FASEB J. Mp Biomedicals
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published as doi: 10.1096/fj.07-097774.
 (The FASEB Journal. 2008;22:2350-2356.)
© 2008 FASEB
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Supplemental Data
Right arrow All Versions of this Article:
fj.07-097774v1
22/7/2350    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Parasassi, T.
Right arrow Articles by Ursini, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Parasassi, T.
Right arrow Articles by Ursini, F.

Low density lipoprotein misfolding and amyloidogenesis

Tiziana Parasassi*,1, Marco De Spirito{dagger}, Giampiero Mei{ddagger}, Roberto Brunelli§, Giulia Greco*, Laura Lenzi*, Giuseppe Maulucci{dagger}, Eleonora Nicolai{ddagger}, Massimiliano Papi{dagger}, Giuseppe Arcovito{dagger}, Silvio C. E. Tosatto and Fulvio Ursini||

* Istituto di Neurobiologia e Medicina Molecolare, CNR, Rome, Italy;

{dagger} Istituto di Fisica, Facoltà di Medicina e Chirurgia, Università Cattolica del Sacro Cuore, Rome, Italy;

{ddagger} Dipartimento di Medicina Sperimentale e Scienze Biochimiche, Università di Roma "Tor Vergata," Rome, Italy;

§ Dipartimento di Scienze Ginecologiche, Perinatologia e Puericultura, Università di Roma "La Sapienza," Rome, Italy;

Dipartimento di Biologia and Centro per le Biotecnologie Innovative (CRIBI), and

|| Dipartimento di Chimica Biologica, Università di Padova, Italy

1Correspondence: Istituto di Neurobiologia e Medicina Molecolare, CNR, Via del Fosso del Cavaliere 100, 00133 Rome, Italy. E-mail: t.parasassi{at}inmm.cnr.it


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In early atherogenesis, subendothelial retention of lipidic droplets is associated with an inflammatory response-to-injury, culminating in the formation of foam cells and plaque. Low density lipoprotein (LDL) is the main constituent of subendothelial lipidic droplets. The process is believed to occur following LDL modification. Searching for a modified LDL in plasma, electronegative LDL [LDL(–)] was identified and found to be associated with major risk biomarkers. The apoprotein in LDL(–) is misfolded, and we show here that this modification primes the aggregation of native LDL, conforming to the typical pattern of protein amyloidogenesis. After a lag phase, whose length depends on LDL(–) concentration, light scattering and atomic force microscopy reveal early exponential growth of intermediate globules, which evolve into fibrils. These globules are remarkably similar to subendothelial droplets in atheromatous lesions and different from those produced by oxidation or biochemical manipulation. During aggregation, ellipticity and tryptophan fluorescence measurements reveal a domino-style spread of apoprotein misfolding from LDL(–) to all of the LDL. Computational analysis of the apoprotein primary sequence predicts an unstable, aggregation-prone domain in the regulatory {alpha}2 region. Apoprotein misfolding well represents an LDL modification able to transform this cholesterol carrier into a trigger for a response-to-injury in the artery wall.—Parasassi, T., De Spirito, M., Mei, G., Brunelli, R., Greco, G., Lenzi, L., Maulucci, G., Nicolai, E., Papi, M., Arcovito, G., Tosatto, S. C. E., Ursini, F. Low density lipoprotein misfolding and amyloidogenesis.


Key Words: amyloidoses • atherogenesis • computational analysis • electronegative LDL • nucleation


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
THE LONG QUEST FOR A MISSING link between plasma cholesterol levels and cardiovascular disease risk has been focused on the modification of low density lipoprotein (LDL), transforming an otherwise innocuous cholesterol carrier into a damaging agent able to trigger a cellular response to injury, thus initiating the pathological process (1) . According to our current understanding of early atherogenesis, subendothelial retention of lipidic aggregates, containing mostly LDL, elicits an inflammatory response that leads to the formation of foam cells and plaque (2 3 4) .

Several LDL modifications, including treatments with lipases, proteases, or lipid peroxidation, have been attempted in vitro to reproduce early biological events of the disease (5 6 7) . These drastic modifications in LDL actually elicit inflammatory or generally toxic responses in cells in culture and often also result in aggregates that are taken up by macrophages, thus suggesting an initial formation of foam cells. Despite some similarities between in vitro-modified LDL and the components of subendothelial lipidic droplets found in early atherogenesis (8 9) , conclusive evidence for the mechanism behind LDL modification and its relationship with subendothelial aggregates has not been found.

Until now, the electronegative LDL [LDL(–)] has been the only modified lipoprotein detected and characterized in plasma (10) . Its increased concentration in hypercholesterolemia, type 2 diabetes, uremia, exhausting physical exercise, and postprandial lipemia suggested a valid correlation with increased cardiovascular disease risk (11 12 13 14 15) . Possibly linked to increased electronegativity, the major difference between native LDL and LDL(–) is the misfolding of its apoprotein in LDL(–), with a marked increase in β-sheet structure and a significant conformational shift (16 17) .

In the present work, we link the misfolded character of apoprotein B-100 (apoB-100) in LDL(–) to the ability of this particle to trigger LDL aggregation, with a mechanism fully in agreement with that described for 3the multistep process of protein amyloidogenesis (18 19 20) . Apoprotein misfolding and LDL aggregation can well represent the modification able to transform a cholesterol carrier into a trigger for a response to injury in the artery wall.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
LDL was isolated from venous blood following routine procedures (21) . Blood donors were healthy, normolipemic, and normal-fasting volunteers in the 29- to 54-yr age range. Isolation of LDL(–) from total LDL was accomplished through preparative anion-exchange chromatography in an ÁKTA-FPLC system (GE Healthcare, Amersham Biosciences, Piscataway, NJ, USA) using a MonoQ 5/50 GL column and a multistep gradient from 0 to 0.3 M NaCl. Fractions corresponding to native LDL (nLDL) and to LDL(–) were pooled, and salts were removed by overnight dialysis against Chelex-treated argon-purged PBS. A total of 31 LDL(–) and nLDL separations were performed, from plasma of 9 subjects. Mixtures containing different concentrations of nLDL and LDL(–) were prepared based on protein determination using Bradford reagent. Aggregation and fibrillation were achieved by simply maintaining LDL samples at 37°C, either in capped Eppendorf tubes, directly in glass tubes for scattering measurements, or, alternatively, in quartz cuvettes for circular dichroism and fluorescence spectroscopy. Given the reported LDL artifactual aggregation after vortexing or stirring (5) , care was taken not to subject LDL samples to agitation. Sample aliquots collected during aggregation served for atomic force microscopy (AFM). Because of the relatively long incubation time required to detect aggregation and fibrillation, few samples were allowed to aggregate in sterile conditions [i.e., after filtering LDL(–) and nLDL through 0.2 µm Millipore (Billerica, MA, USA) filters and using sterilized tubes or cuvettes]. This last treatment did not change the course of observed phenomena.

AFM images were acquired using an SPMagic SX atomic force microscope (Elbatech, Marciana, Livorno, Italy) in the tapping operation mode. The microscope probe consisted of an ultrasharp silicon nitride cantilever of nominal force constant k = 40 N/m with a tip radius < 10 nm (MikroMash, Tallinn, Estonia). Image analysis was performed using WSxM software (Nanotec Electrónica, Madrid, Spain).

The light scattered from 1 mg/ml of LDL protein (corresponding to 155 mg/dl cholesterol) was monitored by means of a commercial light-scattering ALV spectrometer setup (ALV, Langen, Germany) consisting of a CGS-5000 rotating arm goniometer, a photomultiplier tube (EMI, Ruislip, UK), an ALV 5000 multitau digital correlator operating at a sampling time of 200 ns, and an Innova 70 argon ion laser (Coherent, Santa Clara, CA, USA) operated at 488 nm and 100 mW. The scattering cell was immersed in a refractive index matching fluid (toluene) kept at 37 ± 0.1°C. Light-scattering data were collected simultaneously from a scattering volume of ~100 µm3, and analyzed using ad hoc software. Angular runs were performed with angles logarithmically scaled in sin({theta}), where {theta} is the scattering angle, ranging from 30° to 150°.

Light-scattering data analysis allows for the recovery of the principal features of particle aggregation and the fractal nature of inner cluster structure, this being quantified by measuring the fractal dimension (df) (21) . If aggregates are larger than the incident wavelength {lambda}, the scattering intensity is related to df by:

Formula 1(1)
where q is the scattering wavevector [q=(4{pi}n/{lambda})sin({theta}/2)] and n is the refractive index. The fractal df can also be recovered by analyzing AFM images according to a perimeter-area relationship (22) :

Formula 2(2)
where the area A is the number of pixels making up a given aggregate, the perimeter P is a count of the number of pixel edges, and k is a scaling constant. In light-scattering experiments, when aggregates start to polymerize into fibrils, interference phenomena occur and Eq. 1 does not hold any longer (23) , so that df can only be calculated from AFM images.

Fluorescence emission spectra of tryptophan residues in LDL (0.1 µM as the apoprotein concentration) were acquired using a K2 fluorimeter (ISS Inc., Champaign, IL, USA) equipped with a xenon arc lamp and photon counting electronics (PX01; ISS Inc.). For thioflavin-T (ThT), fluorescence excitation spectra of unlabeled LDL samples (0.1 µM) were acquired, which were later subtracted as the blank. Samples were then labeled by adding 1 µM ThT (final concentration). Excitation spectra were acquired from 300 to 450 nm, using emission at 480 ± 8 nm. The cell holder compartment was maintained at 37.0 ± 0.1°C.

Circular dichroism spectra from the different LDL samples (0.1 µM as the apoprotein concentration) were recorded on a spectropolarimeter (Jasco, Tokyo, Japan) using a 0.1 cm quartz cuvette. The cell holder compartment was maintained at 37 ± 0.1°C. Six spectra were averaged for each measurement, and the blank was subtracted.

Aggregation propensity for the apoB-100 primary sequence was predicted based on local pairing preferences using Pasta (24 25) . Pasta determines the most likely pairing arrangement, parallel or antiparallel, for different copies of the same protein from the relative frequency of all possible beta sheet arrangements between parts of the sequence. Secondary structure and disorder predictions were performed using a consensus method (26) and Spritz (27) , respectively. The results were visualized using the Jalview multiple alignment editor (28) . The consensus secondary structure prediction confirmed a preference, in the native state, for an {alpha}-helical conformation of those sequences that appear to be prone to aggregation.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
LDL(–) is a naturally occurring LDL aggregation trigger
LDL(–) isolated from normal human plasma and added to with unmodified native LDL (nLDL) was able to initiate aggregation in a concentration-dependent manner. Indeed, below 5 mol% LDL(–), no aggregation could be detected. In Fig. 1 , we report the static light-scattering intensity of samples composed of different mixtures of nLDL and LDL(–) as a function of time at 37°C. Consistent with a nucleation event, the lag phase of aggregation was dependent on LDL(–) concentration (inset in Fig. 1 ). Following the lag phase, aggregation increased exponentially, eventually reaching a plateau. At this stage, in a sample composed of 20 mol% LDL(–) in nLDL, AFM images revealed the presence of globular clusters (Fig. 2 A, D). These intermediate clusters displayed a relatively homogeneous distribution, with an average diameter of 360 ± 50 nm.


Figure 1
View larger version (24K):
[in this window]
[in a new window]

  		Figure 1. LDL(–) triggers aggregation of LDL. Light-scattering intensity, recovered at {theta} = 90°, as a function of the incubation time at 37°C, in samples composed of different concentrations of LDL(–) in nLDL (mol% of apoprotein). Final total LDL protein concentration was 1 mg/ml, corresponding to 155 mg/dl cholesterol. Inset: length of the lag phase plotted vs. the mol% of LDL(–) in nLDL; dotted line represents a linear regression.


Figure 2
View larger version (66K):
[in this window]
[in a new window]

  		Figure 2. LDL fibrillogenesis. A–F) AFM images of a sample composed of 20% LDL(–) in nLDL (mol % of apoprotein) taken at different times during aggregation. Protofibrillar aggregates (A, D), fibrils (B, C, E) and thicker late fibrils (F). Images D and E are details magnified from images A and B, respectively. G) Light-scattering intensity, recovered at {theta} = 90°, as a function of the incubation time at 37°C of the 20% LDL(–) in nLDL sample (circles) and of nLDL alone (triangles). H) Representative fluorescence excitation spectra of the amyloid-selective fluorescent probe ThT in nLDL (continuous line) and in the 20% mixture of LDL(–) in nLDL at time 0 (dot), 30 h (dot-dash), and at 72 h (dash).

Multistep LDL aggregation: from globules to fibrils
LDL globular aggregates evolve further with time. Over a period of up to 3 days, globular clusters aggregate into fern-like structures, which progressively thicken with time (Fig. 2B, C, E, F ). AFM images revealed fibrils, composed of a linear alignment of globules, while the thicker fibers’ forms suggested a parallel addition of filaments (Fig. 2F ). The earliest developing globular aggregates appear to constitute all fibrillation stages and are still clearly identifiable in the final thick fibers. A transmission microscopy movie of fibrils in buffer showed that they are mobile and flexible, thus ruling out possible artifacts that commonly occur in fixed and dehydrated samples (Supplemental Material).

Consistent with these morphological modifications, the light-scattering intensity increased over the full aggregation process (Fig. 2G ). LDL aggregation dynamics involved an initial lag phase related to the attainment of a nucleation threshold; exponential growth of intermediate globular clusters; a second lag phase, required for the fibers’ eventual nucleation; and, finally, a phase of rapid fibril growth (Fig. 2G ).

The amyloid nature of LDL final aggregates was tested using the ThT, considered an amyloid-selective fluorescent probe (29 30) . Indeed, during the aggregation of a mixture of nLDL and LDL(–), the fluorescence signal increased, particularly in mature fibrils (Fig. 2H ), while no increase was observed in the absence of LDL(–).

Elementary LDL units maintain their individuality during aggregation
Together with the homogeneous distribution reported above, analysis of light-scattering data showed that intermediate globular clusters also displayed a df = 2.2 ± 0.1 (Fig. 3 A). The fractal structure of globular clusters and their aggregation kinetics are consistent with the absence of fusion during aggregation, indicating instead that elementary units, with a diameter of 22.6 ± 0.4 nm, maintain their individuality (21) .


Figure 3
View larger version (8K):
[in this window]
[in a new window]

  		Figure 3. Structural evolution of aggregates. The linear behavior of a log-log plot of I(q) vs. q, or of P vs. A, defines a fractal structure. From the plot’s slope, df can be calculated using Eqs. 1 and 2 . Data of intermediate globule scattering (A) yield df = 2.2 ± 0.1, typical of structures grown by a reaction-limited aggregation. AFM images of early (B) and late (C) fibrils yield df = 1.2 ± 0.1 and df = 1.5 ± 0.1, respectively, characteristic of fern-like structures, which thicken with time.

A linear arrangement of these intermediate globules gives rise to fibrils, clearly evident in AFM images. Consistently, the df of fibrils, using Eq. 2 , shifts to values approaching linearity, specifically 1.2 in early fibrils (Fig. 3B ) and 1.5 in later, thicker fibers (Fig. 3C ).

Domino propagation of LDL(–) misfolding
A peculiar feature in protein amyloidogenesis is the spread of misfolding from the triggering protein to the overall aggregating proteins (20) . This characteristic was also observed in LDL. In comparison with nLDL, apoB-100 in LDL(–) has already been shown to possess a decreased ellipticity, related to an {alpha}-helix to β-sheet transition, and a decreased fluorescence emission of tryptophan residues, suggestive of an exposure to the aqueous environment of these residues that are otherwise buried in the protein core (16) . Therefore, with respect to the single components, the apoB-100 in a 1:1 mixture of LDL(–) and nLDL displayed both an intermediate ellipticity and tryptophan fluorescence (Fig. 4 ). Consistent with a domino-type process, as aggregation progressed, this misfolded characteristic spread to the particles overall, with a further decrease in ellipticity and in tryptophan fluorescence (Fig. 4A, B , respectively).


Figure 4
View larger version (23K):
[in this window]
[in a new window]

  		Figure 4. Domino propagation during aggregation of the misfolded structure and conformation of the apoprotein in LDL(–). A) Circular dichroism spectra of nLDL, LDL(–), and the 50% mixture, as labeled in the figure, at time 0 (continuous line) and during aggregation at 12, 28, 34, 44, and 60 h intervals (dash-dot). B) Tryptophan fluorescence emission spectra (295±8 nm excitation) of nLDL, LDL(–), and of the 50% mixture, as labeled in the figure, at time 0 (continuous line) and during aggregation at 12, 30, 54, and 90 h intervals (dot-dash).

Computational analysis predicts an aggregation-prone region in apoB-100
The tendency toward structural changes in apoB-100 was investigated by scoring the propensity to a conformational switch, from {alpha}-helix to β-fibrillar aggregates, for each amino acid along the primary sequence (24 25) . The aggregation profile predicted 3 aggregation-prone regions, with the central one (residues 2191–2326) displaying by far the highest propensity (0.0110 vs. 0.0027 and 0.0015 in the other 2 side regions; Fig. 5 ). Conversely, the preference for an {alpha}-helical conformation in its native state was confirmed by the secondary structure prediction (26) . This region falls into the amphipathic {alpha}-helix cluster in the {alpha}2 domain, reported to range from residues 2075 to 2575 (31) ; the region is also predicted (27) to be largely disordered (i.e., capable of adaptation depending on its environment). Taken together, these predictions suggest a structure that is highly prone to a conformational switch from a native {alpha}-helical arrangement toward a parallel inregister (cross-β) aggregation with other protein molecules, through a central portion of the amphipathic {alpha}-helix cluster in the {alpha}2 domain.


Figure 5
View larger version (13K):
[in this window]
[in a new window]

  		Figure 5. Aggregation propensity in apoB-100 primary sequence. The relative aggregation propensity is shown as a function of the residue number.

Amyloid-β peptide accelerates while a chaperon inhibits LDL aggregation
The specific role of a misfolded protein in LDL(–) in promoting and propagating aggregation was further supported by the observation that a different β-sheet-rich protein could substitute for LDL(–). A proper aliquot of a stock 6 mM solution of the amyloid-β42 peptide in ethanol was added to isolated native LDL, to a final concentration of 5 mol% relative to the nLDL apoprotein. This small amount of amyloid-β42 was sufficient to trigger a massive aggregation, with a lag phase even shorter than that seen in the sample composed of 50% LDL(–) in nLDL (Fig. 6 ). On the other hand, a chaperon protein was expected to prevent aggregation, demonstrated by using {alpha}-crystallin, a protein belonging to the small heat shock protein family, which prevents stress-induced aggregation of partly denatured proteins by either assisting refolding or hindering the interaction between misfolded proteins (32) . In the 50% mixture of LDL(–) in nLDL, {alpha}-crystallin was able to completely inhibit aggregation (Fig. 6) .


Figure 6
View larger version (20K):
[in this window]
[in a new window]

  		Figure 6. LDL amyloidogenesis is triggered by amyloid-β42 and inhibited by a chaperon. Light-scattering intensity ({theta}=90°) vs. time at 37°C of nLDL containing 5 mol% of the β-sheet-rich amyloid-β42 (squares); the 50% mixture of LDL(–) in nLDL (circles) containing 0.3 mg/ml of the chaperon {alpha}-crystallin (triangles); nLDL only (inverted triangles).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
High levels of plasma LDL represent a major risk factor for vascular disease and are the source of most of the cholesterol that accumulates in the arterial wall. Subendothelial retention of LDL is a key pathogenic process, with lipidic droplets becoming evident in early atherosclerosis (1 2 3 4) . Subsequently, a series of biological responses to this retained material leads to an inflammatory reaction. Nevertheless, a mechanistic link between plasma cholesterol level and a modification of LDL accounting for the transformation of this cholesterol carrier into a harmful agent able to spark a cellular response to retention is still lacking. In this study, we provide evidence that in the presence of LDL(–), the electronegative subclass where the apoprotein is misfolded, LDL can undergo an amyloidogenic aggregation. Our results therefore suggest that apoprotein misfolding is an actual LDL modification that is required to set off the pathogenic mechanism through the formation of amyloid aggregates.

In common with several other generally much smaller proteins or peptides, notwithstanding a massive presence of lipids in close interaction with the apoprotein, the LDL particle undergoes aggregation according to all amyloidogenesis paradigms (18 19 20 , 33) , summarized as follows: an initial structural modification, mostly a β-sheet vs. {alpha}-helix increase, renders the protein prone to aggregation through β/β-strand interaction. Despite being composed of different proteins, amyloid deposits share similar formation dynamics where a lag phase is followed by a period of rapid growth, a behavior typical of nucleated processes such as crystallization. With a seeding mechanism, an increased concentration of misfolded proteins, or of preformed aggregates, shortens the lag phase. When the initial nucleation threshold has been reached, amyloidogenesis proceeds spontaneously by the formation of intermediate structures which later evolve into amyloid-like structures. During this process, the misfolding spreads to all aggregating proteins. Finally, chaperon proteins can effectively inhibit aggregation by either assisting refolding or hindering the interaction between misfolded proteins. Typically, amyloids can be probed by increased ThT fluorescence.

In contrast to other studies on protein aggregation and, specifically, on LDL aggregation, no physical, chemical, or biochemical manipulations were necessary to initiate the formation of LDL aggregates. Indeed, the initial trigger is naturally occurring LDL(–), physiologically present in human plasma at different concentrations. In healthy subjects, its concentration is below 10% of LDL whereas higher concentrations have been reported to correlate positively with other biomarkers of cardiovascular disease risk (11 12 13 14 15) .

In a previous study (21) , we demonstrated a fairly similar initial aggregation pattern using LDL purified from plasma that had been incubated overnight at 37°C. A method for analyzing aggregates that we developed during this earlier study was also used in the present work for the characterization of intermediate globules, and the results are extremely consistent. Both in the aggregation triggered by LDL(–), described here and in that occurring using LDL isolated from incubated plasma, globules formed after a reaction-limited interaction (22) . In this mechanism, not all interactions yield aggregates because the sticking probability between elementary units is <1; favorable collisions involve specific "sticky" sites (i.e., a recognition between selected sites on the particle surface). The predictions of a reaction-limited aggregation mechanism correlate with the determined df, which clearly indicates the absence of fusion and a complex arrangement of the elementary particles, of ~22 nm in size, within the clusters. We also detected decreased ellipticity in LDL purified from incubated plasma (21) , again suggestive of a misfolding-driven aggregation. We can, therefore, reasonably infer that misfolded LDL(–) are produced when plasma is incubated for several hours at 37°C, mimicking the remodeling that occurs in LDL during its intravascular lifetime.

With the continued evolution of aggregation to form fibrils, intermediate globules are maintained and appear as longitudinal linear assemblages in a pearl-necklace fashion. In addition, early globules are still evident as elementary constituents in the later, thicker fibers, formed after a lateral association of earlier fibrils. On the whole, it can be said that LDL aggregation triggered by LDL(–) definitely does not appear to be a random process. In our previous study (21) , we also explored, in detail, aggregates formed by oxidation, because aggregation due to LDL oxidative modification is considered representative of aggregation following lipolytic and proteolytic treatments (8) . We found that oxidized LDL fuses, reaching a final diameter of 34 ± 0.2 nm, without evolving into larger aggregates. This comparison, together with the subsequent formation of ordered fibrils, suggests that the complex properties of aggregates triggered by LDL(–) can be associated with specific biological responses.

The interesting possibility of adding atherosclerosis to the list of already recognized amyloid diseases is reinforced by the finding that LDL globular aggregates described here are remarkably similar to the subendothelial droplets found in atheromatous plaque, whose average diameter ranges from 100 to 400 nm. Indeed, extracellular aggegates larger than 100 nm were visualized on rabbit and human vessels by using both fine freeze-etching and transmission electron microscopy (4 , 8) . In contrast, aggregates produced in vitro by proteolysis, lipolysis, or massive oxidation have diameters less than twice that of native LDL (5 6) , which therefore suggests a fusion process, possibly lipid-driven, rather than aggregation (21) .

The reaction-limited aggregation mechanism as well as the propagation of misfolding to all aggregating LDL fit well with the finding that along the apoB-100 sequence, 3 sites prone to an {alpha}-helix to β-fibrillar switch can be predicted. The central site appears particularly relevant both for its extreme propensity to misfolding and for its location in the {alpha}2 domain. In fact, this domain was previously proposed as a flexible spring-like structure, able to regulate LDL particle size and possibly involved in a shape/density modulation of the entire particle (31 , 34) through protein conformational changes in response to different lipid content. Indeed, the most aggregation-prone region falls in an {alpha}-{alpha} superhelix, forming a highly curved solenoid structure whose shape recalls a handle. Contact with underlying lipids occurs on both sides of the handle, suggesting a mechanical function in accommodating the protein on the particle surface (Fig. 7 ). In this respect, we note that the electronegative LDL(–) subclass is predominantly found in the small, dense LDL fraction (10 11 , 36) . The present prediction of an extreme propensity to aggregation in the apoB-100’s {alpha}2 domain therefore suggests at least one of the possible protein sites for an initial nucleation.


Figure 7
View larger version (41K):
[in this window]
[in a new window]

  		Figure 7. Structural model of the aggregation-prone region in apoB-100. The structure for residues 1931 to 2435 in apoB-100, modeled in ref. 35 as a single domain using PDB 1gw5 as a template, is shown in the same orientation and color as in Fig. 2 E of that article. The region most prone to aggregation is highlighted in red. For clarity, dotted blue circles have been added to highlight the hydrophobic patches predicted as putative contact sites with the lipid core (35) , a schematic representation of the lipid core, and the relative direction of the N and C termini of apoB-100. The image was made using PyMol (http://pymol.sourceforge.net/).

Modified lipoproteins retained within the subendothelium in the form of droplets are regarded as an essential stimulus for the activation of endothelium and for macrophages, T-cells, and mast cells recruitment, as well as for the subsequent migration of smooth muscle cells (1 2 3 4) . In human plasma, concentrations of the misfolded LDL(–) above 10% were reported to correlate positively with other biomarkers of cardiovascular disease risk (11 12 13 14 15) . In this context, the actual occurrence of LDL aggregation in vivo reasonably depends on several other factors promoting or inhibiting aggegation.

Finally, drawing an analogy between LDL globular early aggregates and the protein aggregation that occurs in general amyloidoses, where prefibrillar aggregates are seen as the most detrimental (37) , the former could emerge as the actual agonists triggering the response-to-retention in early atherogenesis.


   ACKNOWLEDGMENTS
 
This work was supported by grants from Indena SpA to T.P. and G.G., by the Italian Ministry for Education, Universities and Research (grant "Rientro dei cervelli") to S.C.E.T., and by Università Cattolica del Sacro Cuore to M.D.S., G.M., M.P., and G.A.

Received for publication October 16, 2007. Accepted for publication January 31, 2008.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Ross, R. (1999) Atherosclerosis: an inflammatory disease. New Engl. J. Med. 340,115-126[Free Full Text]
  2. Skålén, K., Gustafsson, M., Rydberg, E. K., Hultén, L. M., Wiklund, O., Innerarity, T. L., Borén, J. (2002) Subendothelial retention of atherogenic lipoproteins in early atherosclerosis. Nature 417,750-754[CrossRef][Medline]
  3. Williams, K. J., Tabas, I. (2005) Lipoprotein retention–and clues for atheroma regression. Arterioscler. Thromb. Vasc. Biol. 25,1536-1540[Abstract/Free Full Text]
  4. Nievelstein, P. F., Fogelman, A. M., Mottino, G., Frank, J. S. (1991) Lipid accumulation in rabbit aortic intima 2 hours after bolus infusion of low density lipoprotein. A deep-etch and immunolocalization study of ultrarapidly frozen tissue. Arterioscler. Thromb. 11,1795-1805[Abstract/Free Full Text]
  5. Pentikäinen, M. O., Lehtonen, E. M. P., Kovanen, P. T. (1996) Aggregation and fusion of modified low density lipoprotein. J. Lipid. Res. 37,2638-2649[Abstract]
  6. Öörni, K., Sneck, M., Brömme, D., Pentikäinen, M. O., Lindstedt, K. A., Mäyränpää, M., Aitio, H., Kovanen, P. T. (2004) Cysteine protease cathepsin F is expressed in human atherosclerotic lesions, is secreted by cultured macrophages, and modifies low density lipoprotein particles in vitro. J. Biol. Chem. 279,34776-34784[Abstract/Free Full Text]
  7. Öörni, K., Kovanen, P. T. (2007) PLA2-V. A real player in atherogenesis. Arterioscler. Thromb. Vasc. Biol. 27,445-447[Free Full Text]
  8. Öörni, K., Pentikäinen, M. O., Ala-Korpela, M., Kovanen, P. T. (2000) Aggregation, fusion, and vesicle formation of modified low density lipoprotein particles: molecular mechanisms and effects on matrix interactions. J. Lipid Res. 41,1703-1714[Abstract/Free Full Text]
  9. Chao, F.-F., Blanchette-Mackie, E. J., Chen, Y.-J., Dickens, B. F., Berlin, E., Amende, L. M., Skarlatos, S. I., Gamble, W., Resau, J. H., Mergner, W. T., Kruth, H. S. (1990) Characterization of two unique cholesterol-rich lipid particles isolated from human atherosclerotic lesions. Am. J. Pathol. 136,169-179[Abstract]
  10. Sevanian, A., Asatryan, L., Ziouzenkova, O. (1999) Low density lipoprotein (LDL) modification: basic concepts and relationship to atherosclerosis. Blood Purif. 17,66-78[CrossRef][Medline]
  11. Demuth, K., Myara, I., Chappey, B., Vedie, B., Pech-Amsellem, M. A., Haberland, M. E., Moatti, N. (1966) A cytotoxic electronegative LDL subfraction is present in human plasma. Arterioscler. Thromb. Vasc. Biol. 16,773-783[Medline]
  12. Sánchez-Quesada, J. L., Otal-Entraigas, C., Franco, M., Jorba, O., González-Sastre, F., Blanco-Vaca, F., Ordóñez-Llanos, J. (1999) Effect of simvastatin treatment on the electronegative low-density lipoprotein present in patients with familial hypercholesterolemia. Am. J. Cardiol. 84,655-659[CrossRef][Medline]
  13. Sánchez-Quesada, J. L., Benítez, S., Otal, C., Franco, M., Blanco-Vaca, F., Ordóñez-Llanos, J. (2002) Density distribution of electronegative LDL (LDL(–)) in normolipemic and hyperlipemic subjects. J. Lipid Res. 43,699-705[Abstract/Free Full Text]
  14. Sanchez-Quesada, J. L., Pérez, A., Caixàs, A., Ordónmez-Llanos, J., Carreras, G., Payés, A., González-Sastre, F., de Leiva, A. (1996) Electronegative low density lipoprotein subform is increased in patients with short-duration IDDM and is closely related to glycaemic control. Diabetologia 39,1469-1476[CrossRef][Medline]
  15. Ursini, F., Sevanian, A. (2002) Postprandial oxidative stress. Biol. Chem. 383,599-605[CrossRef][Medline]
  16. Parasassi, T., Bittolo-Bon, G., Brunelli, R., Cazzolato, G., Krasnowska, E. K., Mei, G., Sevanian, A., Ursini, F. (2001) Loss of apoB-100 secondary structure and conformation in hydroperoxide rich, electronegative LDL(–). Free Radic. Biol. Med. 31,82-89[CrossRef][Medline]
  17. Ursini, F., Davies, K. J., Maiorino, M., Parasassi, T., Sevanian, A. (2002) Atherosclerosis: another protein misfolding disease?. Trends Mol. Med. 8,370-374[CrossRef][Medline]
  18. Chiti, F., Dobson, C. M. (2006) Protein misfolding, functional amyloid, and human disease. Annu. Rev. Biochem. 75,333-366[CrossRef][Medline]
  19. Stefani, M., Dobson, C. M. (2003) Protein aggregation and aggregate toxicity: new insights into protein folding, misfolding diseases and biological evolution. J. Mol. Med. 81,678-699[CrossRef][Medline]
  20. Dobson, C. M. (2003) Protein folding and misfolding. Nature 426,884-890[CrossRef][Medline]
  21. De Spirito, M., Brunelli, R., Mei, G., Bertani, F. R., Ciasca, G., Greco, G., Papi, M., Arcovito, G., Ursini, F., Parasassi, T. (2006) Low density lipoprotein aged in plasma forms clusters resembling subendothelial droplets: aggregation via surface sites. Biophys. J. 90,4239-4247[CrossRef][Medline]
  22. Weitz, D. A., Lin, M. Y., Huang, J. S. (1987) Fractal and scaling in kinetic colloid aggregation. Safran, S. A. Dark, N. A. eds. Physics of Complex and Supermolecular Fluids ,509-545 Wiley-Interscience New York.
  23. Teixeira, J. (1988) Small-angle scattering by fractal systems. J. Appl. Cryst. 21,781-785[CrossRef]
  24. Trovato, A., Chiti, F., Maritan, A., Seno, F. (2006) Insight into the structure of amyloid fibrils from the analysis of globular proteins. PLoS Comput. Biol. 2,1608-1618
  25. Trovato, A., Seno, F., Tosatto, S. C. E. (2007) The PASTA server for protein aggregation prediction. Protein Eng. Des. Sel. 20,521-523[Abstract/Free Full Text]
  26. Albrecht, M., Tosatto, S. C. E., Lengauer, T., Valle, G. (2003) Simple consensus procedures are effective and sufficient in secondary structure prediction. Protein Eng. 16,459-462[Abstract/Free Full Text]
  27. Vullo, A., Bortolami, O., Pollastri, G., Tosatto, S. C. E. (2006) Spritz: a server for the prediction of intrinsically disordered regions in protein sequences using kernel machines. Nucleic Acids Res. 34,W164-W168[Abstract/Free Full Text]
  28. Clamp, M., Cuff, J., Searle, S. M., Barton, G. J. (2004) The Jalview Java alignment editor. Bioinformatics 20,426-427[Abstract/Free Full Text]
  29. Levine, H. (1999) Quantitation of β-sheet amyloid fibril structures with Thioflavin T. Methods Enzymol. G309,274-284[Medline]
  30. Bader, R., Bamford, R., Zurdo, J., Luisi, B. F., Dobson, C. M. (2006) Probing the mechanism of amyloidogenesis through a tandem repeat of the PI3-SH3 domain suggests a generic model for protein aggregation and fibril formation. J. Mol. Biol. 356,189-208[CrossRef][Medline]
  31. Segrest, J. P., Jones, M. K., De Loof, H., Dashti, N. (2001) Structure of apolipoprotein B-100 in low density lipoproteins. J. Lipid Res. 42,1346-1367[Abstract/Free Full Text]
  32. Sun, Y., MacRae, T. H. (2005) The small heat shock proteins and their role in human disease. FEBS J. 272,2613-2627[CrossRef][Medline]
  33. Lee, C. C., Nayak, A., Sethuraman, A., Belfort, G., McRae, G. J. (2007) A three-stage kinetic model of amyloid fibrillation. Biophys. J. 92,3448-3458[CrossRef][Medline]
  34. Hevonoja, T., Pentikäinen, M. O., Hyvönen, M. T., Kovanen, P. T., Ala-Korpela, M. (2000) Structure of low density lipoprotein (LDL) particles: basis for understanding molecular changes in modified LDL. Biochim. Biophys. Acta 1488,189-210[Medline]
  35. Krisko, A., Etchebest, C. (2007) Theoretical model of human apolipoprotein B100 tertiary structure. Proteins 66,342-358[CrossRef][Medline]
  36. Sevanian, A., Hwang, J., Hodis, H., Cazzolato, G., Avogaro, P., Bittolo-Bon, G. (1996) Contribution of an in vivo oxidized LDL to LDL oxidation and its association with dense LDL subpopulations. Arterioscler. Thromb. Vasc. Biol. 16,784-793[Abstract/Free Full Text]
  37. Baglioni, S., Casamenti, F., Bucciantini, M., Luheshi, L. M., Taddei, N., Chiti, F., Dobson, C. M., Stefani, M. (2006) Prefibrillar amyloid aggregates could be generic toxins in higher organisms. J. Neurosci. 26,8160-8167[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Supplemental Data
Right arrow All Versions of this Article:
fj.07-097774v1
22/7/2350    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Parasassi, T.
Right arrow Articles by Ursini, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Parasassi, T.
Right arrow Articles by Ursini, F.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS