) Google Scholar). Continuous chemical shift changes of selected residues followed by the disappearance of the most strongly shifting resonances, especially at 20 °C, were observed. In addition, the signal:noise ratio rapidly decreased at high MT concentrations, indicative of an increasing amount of K18 or K19 bound to MTs. When bound to MTs, the NMR signals of K18 or K19 are not observable because of very fast relaxation. At 5 °C the most strongly shifting resonances in K18 were found mainly in the environment of lysine and histidine residues: Leu253-Lys254, His268-Gln269, Lys298-Val300, His329-Lys331, and His362 (Fig. 6A). This indicates that each repeat contains a binding site for anchoring K18 to MTs. In agreement with the repetitive nature of the repeat region, these anchors are primarily located at equivalent positions in each repeat, just N-terminal to the PGGG motif. The same lysine and histidines resonances were affected at 20 °C, although to a lesser extent. At 20 °C the most pronounced chemical shift changes were observed for r
Referência(s)