Interactions between molecules are fundamental in biology. They occur also between amyloidogenic ... more Interactions between molecules are fundamental in biology. They occur also between amyloidogenic peptides or proteins that are associated with different amyloid diseases, which makes it important to study the mutual influence of two polypeptides on each other's properties in mixed samples. However, addressing this research question with imaging techniques faces the challenge to distinguish different polypeptides without adding artificial probes for detection. Here, we show that nanoscale infrared spectroscopy in combination with 13 C, 15 N-labeling solves this problem. We studied aggregated amyloid-β peptide (Aβ) and its interaction with an inhibitory peptide (NCAM1-PrP) using scattering-type scanning near-field optical microscopy. Although having similar secondary structure, labeled and unlabeled peptides could be distinguished by comparing optical phase images taken at wavenumbers characteristic for either the labeled or the unlabeled peptide. NCAM1-PrP seems to be able to associate with or to dissolve existing Aβ fibrils because pure Aβ fibrils were not detected after mixing.
A new pulse sequence for the direct measurement of heteroof relaxation rates should be measured. ... more A new pulse sequence for the direct measurement of heteroof relaxation rates should be measured. One of the most nuclear cross-relaxation rates is presented. The pulse sequence frequently measured rates is the heteronuclear cross-relaxuses proton detection of transient carbon magnetization with sensiation rate s between an a carbon and an a proton or between tivity-enhanced transfer of magnetization and pulsed field gradian amide nitrogen and an amide proton. This rate is sensitive ents for coherence selection and water suppression. The heteroto the difference of the spectral density at the frequencies nuclear cross-relaxation rate is measured using either nonselective v H / v X and v H 0 v X and probes the dynamics on the or selective inversion of all or particular protons. A technique for nanosecond time scale. The cross-relaxation rate is usually the suppression of spin diffusion is also applied. The pulse secalculated indirectly from the measured heteronuclear quence is tested on a peptide with a selectively 13 C-labeled asteady-state NOE and the longitudinal-relaxation rate of the carbon. The directly measured rate is compared with and agrees heteronucleus. Here we present a direct and selective method well with the cross-relaxation rates as traditionally calculated from the steady-state NOE and carbon longitudinal-relaxation rate. for measuring the cross-relaxation rate between a hetero-᭧ 1997 Academic Press nucleus and any of its neighboring protons. The basic idea of this experiment is the same as that described by Solomon (4). The sensitivity of the pulse sequence described here is increased by proton detection of the heteronucleus magneti
Alzheimer's disease involves progressive neuronal loss. Linked to the disease is the amyloid β (A... more Alzheimer's disease involves progressive neuronal loss. Linked to the disease is the amyloid β (Aβ) peptide, a 38−43-amino acid peptide found in extracellular amyloid plaques in the brain. Cyclodextrins are nontoxic, cone-shaped oligosaccharides with a hydrophilic exterior and a hydrophobic cavity making them suitable hosts for aromatic guest molecules in water. β-Cyclodextrin consists of seven α-D-glucopyranoside units and has been shown to reduce the level of fibrillation and neurotoxicity of Aβ. We have studied the interaction between Aβ and a β-cyclodextrin dimer, consisting of two β-cyclodextrin monomers connected by a flexible linker. The β-cyclodextrin monomer has been found to interact with Aβ(1−40) at sites Y10, F19, and/or F20 with a dissociation constant (K D) of 3.9 ± 2.0 mM. Here 1 H− 15 N and 1 H− 13 C heteronuclear single-quantum correlation nuclear magnetic resonance (NMR) spectra show that in addition, the β-cyclodextrin monomer and dimer bind to the histidines. NMR translational diffusion experiments reveal the increased affinity of the β-cyclodextrin dimer (apparent K D of 1.1 ± 0.5 mM) for Aβ(1−40) compared to that of the β-cyclodextrin monomer. Kinetic aggregation experiments based on thioflavin T fluorescence indicate that the dimer at 0.05−5 mM decreases the lag time of Aβ aggregation, while a concentration of 10 mM increases the lag time. The β-cyclodextrin monomer at a high concentration decreases the lag time of the aggregation. We conclude that cyclodextrin monomers and dimers have specific, modulating effects on the Aβ(1−40) aggregation process. Transmission electron microscopy shows that the regular fibrillar aggregates formed by Aβ(1−40) alone are replaced by a major fraction of amorphous aggregates in the presence of the β-cyclodextrin dimer.
Self-assembly of amyloid β (Aβ) peptide molecules into large aggregates is a naturally occurring ... more Self-assembly of amyloid β (Aβ) peptide molecules into large aggregates is a naturally occurring process driven in aqueous solution by a dynamic interplay between hydrophobic interactions among Aβ molecules, which promote aggregation, and steric and overall electrostatic hindrance, which stifles it. Aβ self-association is entropically unfavorable, as it implies order increase in the system, but under favorable kinetic conditions, the process proceeds at appreciable rates, yielding Aβ aggregates of different sizes and structures. Despite the great relevance and extensive research efforts, detailed kinetic mechanisms underlying Aβ aggregation remain only partially understood. In this study, fluorescence correlation spectroscopy (FCS) and Thioflavin T (ThT) were used to monitor the time dependent growth of structured aggregates and characterize multiple components during the aggregation of Aβ peptides in a heterogeneous aqueous solution. To this aim, we collected data during a relatively large number of observation periods, 30 consecutive measurements lasting 10 s each, at what we consider to be a constant time point in the slow aggregation process. This approach enabled monitoring the formation of nanomolar concentrations of structured amyloid aggregates and demonstrated the changing distribution of amyloid aggregate sizes throughout the aggregation process. We identified aggregates of different sizes with molecular weight from 260 to more than 1 × 10(6) kDa and revealed the hitherto unobserved kinetic turnover of intermediates during Aβ aggregation. The effect of different Aβ concentrations, Aβ:ThT ratios, differences between the 40 (Aβ40) and 42 (Aβ42) residue long variants of Aβ, and the effect of stirring were also examined.
For NMR probes equipped with pulsed field gradient coils, which are not optimized for gradient li... more For NMR probes equipped with pulsed field gradient coils, which are not optimized for gradient linearity, the precision and accuracy of experimentally measured translational diffusion coefficients are limited by the linearity of the gradient pulses over the sample volume. This study shows that the accuracy and precision of measured diffusion coefficients by the Stejskal-Tanner spin-echo pulsed field gradient experiment can be significantly improved by mapping the gradient z-profile and by using the mapped calibration parameters in the data analysis. For practical applications the gradient distribution may be approximated by a truncated linear distribution defined by minimum and maximum values of the gradient. By including the truncated linear gradient distribution function in the Stejskal-Tanner equation, the systematic deviation between the fitted curve and the experimental attenuation curve decreases by an order of magnitude. The gradient distribution may be calibrated using an intense NMR signal from a sample with a known diffusion coefficient. The diffusion coefficient of an unknown sample may then be determined from a two-parameter fit, using the known gradient distribution function.
One of the grand challenges of biophysical chemistry is to understand the principles that govern ... more One of the grand challenges of biophysical chemistry is to understand the principles that govern protein aggregation leading to amyloid fibrils, which is a highly complex and sensitive process. This review provides a comprehensive overview of how amyloid aggregation is affected by the variousin vivoconstituents and conditions.
NMR spectroscopy combined with paramagnetic relaxation agents was used to study the positioning o... more NMR spectroscopy combined with paramagnetic relaxation agents was used to study the positioning of the 40-residue Alzheimer Amyloid b-peptide Ab(1-40) in SDS micelles. 5-Doxyl stearic acid incorporated into the micelle or Mn 2+ ions in the aqueous solvent were used to determine the position of the peptide relative to the micelle geometry. In SDS solvent, the two a-helices induced in Ab(1-40), comprising residues 15-24, and 29-35, respectively, are surrounded by flexible unstructured regions. NMR signals from these unstructured regions are strongly attenuated in the presence of Mn 2+ showing that these regions are positioned mostly outside the micelle. The central helix (residues 15-24) is significantly affected by 5-doxyl stearic acid however somewhat less for residues 16, 20, 22 and 23. This a-helix therefore resides in the SDS headgroup region with the face with residues 16, 20, 22 and 23 directed away from the hydrophobic interior of the micelle. The C-terminal helix is protected both from 5-doxyl stearic acid and Mn 2+ , and should be buried in the hydrophobic interior of the micelle. The SDS micelles were characterized by diffusion and 15 N-relaxation measurements. Comparison of experimentally determined translational diffusion coefficients for SDS and Ab(1-40) show that the size of SDS micelle is not significantly changed by interaction with Ab(1-40).
The downstream prion-like Doppel (Dpl) protein is a homologue related to the prion protein (PrP).... more The downstream prion-like Doppel (Dpl) protein is a homologue related to the prion protein (PrP). Dpl is expressed in the brains of mice that do not express PrP, and Dpl is known to be toxic to neurons. One mode of toxicity has been suggested to involve direct membrane interactions. PrP under certain conditions of cell trafficking retains an uncleaved signal peptide, which may also hold for the much less studied Dpl. For a peptide with a sequence derived from the N-terminal part (1-30) of mouse Dpl (mDpl(1-30)) CD spectroscopy shows about 40% R-helical structure in DHPC and SDS micelles. In aqueous solution it is mostly a random coil. The three-dimensional solution structure was determined by NMR for mDpl(1-30) associated with DHPC micelles. 2D 1 H NMR spectra of the peptide in q) 0.25 DMPC/DHPC bicelles only showed signals from the unstructured termini, indicating that the structured part of the peptide resides within the lipid bilayer. Together with 2 H 2 O exchange data in the DHPC micelle solvent, these results show an R-helix protected from solvent exchange between residues 7 and 19, and suggest that the R-helical segment can adopt a transmembrane localization also in a membrane. Leakage studies with entrapped calcein in large unilamellar phospholipid vesicles showed that the peptide is almost as membrane perturbing as melittin, known to form pores in membranes. The results suggest a possible channel formation mechanism for the unprocessed Dpl protein, which may be related to toxicity through direct cell membrane interaction and damage.
Non-invasive biological information about residual neuroblastoma tumour tissue could allow treatm... more Non-invasive biological information about residual neuroblastoma tumour tissue could allow treatment monitoring without the need for repeated biopsies. Magnetic resonance spectroscopy (MRS) can be performed with standard MR-scanners, providing specific biochemical information from selected tumour regions. By proton 1H-MRS, lipids, certain amino acids and lactate can be detected and their relative concentrations estimated in vivo. Using experimental models of neuroblastoma, we have described the potential of 1H-MRS for the prediction of tumour tissue viability and treatment response. Whereas viable neuroblastoma tissue is dominated by the choline 1H-MRS resonance, cell death as a consequence of spontaneous necrosis or successful treatment with chemotherapy, angiogenesis inhibitors, or NSAIDs is associated with decreased choline content. Therapy-induced neuroblastoma cell death is also associated with enhanced 1H-MRS resonances from mobile lipids and polyunsaturated fatty acids. The mobile lipid/choline ratio correlates significantly with cell death and based on the dynamics of this ratio tumour regression or continued growth (drug resistance) after chemotherapy can be predicted in vivo. The implications of these findings are discussed with focus on the potentials and limitations of introducing 1H-MRS for clinical assessment of treatment response in children with neuroblastoma. Biochemical monitoring of neuroblastoma with 1H-MRS could enable tailoring of individual therapy as well as provide early pharmacodynamic evaluation of novel therapeutic modalities.
A new early-onset form of Alzheimer's disease (AD) was described recently where a point ... more A new early-onset form of Alzheimer's disease (AD) was described recently where a point mutation was discovered in codon 693 of the β-amyloid (Aβ) precursor protein gene, the Arctic mutation. The mutation translates into a single amino acid substitution, glutamic acid→glycine, in position 22 of the Aβ peptide. The mutation carriers have lower plasma levels of Aβ than normal, while in vitro studies show that Aβ1–40E22G protofibril formation is significantly enhanced. We have explored the nature of the Aβ1–40E22G peptide in more detail, in particular the protofibrils. Using size-exclusion chromatography (SEC) and circular dichroism spectroscopy (CD) kinetic and secondary structural characteristics were compared with other Aβ1–40 peptides and the Aβ12–28 fragment, all having single amino acid substitutions in position 22. We have found that Aβ1–40E22G protofibrils are a group of comparatively stabile β-sheet-containing oligomers with a heterogeneous size distribution, ranging from >100 kDa to >3000 kDa. Small Aβ1–40E22G protofibrils are generated about 400 times faster than large ones. Salt promotes their formation, which significantly exceeds all the other peptides studied here, including the Dutch mutation Aβ1–40E22Q. Position 22 substitutions had significant effects on aggregation kinetics of Aβ1–40 and in Aβ12–28, although the qualitative aspects of the effects differed between the native peptide and the fragment, as no protofibrils were formed by the fragments. The rank order of protofibril formation of Aβ1–40 and its variants was the same as the rank order of the length of the nucleation/lag phase of the Aβ12–28 fragments, E22V>E22A≫E22G>E22Q≫E22, and correlated with the degree of hydrophobicity of the position 22 substituent. The molecular mass of peptide monomers and protofibrils were estimated better in SEC studies using linear rather than globular calibration standards. The characteristics of the Aβ1–40E22G suggest an important role for the peptide in the neuropathogenesis in the Arctic form of AD.
The dynamics of monomeric Alzheimer Ab(1-40) in aqueous solution was studied using heteronuclear ... more The dynamics of monomeric Alzheimer Ab(1-40) in aqueous solution was studied using heteronuclear NMR experiments. 15 N NMR relaxation rates of amide groups report on the dynamics in the peptide chain and make it possible to estimate structural propensities from temperature-dependent relaxation data and chemical shifts change analysis. The persistence length of the polypeptide chain was determined using a model in which the influence of neighboring residue relaxation is assumed to decay exponentially as a function of distance. The persistence length of the Ab(1-40) monomer was found to decrease from eight to three residues when temperature was increased from 3 to 18°C. At 3°C the peptide shows structural propensities that correlate well with the suggested secondary structure regions of the peptide to be present in the fibrils, and with the a-helical structure in membrane-mimicking systems. Our data leads to a structural model for the monomeric soluble b-peptide with six different regions of secondary structure propensities. The peptide has two regions with b-strand propensity (residues 16-24 and 31-40), two regions with high PII-helix propensity (residues 1-4 and 11-15) and two unstructured regions with higher mobility (residues 5-10 and 25-30) connecting the structural elements.
Journal of Magnetic Resonance, Series B, Apr 1, 1996
The spectral-density mapping of a 13 C a-1 H a vector of Leu 10 in mension to the relaxation data... more The spectral-density mapping of a 13 C a-1 H a vector of Leu 10 in mension to the relaxation data, although the information the 22-residue peptide hormone motilin [P. Allard, J. Jarvet, A. content of the six relaxation rates in the time domain is Ehrenberg, and A. Grä slund, J. Biomol. NMR 5, 133-146 (1995)] the same as for the spectral-density values in the frequency is extended in this paper to three polarizing fields 9.4, 11.7, and domain. The correspondence between the two representa-14.1 T in order to improve the accuracy of the calculated spectraltions is somewhat analogous to the relation between the freedensity function J(v) and to extend the sampling range up to induction decay and the spectrum in NMR spectroscopy. 750 MHz. The problem with a usually large relative error in J(v H) When experiments are done at different magnetic fields, one is eliminated since the generally more precise J(v H 0 v C) and can calculate the slope of the spectral-density function at J(v H / v C) determined at other fields appear at nearly the same different sampling frequencies. The slope of the spectralfrequencies. The fitting of dynamic models to the points of spectral density was made with error weighting, and the influence of J(v H) density function at different frequencies is very useful when was found to be negligible. Therefore, the high-frequency part comparison is made between experimental data and dynamiof the spectral-density function is determined essentially without cal models. influence from the two transverse-type relaxation rates. In the case With the spectral-density mapping technique, the meaof a carbon-proton vector, the relaxation is mainly determined surement of the spectral-density function is separated from by dipolar interaction and is only weakly influenced by other relaxthe fitting of dynamic models, which is not true, e.g., when ation mechanisms, which makes it particularly suitable for the using the model-free approach of Lipari and Szabo (2, 3). spectral-density mapping technique. The measured relaxation To obtain spectral-density values, which can discriminate rates in the time domain are transformed into the frequency dobetween different motional models, the precision of experimain by spectral-density mapping, and the slopes in different frequency regions are important parameters when comparing experimental relaxation rates must be high. Less-accurate data are mental data with theoretical models of motion. Using an adjustable generally used as input for the model-free analysis (2, 3) internuclear distance r eff , combined with the model-free approach, which still leads to reasonable values for the overall rotait is possible to obtain a reasonable fit to measured spectral-density tional correlation time t m and for the general order parameter points at J(0) and around J(v C). At the same time, however, the S. For peptides as well as for proteins, the correlation time high-frequency slope of the spectral-density function defined by for overall rotational motion is typically on the order of J(v H 0 v C) and J(v H / v C) could not be reproduced. ᭧ 1996 nanoseconds, and, as a consequence, the numerical values Academic Press, Inc. of different spectral-density points differ in size by at least an order of magnitude at the magnetic fields used here. When the model-free approach was introduced (2, 3), the
Interactions between molecules are fundamental in biology. They occur also between amyloidogenic ... more Interactions between molecules are fundamental in biology. They occur also between amyloidogenic peptides or proteins that are associated with different amyloid diseases, which makes it important to study the mutual influence of two polypeptides on each other's properties in mixed samples. However, addressing this research question with imaging techniques faces the challenge to distinguish different polypeptides without adding artificial probes for detection. Here, we show that nanoscale infrared spectroscopy in combination with 13 C, 15 N-labeling solves this problem. We studied aggregated amyloid-β peptide (Aβ) and its interaction with an inhibitory peptide (NCAM1-PrP) using scattering-type scanning near-field optical microscopy. Although having similar secondary structure, labeled and unlabeled peptides could be distinguished by comparing optical phase images taken at wavenumbers characteristic for either the labeled or the unlabeled peptide. NCAM1-PrP seems to be able to associate with or to dissolve existing Aβ fibrils because pure Aβ fibrils were not detected after mixing.
A new pulse sequence for the direct measurement of heteroof relaxation rates should be measured. ... more A new pulse sequence for the direct measurement of heteroof relaxation rates should be measured. One of the most nuclear cross-relaxation rates is presented. The pulse sequence frequently measured rates is the heteronuclear cross-relaxuses proton detection of transient carbon magnetization with sensiation rate s between an a carbon and an a proton or between tivity-enhanced transfer of magnetization and pulsed field gradian amide nitrogen and an amide proton. This rate is sensitive ents for coherence selection and water suppression. The heteroto the difference of the spectral density at the frequencies nuclear cross-relaxation rate is measured using either nonselective v H / v X and v H 0 v X and probes the dynamics on the or selective inversion of all or particular protons. A technique for nanosecond time scale. The cross-relaxation rate is usually the suppression of spin diffusion is also applied. The pulse secalculated indirectly from the measured heteronuclear quence is tested on a peptide with a selectively 13 C-labeled asteady-state NOE and the longitudinal-relaxation rate of the carbon. The directly measured rate is compared with and agrees heteronucleus. Here we present a direct and selective method well with the cross-relaxation rates as traditionally calculated from the steady-state NOE and carbon longitudinal-relaxation rate. for measuring the cross-relaxation rate between a hetero-᭧ 1997 Academic Press nucleus and any of its neighboring protons. The basic idea of this experiment is the same as that described by Solomon (4). The sensitivity of the pulse sequence described here is increased by proton detection of the heteronucleus magneti
Alzheimer's disease involves progressive neuronal loss. Linked to the disease is the amyloid β (A... more Alzheimer's disease involves progressive neuronal loss. Linked to the disease is the amyloid β (Aβ) peptide, a 38−43-amino acid peptide found in extracellular amyloid plaques in the brain. Cyclodextrins are nontoxic, cone-shaped oligosaccharides with a hydrophilic exterior and a hydrophobic cavity making them suitable hosts for aromatic guest molecules in water. β-Cyclodextrin consists of seven α-D-glucopyranoside units and has been shown to reduce the level of fibrillation and neurotoxicity of Aβ. We have studied the interaction between Aβ and a β-cyclodextrin dimer, consisting of two β-cyclodextrin monomers connected by a flexible linker. The β-cyclodextrin monomer has been found to interact with Aβ(1−40) at sites Y10, F19, and/or F20 with a dissociation constant (K D) of 3.9 ± 2.0 mM. Here 1 H− 15 N and 1 H− 13 C heteronuclear single-quantum correlation nuclear magnetic resonance (NMR) spectra show that in addition, the β-cyclodextrin monomer and dimer bind to the histidines. NMR translational diffusion experiments reveal the increased affinity of the β-cyclodextrin dimer (apparent K D of 1.1 ± 0.5 mM) for Aβ(1−40) compared to that of the β-cyclodextrin monomer. Kinetic aggregation experiments based on thioflavin T fluorescence indicate that the dimer at 0.05−5 mM decreases the lag time of Aβ aggregation, while a concentration of 10 mM increases the lag time. The β-cyclodextrin monomer at a high concentration decreases the lag time of the aggregation. We conclude that cyclodextrin monomers and dimers have specific, modulating effects on the Aβ(1−40) aggregation process. Transmission electron microscopy shows that the regular fibrillar aggregates formed by Aβ(1−40) alone are replaced by a major fraction of amorphous aggregates in the presence of the β-cyclodextrin dimer.
Self-assembly of amyloid β (Aβ) peptide molecules into large aggregates is a naturally occurring ... more Self-assembly of amyloid β (Aβ) peptide molecules into large aggregates is a naturally occurring process driven in aqueous solution by a dynamic interplay between hydrophobic interactions among Aβ molecules, which promote aggregation, and steric and overall electrostatic hindrance, which stifles it. Aβ self-association is entropically unfavorable, as it implies order increase in the system, but under favorable kinetic conditions, the process proceeds at appreciable rates, yielding Aβ aggregates of different sizes and structures. Despite the great relevance and extensive research efforts, detailed kinetic mechanisms underlying Aβ aggregation remain only partially understood. In this study, fluorescence correlation spectroscopy (FCS) and Thioflavin T (ThT) were used to monitor the time dependent growth of structured aggregates and characterize multiple components during the aggregation of Aβ peptides in a heterogeneous aqueous solution. To this aim, we collected data during a relatively large number of observation periods, 30 consecutive measurements lasting 10 s each, at what we consider to be a constant time point in the slow aggregation process. This approach enabled monitoring the formation of nanomolar concentrations of structured amyloid aggregates and demonstrated the changing distribution of amyloid aggregate sizes throughout the aggregation process. We identified aggregates of different sizes with molecular weight from 260 to more than 1 × 10(6) kDa and revealed the hitherto unobserved kinetic turnover of intermediates during Aβ aggregation. The effect of different Aβ concentrations, Aβ:ThT ratios, differences between the 40 (Aβ40) and 42 (Aβ42) residue long variants of Aβ, and the effect of stirring were also examined.
For NMR probes equipped with pulsed field gradient coils, which are not optimized for gradient li... more For NMR probes equipped with pulsed field gradient coils, which are not optimized for gradient linearity, the precision and accuracy of experimentally measured translational diffusion coefficients are limited by the linearity of the gradient pulses over the sample volume. This study shows that the accuracy and precision of measured diffusion coefficients by the Stejskal-Tanner spin-echo pulsed field gradient experiment can be significantly improved by mapping the gradient z-profile and by using the mapped calibration parameters in the data analysis. For practical applications the gradient distribution may be approximated by a truncated linear distribution defined by minimum and maximum values of the gradient. By including the truncated linear gradient distribution function in the Stejskal-Tanner equation, the systematic deviation between the fitted curve and the experimental attenuation curve decreases by an order of magnitude. The gradient distribution may be calibrated using an intense NMR signal from a sample with a known diffusion coefficient. The diffusion coefficient of an unknown sample may then be determined from a two-parameter fit, using the known gradient distribution function.
One of the grand challenges of biophysical chemistry is to understand the principles that govern ... more One of the grand challenges of biophysical chemistry is to understand the principles that govern protein aggregation leading to amyloid fibrils, which is a highly complex and sensitive process. This review provides a comprehensive overview of how amyloid aggregation is affected by the variousin vivoconstituents and conditions.
NMR spectroscopy combined with paramagnetic relaxation agents was used to study the positioning o... more NMR spectroscopy combined with paramagnetic relaxation agents was used to study the positioning of the 40-residue Alzheimer Amyloid b-peptide Ab(1-40) in SDS micelles. 5-Doxyl stearic acid incorporated into the micelle or Mn 2+ ions in the aqueous solvent were used to determine the position of the peptide relative to the micelle geometry. In SDS solvent, the two a-helices induced in Ab(1-40), comprising residues 15-24, and 29-35, respectively, are surrounded by flexible unstructured regions. NMR signals from these unstructured regions are strongly attenuated in the presence of Mn 2+ showing that these regions are positioned mostly outside the micelle. The central helix (residues 15-24) is significantly affected by 5-doxyl stearic acid however somewhat less for residues 16, 20, 22 and 23. This a-helix therefore resides in the SDS headgroup region with the face with residues 16, 20, 22 and 23 directed away from the hydrophobic interior of the micelle. The C-terminal helix is protected both from 5-doxyl stearic acid and Mn 2+ , and should be buried in the hydrophobic interior of the micelle. The SDS micelles were characterized by diffusion and 15 N-relaxation measurements. Comparison of experimentally determined translational diffusion coefficients for SDS and Ab(1-40) show that the size of SDS micelle is not significantly changed by interaction with Ab(1-40).
The downstream prion-like Doppel (Dpl) protein is a homologue related to the prion protein (PrP).... more The downstream prion-like Doppel (Dpl) protein is a homologue related to the prion protein (PrP). Dpl is expressed in the brains of mice that do not express PrP, and Dpl is known to be toxic to neurons. One mode of toxicity has been suggested to involve direct membrane interactions. PrP under certain conditions of cell trafficking retains an uncleaved signal peptide, which may also hold for the much less studied Dpl. For a peptide with a sequence derived from the N-terminal part (1-30) of mouse Dpl (mDpl(1-30)) CD spectroscopy shows about 40% R-helical structure in DHPC and SDS micelles. In aqueous solution it is mostly a random coil. The three-dimensional solution structure was determined by NMR for mDpl(1-30) associated with DHPC micelles. 2D 1 H NMR spectra of the peptide in q) 0.25 DMPC/DHPC bicelles only showed signals from the unstructured termini, indicating that the structured part of the peptide resides within the lipid bilayer. Together with 2 H 2 O exchange data in the DHPC micelle solvent, these results show an R-helix protected from solvent exchange between residues 7 and 19, and suggest that the R-helical segment can adopt a transmembrane localization also in a membrane. Leakage studies with entrapped calcein in large unilamellar phospholipid vesicles showed that the peptide is almost as membrane perturbing as melittin, known to form pores in membranes. The results suggest a possible channel formation mechanism for the unprocessed Dpl protein, which may be related to toxicity through direct cell membrane interaction and damage.
Non-invasive biological information about residual neuroblastoma tumour tissue could allow treatm... more Non-invasive biological information about residual neuroblastoma tumour tissue could allow treatment monitoring without the need for repeated biopsies. Magnetic resonance spectroscopy (MRS) can be performed with standard MR-scanners, providing specific biochemical information from selected tumour regions. By proton 1H-MRS, lipids, certain amino acids and lactate can be detected and their relative concentrations estimated in vivo. Using experimental models of neuroblastoma, we have described the potential of 1H-MRS for the prediction of tumour tissue viability and treatment response. Whereas viable neuroblastoma tissue is dominated by the choline 1H-MRS resonance, cell death as a consequence of spontaneous necrosis or successful treatment with chemotherapy, angiogenesis inhibitors, or NSAIDs is associated with decreased choline content. Therapy-induced neuroblastoma cell death is also associated with enhanced 1H-MRS resonances from mobile lipids and polyunsaturated fatty acids. The mobile lipid/choline ratio correlates significantly with cell death and based on the dynamics of this ratio tumour regression or continued growth (drug resistance) after chemotherapy can be predicted in vivo. The implications of these findings are discussed with focus on the potentials and limitations of introducing 1H-MRS for clinical assessment of treatment response in children with neuroblastoma. Biochemical monitoring of neuroblastoma with 1H-MRS could enable tailoring of individual therapy as well as provide early pharmacodynamic evaluation of novel therapeutic modalities.
A new early-onset form of Alzheimer's disease (AD) was described recently where a point ... more A new early-onset form of Alzheimer's disease (AD) was described recently where a point mutation was discovered in codon 693 of the β-amyloid (Aβ) precursor protein gene, the Arctic mutation. The mutation translates into a single amino acid substitution, glutamic acid→glycine, in position 22 of the Aβ peptide. The mutation carriers have lower plasma levels of Aβ than normal, while in vitro studies show that Aβ1–40E22G protofibril formation is significantly enhanced. We have explored the nature of the Aβ1–40E22G peptide in more detail, in particular the protofibrils. Using size-exclusion chromatography (SEC) and circular dichroism spectroscopy (CD) kinetic and secondary structural characteristics were compared with other Aβ1–40 peptides and the Aβ12–28 fragment, all having single amino acid substitutions in position 22. We have found that Aβ1–40E22G protofibrils are a group of comparatively stabile β-sheet-containing oligomers with a heterogeneous size distribution, ranging from >100 kDa to >3000 kDa. Small Aβ1–40E22G protofibrils are generated about 400 times faster than large ones. Salt promotes their formation, which significantly exceeds all the other peptides studied here, including the Dutch mutation Aβ1–40E22Q. Position 22 substitutions had significant effects on aggregation kinetics of Aβ1–40 and in Aβ12–28, although the qualitative aspects of the effects differed between the native peptide and the fragment, as no protofibrils were formed by the fragments. The rank order of protofibril formation of Aβ1–40 and its variants was the same as the rank order of the length of the nucleation/lag phase of the Aβ12–28 fragments, E22V>E22A≫E22G>E22Q≫E22, and correlated with the degree of hydrophobicity of the position 22 substituent. The molecular mass of peptide monomers and protofibrils were estimated better in SEC studies using linear rather than globular calibration standards. The characteristics of the Aβ1–40E22G suggest an important role for the peptide in the neuropathogenesis in the Arctic form of AD.
The dynamics of monomeric Alzheimer Ab(1-40) in aqueous solution was studied using heteronuclear ... more The dynamics of monomeric Alzheimer Ab(1-40) in aqueous solution was studied using heteronuclear NMR experiments. 15 N NMR relaxation rates of amide groups report on the dynamics in the peptide chain and make it possible to estimate structural propensities from temperature-dependent relaxation data and chemical shifts change analysis. The persistence length of the polypeptide chain was determined using a model in which the influence of neighboring residue relaxation is assumed to decay exponentially as a function of distance. The persistence length of the Ab(1-40) monomer was found to decrease from eight to three residues when temperature was increased from 3 to 18°C. At 3°C the peptide shows structural propensities that correlate well with the suggested secondary structure regions of the peptide to be present in the fibrils, and with the a-helical structure in membrane-mimicking systems. Our data leads to a structural model for the monomeric soluble b-peptide with six different regions of secondary structure propensities. The peptide has two regions with b-strand propensity (residues 16-24 and 31-40), two regions with high PII-helix propensity (residues 1-4 and 11-15) and two unstructured regions with higher mobility (residues 5-10 and 25-30) connecting the structural elements.
Journal of Magnetic Resonance, Series B, Apr 1, 1996
The spectral-density mapping of a 13 C a-1 H a vector of Leu 10 in mension to the relaxation data... more The spectral-density mapping of a 13 C a-1 H a vector of Leu 10 in mension to the relaxation data, although the information the 22-residue peptide hormone motilin [P. Allard, J. Jarvet, A. content of the six relaxation rates in the time domain is Ehrenberg, and A. Grä slund, J. Biomol. NMR 5, 133-146 (1995)] the same as for the spectral-density values in the frequency is extended in this paper to three polarizing fields 9.4, 11.7, and domain. The correspondence between the two representa-14.1 T in order to improve the accuracy of the calculated spectraltions is somewhat analogous to the relation between the freedensity function J(v) and to extend the sampling range up to induction decay and the spectrum in NMR spectroscopy. 750 MHz. The problem with a usually large relative error in J(v H) When experiments are done at different magnetic fields, one is eliminated since the generally more precise J(v H 0 v C) and can calculate the slope of the spectral-density function at J(v H / v C) determined at other fields appear at nearly the same different sampling frequencies. The slope of the spectralfrequencies. The fitting of dynamic models to the points of spectral density was made with error weighting, and the influence of J(v H) density function at different frequencies is very useful when was found to be negligible. Therefore, the high-frequency part comparison is made between experimental data and dynamiof the spectral-density function is determined essentially without cal models. influence from the two transverse-type relaxation rates. In the case With the spectral-density mapping technique, the meaof a carbon-proton vector, the relaxation is mainly determined surement of the spectral-density function is separated from by dipolar interaction and is only weakly influenced by other relaxthe fitting of dynamic models, which is not true, e.g., when ation mechanisms, which makes it particularly suitable for the using the model-free approach of Lipari and Szabo (2, 3). spectral-density mapping technique. The measured relaxation To obtain spectral-density values, which can discriminate rates in the time domain are transformed into the frequency dobetween different motional models, the precision of experimain by spectral-density mapping, and the slopes in different frequency regions are important parameters when comparing experimental relaxation rates must be high. Less-accurate data are mental data with theoretical models of motion. Using an adjustable generally used as input for the model-free analysis (2, 3) internuclear distance r eff , combined with the model-free approach, which still leads to reasonable values for the overall rotait is possible to obtain a reasonable fit to measured spectral-density tional correlation time t m and for the general order parameter points at J(0) and around J(v C). At the same time, however, the S. For peptides as well as for proteins, the correlation time high-frequency slope of the spectral-density function defined by for overall rotational motion is typically on the order of J(v H 0 v C) and J(v H / v C) could not be reproduced. ᭧ 1996 nanoseconds, and, as a consequence, the numerical values Academic Press, Inc. of different spectral-density points differ in size by at least an order of magnitude at the magnetic fields used here. When the model-free approach was introduced (2, 3), the
Uploads
Papers by Jüri Jarvet