Differential effects on light chain amyloid formation depend on mutations and type of glycosaminoglycans.

Protein Structure and Folding: Differential Effects on Light Chain Amyloid Formation Depend on Mutations and Type of Glycosaminoglycans

J. Biol. Chem. 2015, 290:4953-4965. doi: 10.1074/jbc.M114.615401 originally published online December 23, 2014

Access the most updated version of this article at doi: 10.1074/jbc.M114.615401 Find articles, minireviews, Reflections and Classics on similar topics on the JBC Affinity Sites. Alerts: • When this article is cited • When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts This article cites 50 references, 15 of which can be accessed free at http://www.jbc.org/content/290/8/4953.full.html#ref-list-1

Downloaded from http://www.jbc.org/ at New York Medical College on February 25, 2015

Luis M. Blancas-Mejía, Jared Hammernik, Marta Marin-Argany and Marina Ramirez-Alvarado

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 290, NO. 8, pp. 4953–4965, February 20, 2015 © 2015 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.

Differential Effects on Light Chain Amyloid Formation Depend on Mutations and Type of Glycosaminoglycans* Received for publication, October 2, 2014, and in revised form, December 3, 2014 Published, JBC Papers in Press, December 23, 2014, DOI 10.1074/jbc.M114.615401

Luis M. Blancas-Mejía‡, Jared Hammernik§, Marta Marin-Argany‡, and Marina Ramirez-Alvarado‡¶1 From the Departments of ‡Biochemistry and Molecular Biology and ¶Immunology, Mayo Clinic, Rochester, Minnesota 55905 and the §School of Molecular and Cellular Biology, University of Illinois, Urbana, Illinois 61801 Background: Extracellular amyloid deposits involve glycosaminoglycans (GAGs). Results: Fibrillation of AL proteins was accelerated by heparan sulfate and inhibited by chondroitin sulfate A. Conclusion: Endogenous GAGs can modulate amyloid formation, and their effect is determined by the amyloidogenic properties of AL proteins studied. Significance: Biologically relevant molecules like GAGs play a major role in the amyloidogenicity of AL proteins.

One of the most clinically interesting and perhaps the most complex systemic amyloidosis is amyloid light chain (AL)2

* This work was supported, in whole or in part, by National Institutes of Health Grant R01 GM 071514. This work was also supported by the Mayo Foundation. To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, Mayo Clinic, 200 First St. SW, Rochester, MN 55905. E-mail: [email protected]. 2 The abbreviations used are: AL, amyloid light chain; LC, light chain; GAG, glycosaminoglycan; ThT, thioflavin T; TEM, transmission electron microscopy. 1

FEBRUARY 20, 2015 • VOLUME 290 • NUMBER 8

amyloidosis (1). AL amyloidosis is a fatal, incurable protein misfolding disease characterized by extracellular deposition of immunoglobulin light chains (LCs) into fibrillar aggregates, resulting in multiple-organ failure and death. Despite advancements in the management and treatment of the disease, the molecular determinants for protein misfolding and toxicity in AL amyloidosis are not completely understood (2). In some amyloidoses, such as dialysis-related amyloidosis (caused by ␤2-microglobulin), AA amyloidosis (caused by serum amyloid A protein), or systemic senile amyloidosis (caused by wild type transthyretin), the amyloid precursors are the wild type amino acid sequence. In other amyloidoses, the amyloid precursor amino acid sequence has single mutations (e.g. lysozyme-associated amyloidosis (caused by lysozyme) or familial amyloid polyneuropathy (caused by mutant Transthyretin)). By contrast, the LCs found in AL amyloidosis are always highly heterogeneous in sequence (3, 4). This heterogeneity stems from the fact that LCs are formed by random recombination of multiple gene segments used to generate the required immunoglobulin diversity (e.g. ␬ light chains are synthetized by using one of the 40 V␬ gene segments, together with one of the five possible J␬ gene segments and the single C␬ gene). In addition, somatic hypermutation improves the antibody affinity for the antigen, leading to further sequence variation (5). As a result of the large sequence diversity found in LCs, each pathogenic AL protein presents a different set of mutations. These mutations contribute to the differences found among AL amyloidosis patients and proteins: the different propensities for amyloid formation, organ involvement, and degrees of severity of the disease (6). Cardiac involvement, for example, is found in 50% of AL amyloidosis patients and represents the worst prognosis, with a median survival of less than 1 year (7–9). Somatic mutations do not greatly affect the overall immunoglobulin structure (10 –14), but mutations in AL proteins can reduce the thermodynamic stability and promote amyloid fibril formation (15–17), although we recently reported that an AL protein can be too unstable to form fibrils, suggesting a thermodynamic stability threshold for fibril formation (18). Amyloid deposits are found in the extracellular space and involve a variety of accessory molecules. Proteoglycans are JOURNAL OF BIOLOGICAL CHEMISTRY

4953


Amyloid light chain (AL) amyloidosis is a protein misfolding disease where immunoglobulin light chains sample partially folded states that lead to misfolding and amyloid formation, resulting in organ dysfunction and death. In vivo, amyloid deposits are found in the extracellular space and involve a variety of accessory molecules, such as glycosaminoglycans, one of the main components of the extracellular matrix. Glycosaminoglycans are a group of negatively charged heteropolysaccharides composed of repeating disaccharide units. In this study, we investigated the effect of glycosaminoglycans on the kinetics of amyloid fibril formation of three AL cardiac amyloidosis light chains. These proteins have similar thermodynamic stability but exhibit different kinetics of fibril formation. We also studied single restorative and reciprocal mutants and wild type germ line control protein. We found that the type of glycosaminoglycan has a different effect on the kinetics of fibril formation, and this effect seems to be associated with the natural propensity of each AL protein to form fibrils. Heparan sulfate accelerated AL-12, AL-09, ␬I Y87H, and AL-103 H92D fibril formation; delayed fibril formation for AL-103; and did not promote any fibril formation for AL-12 R65S, AL-103 delP95aIns, or ␬I O18/ O8. Chondroitin sulfate A, on the other hand, showed a strong fibril formation inhibition for all proteins. We propose that heparan sulfate facilitates the formation of transient amyloidogenic conformations of AL light chains, thereby promoting amyloid formation, whereas chondroitin sulfate A kinetically traps partially unfolded intermediates, and further fibril elongation into fibrils is inhibited, resulting in formation/accumulation of oligomeric/protofibrillar aggregates.

Light Chains and Glycosaminoglycans

extracellular space glycoproteins that are bound to glycosaminoglycans (GAGs) (19, 20). GAGs are a group of negatively charged, unbranched, long heteropolysaccharides composed of repeating disaccharide units consisting of a uronic acid and either N-acetylglucosamine or N-acetylgalactosamine. GAGs play multiple roles related to cell signaling and adhesion (21– 23). The structures of the repeating disaccharides determine the classes of GAG, which include heparin, heparan sulfate, chondroitin sulfate A, dermatan sulfate, hyaluronic acid, and keratan sulfate (20, 24). Amyloid fibrils have been found associated with GAGs from proteoglycans in AL amyloidosis (25). In vivo, there is a differential distribution of GAGs in organs and tissues. In normal cardiac tissue, the main GAG is heparan sulfate (⬃25%) with a minor amount of dermatan sulfate (⬃19%) and chondroitin sulfate (⬃15), whereas heart tissue with (AL) amyloid involvement presents comparable amounts of dermatan sulfate (⬃47%) and heparan sulfate (⬃46%) but no detectable chondroitin sulfate. In addition, cardiac tissue of AL amyloidosis patients have a ⬃10-fold increase in the total amount of GAG (mg/g of dry tissue) compared with healthy controls (26), suggesting that the disease process in AL amyloidosis somehow contributes to a change in GAG levels. Early in vitro studies showed an interaction between light chain proteins and various GAGs (27). Trinkaus-Randall et al. (28) later showed that cardiac fibroblast cells incubated with AL light chains internalized the protein, up-regulated GAG production, and secreted highly sulfated

4954 JOURNAL OF BIOLOGICAL CHEMISTRY

GAGs. Sulfation also plays a role enhancing fibrillogenesis of AL light chains in vitro (29, 30). In addition, the type of GAG plays a role as well. We previously demonstrated that dermatan sulfate accelerated seeded amyloid fibril formation of AL protein AL-09, whereas chondroitin sulfate A inhibited fibril formation and yielded spherical intermediates (31). We later showed that GAGs enhance amyloid fibril formation for another AL protein (AL-103) by a transient electrostatic interaction with the soluble precursor protein (29). Taken together, previous reports indicate that the presence of GAGs plays a role in the pathogenesis and the process of amyloid formation in AL amyloidosis, and this role is highly dependent of the type of GAG present in the tissue and the degree of sulfation (oversulfated or desulfated heparin). In this study, we explore the effect of GAGs (heparin, heparan sulfate, chondroitin sulfate A, and dermatan sulfate; Fig. 1) on the kinetics of fibril formation of three AL proteins with similar thermodynamic stability and different kinetics of fibril formation (AL-09, AL-12, AL-103). In particular, we are interested in understanding the effect of relevant somatic mutations in amyloid formation in the presence of GAGs. With this purpose, we studied four single restorative mutants (AL-09 H87Y, AL-12 R65S, AL-103 H92D, AL-103 delP95aIns), one double restorative mutant (AL-09 I34N/H87Y), one reciprocal mutant (␬I Y87H), and the non-amyloidogenic germ line (wild type) control ␬I O18/O8 (32, 33). We hypothesize that the interaction between AL light chains and GAGs depends on (i) the type VOLUME 290 • NUMBER 8 • FEBRUARY 20, 2015


FIGURE 1. Chemical structure of sulfated glycosaminoglycans consensus sequences (A–D) and the control branched glycan dextran sulfate (E) and dextran sulfate (F). Negatively charged groups (sulfate and carboxylic acid groups) on the saccharide sequences are colored in red. The average number of negative charges per saccharide unit is displayed by the numbers in parentheses in light blue.

Light Chains and Glycosaminoglycans of GAG, (ii) the propensity of each protein to populate partial unfolded (altered) amyloidogenic states, and (iii) the presence of kinetic traps in the protein folding/misfolding pathway.


JOURNAL OF BIOLOGICAL CHEMISTRY

4955


MATERIALS AND METHODS Chemicals—Water was Milli-Q grade. Yeast extract and tryptone were from Difco. Other reagents were from Sigma-Aldrich. Glycosaminoglycans—Heparin (Sigma) was obtained from porcine intestinal mucosa with typical activity of ⱖ140 USP units/mg and average molecular mass of ⬃16,000 kDa. Chondroitin sulfate A (Calbiochem) was obtained from bovine trachea and composed of 80% chondroitin sulfate A with the balance being chondroitin sulfate C and D. The average molecular mass was ⬃50,000 kDa. Heparan sulfate (Calbiochem) was obtained from bovine intestinal mucosa with average molecular mass of ⬃7,500 kDa. Dermatan sulfate (Calbiochem) was obtained from porcine intestinal mucosa. The average molecular mass was ⬃30,000 kDa. Dextran sulfate (Sigma-Aldrich) was obtained from Leuconostoc spp. The average molecular mass range listed was ⬃9,000 –20,000 kDa. Dextran (SigmaAldrich) was also obtained from Leuconostoc mesenteroides. The average molecular mass was ⬃50,000 kDa. All GAGs were used without further purification. Cloning, Expression, Extraction, and Purification of Recombinant AL-09, AL-12, AL-103, ␬I O18/O8, and Mutant Proteins— AL-09, AL-12, and AL-103 are patient-derived proteins belonging to the germ line gene product, ␬I O18/O8 (also known as IGKV 1–33), whose sequences were obtained from patients with cardiac involvement (4). The AL-103 variable domain sequence was previously obtained from a patient who presented heart, liver, and tongue AL deposits (GenBankTM accession number AY701640); the AL-09 (GenBankTM accession number AF490909) and the AL-12 (GenBankTM accession number AF490912) variable domain protein sequences were obtained from patients with cardiac deposits. ␬I O18/O8 germ line (IGKV 1–33) DNA was generated (34) by mutating the cDNA of AL-103 (␬I O18/O8 sequence deposited under GenBankTM accession number EF640313). Restorative mutants of AL-09, AL-12, and AL-103 and reciprocal mutants of ␬I O18/O8 were generated by using the QuikChangeTM multisite-directed mutagenesis kit (Stratagene, La Jolla, CA) as described previously (32, 33). Protein expression was performed as reported previously (10, 34, 35). Briefly, all plasmids were transformed into Escherichia coli BL21 (DE3) Gold competent cells (Stratagene, La Jolla, CA), and protein expression was induced with 0.8 mM isopropyl-␤-D-thiogalactopyranoside at an A600 nm of 0.7. After 17–20 h of postinduction growth, the bacteria were collected, pelleted, and frozen at ⫺20 °C. The overexpressed AL-09 I34N/H87Y, AL-12, AL-12 R65S, and ␬I O18/O8 proteins were extracted from the periplasmic space of the bacteria by breaking the cells through one freezethaw cycle using phosphate-buffered saline (PBS). The periplasmic fraction was then dialyzed against 10 mM Tris-HCl, 0.02% NaN3, pH 7.4. AL-09, AL-09 H87Y, AL-103, AL-103 H92D, AL-103 del95aIns, and ␬I Y87H were found as insoluble inclusion bodies. The thawed cell pellet was resuspended into 10 mM Tris-

HCl (pH 7.4) and lysed by ultrasonication. Proteins were extracted from solubilized inclusion bodies using 6.0 M urea and immediately dialyzed overnight against 10 mM Tris-HCl (pH 7.4) to remove the urea. All protein extracts were filtered, and the monomeric protein was isolated by size exclusion chromatography on a HiLoad 16/60 Superdex 75 column on an AKTA FPLC (GE Healthcare) system, in 10 mM Tris-HCl, 0.02% NaN3, pH 7.4. The filtration and size exclusion chromatography remove traces of any preexistent aggregate. Protein purity and homogeneity of protein stocks were determined by SDS-PAGE and analytical size exclusion chromatography (see below). Protein concentration was determined by UV absorption using an extinction coefficient of ⑀ ⫽ 14,890 M⫺1 cm⫺1 for ␬I O18/O8, AL-09 I34N/H87Y, AL-103, and AL-103 restorative mutants and ⑀ ⫽ 13,610 cm⫺1 M⫺1 for AL-09, AL-12, AL-12 R65S, and ␬I Y87H, calculated from the amino acid sequence. Pure fractions were combined, concentrated, flash-frozen, and stored at ⫺80 °C. Proteins were thawed at 4 °C, filtered, and ultracentrifuged before they were used for each study. Sample Preparation for in Vitro Fibril Formation Assay— Proteins were thawed at 4 °C, filtered using 0.45-␮m membranes, and equilibrated for 24 h at 4 °C. Filtered, equilibrated proteins were ultracentrifuged at a speed of 90,000 rpm for 3.3 h (enrichment for dimeric species) in a NVT-90 rotor on an Optima L-100 XP centrifuge (Beckman Coulter). This step was carried out to remove any preformed aggregates from the soluble protein, as reported previously (36). Postultracentrifugation far-UV CD spectra and thermal unfolding data were collected and analytical size exclusion chromatography was carried out to ensure the integrity and oligomeric state of the proteins before the initiation of the fibril formation reaction. Far-UV Circular Dichroism and Thermal Unfolding—Circular dichroism (CD) spectroscopy was used to confirm that all proteins in this study retain their native secondary structure at the beginning of the fibril formation reaction. Far-UV CD spectra from 260 to 200 nm (1-nm bandwidth) were acquired at 4 °C, on a Jasco spectropolarimeter 810 (JASCO, Inc., Easton, MD) using a 0.2-cm path length quartz cuvette. All samples (20 ␮M) were prepared in 10 mM Tris-HCl, pH 7.4. Thermal unfolding/refolding experiments were carried out following the ellipticity at 217 nm over a temperature range of 4 –90 °C to determine the melting temperature (Tm) as reported previously (36). Temperature was regulated within ⫾0.002 °C using a Peltier system. We previously reported that AL-103 and its restorative mutants exhibited hysteresis and scan rate dependence in the thermal unfolding/refolding transitions. However, we also demonstrated that apparent Tm values obtained from thermal unfolding experiments using a scan rate of 0.5 °C/min are very close to the absolute Tm values (33). Size Exclusion Chromatography—Analytical size exclusion chromatography was carried out at 4 °C using a BioSil 125–5 HPLC (Bio-Rad) size exclusion column on an AKTA FPLC (GE Healthcare). The column was equilibrated with 50 mM Na2HPO4, 50 mM NaH2PO4, 150 mM NaCl, and 0.02% NaN3 at pH 6.8. Chromatographic analyses were carried out at 0.2 ml/min using the same buffer as elution buffer. ␬ light chain variable domains have been characterized previously as weak


y⫽

A1 ⫺ A2 ⫹ A2 1 ⫹ e 共 x ⫺ x 0兲/dx


(Eq. 1)

where A1 is the initial fluorescence value, A2 is the final fluorescence value, x0 is the center (t50 value), and dx is defined as the time constant. A larger t50 value indicates a longer time required to form fibrils. The comparison of the effects of fibril formation in the presence or absence of GAGs was based on paired Student’s t test. Significance was reported at the 95% (p ⬍ 0.05) confidence level. Electron Microscopy—A 3-␮l fibril sample was placed on a 300-mesh copper Formvar/carbon grid (Electron Microscopy Science, Hatfield, PA), and excess liquid was removed. The samples were negatively stained with 2% uranyl acetate, washed twice with H2O, and air-dried. Grids were analyzed on a Philips Tecnai T12 transmission electron microscope at 80 kV (FEI, Hillsboro, OR).

RESULTS Proteins Selected in This Study; Secondary Structure and Thermodynamic Stability of AL Proteins—For this study, restorative and reciprocal mutants have been selected to understand the effect of specific amino acids in the fibril formation properties of these proteins in the presence of different glycosaminoglycans. AL-09 (hereby denoted as highly amyloidogenic) is the fastest amyloid fibril former protein that we have characterized in our laboratory (38). We chose the restorative mutant AL-09 H87Y (hereby denoted as non-amyloidogenic) because it regains most of the thermodynamic stability (reflected in an increased Tm and a lower ⌬Gfolding value) found in the nonamyloidogenic germ line ␬I O18/O8, whereas it significantly delays fibril formation compared with its parent protein AL-09 (reflected in an increased t50 value). On the other hand, the reciprocal ␬I Y87H (hereby denoted as amyloidogenic) presents a decreased stability and accelerated fibril formation compared with ␬I O18/O8 (32). The double restorative mutant AL-09 I34N/H87Y (hereby denoted as non-amyloidogenic) increases the thermodynamic stability slightly beyond that of ␬I O18/O8 and has a larger delay in fibril formation compared with ␬I O18/O8 (32), suggesting that Ile-34 and His-87 play an important role in the amyloid formation process for AL-09. AL-103 (hereby denoted as amyloidogenic) presents similar thermodynamic stability parameters compared with AL-09 but shows a strong kinetic control in protein folding and delayed and restricted amyloid formation kinetics, possibly due to the presence of the somatic mutation P95aIns (33, 36, 38). We chose the restorative mutant AL-103 H92D (hereby denoted as amyloidogenic) because this amino acid substitution decreases the stability with respect to its parent protein AL-103 (suggesting that H92D is a protective mutation for stability), whereas AL-103 delP95aIns (hereby denoted as non-amyloidogenic) dramatically increased the stability of the protein beyond the ␬I O18/O8 stability. These AL-103 mutants present opposite trends for fibril formation; AL-103 H92D is able to form fibrils at pH 7.4 quite readily, whereas AL-103 delP95aIns does not form fibrils after 800 h of incubation (33). AL-12 (hereby denoted as mildly amyloidogenic) is slightly more stable than AL-09 and AL-103 and presents delayed amyloid formation kinetics with respect to AL-09. Unpublished VOLUME 290 • NUMBER 8 • FEBRUARY 20, 2015


homodimers at relatively high concentrations (⬃700 mM), with a dissociation constant that ranges between 300 and 0.2 ␮M as calculated by NMR diffusion experiments (37). For that reason, protein samples (200 ␮l of pure proteins diluted to 20 ␮M) were incubated for 24 h at 4 °C prior to injection to reestablish the dimer-monomer equilibrium. Chromatographic peaks were detected by UV absorbance at 280 nm. Molecular weight and oligomeric states were estimated from elution volumes using a molecular weight calibration curve as reported before (10). All proteins studied under these experimental conditions were monomeric when prepared at 20 ␮M. The use of ultracentrifugation, far-UV CD, and analytical size exclusion chromatography as quality control mechanisms of the oligomeric state of the proteins in this study allowed us to ensure that the fibril formation reaction experiments were conducted with a homogeneous population of monomeric, native light chain proteins. In Vitro Fibril Formation Assays at Physiological pH—In vitro fibrillogenesis reactions were carried out by monitoring the fluorescence intensity of thioflavin T (ThT) that is enhanced when ThT binds to amyloid fibrils. We considered that a fibril formation reaction has occurred when we observe at least 4-fold ThT fluorescence enhancement (⬃200,000 arbitrary units). Samples of each filtered and ultracentrifuged protein were prepared in 10 mM Tris-HCl buffer containing 150 mM NaCl, 10 ␮M ThT, 0.02% NaN3, at pH 7.4. This protein solution was mixed 1:1 (v/v) with a solution containing each GAG at a concentration of 2.0 mg/ml to reach a final volume of 260 ␮l. All GAGs were dissolved in 10 mM Tris-HCl buffer containing 150 mM NaCl, 10 ␮M ThT, 0.02% NaN3, pH 7.4, and filtered through a 0.22-␮m membrane. In all instances, the final protein concentration was 20 ␮M, and the final GAG concentration was 1.0 mg/ml. Controls containing AL proteins alone were prepared in the same buffer without GAGs as well as in control wells containing GAGs without protein. By performing experiments in parallel, any differences in fibril formation (measured as t50 values, or the time it takes to complete 50% of the fibril formation reaction) could be attributed directly to the presence of the specific GAG while ensuring the reproducibility of the observations. All fibril formation assays were performed in triplicate using black 96-well polystyrene plates (Greiner, Monroe, NC) sealed with plate sealers (Nunc, Roskilde, Denmark), covered with a black polystyrene cover sealed with tape and incubated at 37 °C with continuous orbital shaking (300 rpm) in a New Brunswick Scientific Innova40 incubator shaker. Fibril formation was monitored daily for 1 month (⬃750 h), following fluorescence on a plate reader (Analyst AD, Molecular Devices, Sunnyvale, CA). The excitation wavelength used was 440 nm, and the emission wavelength was 480 nm. The plate sealer and the tape sealing the cover were replaced daily. The t50 value was obtained by fitting each independent kinetic trace to a sigmoidal function (defined as a Boltzmann function by the Origin software package.


FIGURE 2. Structural and stability properties of the AL proteins prior to amyloid formation reactions. Shown are far-UV CD spectra (A) and thermal unfolding (B) of AL-103 H92D (olive), AL-103 (blue), AL-09 (red), AL-12 (green), ␬I Y87H (violet), AL-12 R65S (orange), AL-09 H87Y (yellow), ␬I O18/O8 (brown), AL-09 I34N/H87Y (cyan), and AL-103 delP95aIns (black). All proteins display ␤-sheet structure with the characteristic two minima (235 and ⬃217 nm) for these proteins. Experimental conditions were as follows: 20 ␮M protein in 10 mM Tris-HCl, pH 7.4. Far-UV CD spectra were acquired at 4 °C. Thermal denaturation experiments were performed from 4 –90 °C at a rate of 0.5 °C min⫺1. All experiments were followed by a refolding experiment (not shown).

TABLE 1 Thermodynamic parameter of AL proteins

Protein AL-103 H92Da AL-103a AL-09b AL-12c ␬I O18/O8 Y87Ha AL-12 R65Sc AL-09 H87Yb ␬I O18/O8a AL-09 I34N/H87Yb AL-103 deIP95aInsa

Amyloidogenicity A A HA MA A NA NA NA NA NA

Tm

Cm 关urea兴

°C

M

33.6 ⫾ 1.0 40.5 ⫾ 0.6 41.1 ⫾ 1.0 46.3 ⫾ 0.4 47.3 ⫾ 0.4 49.0 ⫾ 0.2 54.6 ⫾ 0.6 54.7 ⫾ 0.3 58.0 ⫾ 0.1 59.9 ⫾ 0.5

1.98 2.41 1.9 2.3 2.98 2.4 3.29 3.98 4.52 4.62

⌬G folding at 4 °C kcal/mol

⫺3.4 ⫾ 0.2 ⫺4.0 ⫾ 0.3 ⫺3.5 ⫾ 0.3 ⫺4.4 ⫾ 0.8 ⫺4.6 ⫾ 0.4 ⫺4.3 ⫾ 0.9 ⫺6.1 ⫾ 0.3 ⫺6.1 ⫾ 0.2 ⫺6.8 ⫾ 1.3 ⫺8.6 ⫾ 0.3

a

Data from Ref. 3. Data from Ref. 32. c Data from M. Marin-Argany et al. (manuscript in preparation). b

observations3 from our laboratory have indicated that AL-12 R65S (hereby denoted as non-amyloidogenic) is slightly more stable than AL-12. Amyloid fibril formation data shows that restoration of Ser-65 delays the fibrillation kinetics at pH 7.0.4 We selected the restorative AL-12 R65S mutant because it removes the positive charge residue, and we hypothesize that this amino acid may mediate some interactions with negatively charged GAGs. To ensure that the proteins in this study were in the native conformation at the beginning of the fibril formation reaction and retained the secondary structure and thermodynamic properties as published previously (3, 32), far-UV CD spectra and thermal unfolding experiments were carried out for all proteins. All proteins showed the characteristic spectrum with two minima for a light chain variable domain ␤-sheet structure (Fig. 2A): one minimum at ⬃217 nm(typical for ␤-sheets) and a second minimum at ⬃235 nm (due to aromatic residues that are optically active in the far-UV region) (39, 40) as reported previously (41– 46). Thermal unfolding transitions of all samples were sigmoidal and fully reversible over a temperature range of 4 –90 °C (Fig. 2B). Tm values obtained for each protein were in agreement with the previous reported values (Table 1) as expected for fully folded ␬I AL proteins (10, 33, 34, 36). As we mentioned earlier in this section, we have selected the proteins 3

4

L. M. Blancas-Mejía, J. Hammernik, M. Marin-Argany, and M. RamirezAlvarado, unpublished observations. M. Marin-Argany, J. Güell-Bosch, L. M. Blancas-Mejıá, S. Villegas, and M. Ramirez-Alvarado, manuscript in preparation.


in this study for their differences in thermodynamic stability and amyloidogenic potential. The data in Fig. 2B and Table 1 confirmed our previous observations. GAGs Have Different Effects in the Fibrillogenesis of AL Proteins at Physiological pH—Heparin, heparan sulfate, Chondroitin sulfate A, dermatan sulfate, and the long branched polysaccharide controls dextran and dextran sulfate were screened for their ability to promote amyloid fibril formation of all proteins under physiological conditions (150 mM NaCl, pH 7.4, 37 °C). Fibrillogenesis was monitored by the binding of the amyloidspecific dye thioflavin T using a standardized methodology published previously (33, 36) and confirmed by transmission electron microscopy (TEM). To ensure all experiments started with identical samples and that changes in the rate of fibril formations can be attributed unequivocally to the GAG added, the fibril formation reactions were carried out in triplicate using the same GAG, buffer, and protein stocks as well as using the same 96-well plate for each protein. Fig. 3 shows the effect of each GAG as a function of ⌬t50 or the difference in the time it takes to complete 50% of the fibril formation reaction (t50 values) between the protein alone and the protein with each GAG. We will highlight only the most important results for our comparative analysis between AL proteins and their mutants (see Figs. 4-9 for the kinetic traces). AL-12 and AL-12 R65S present dramatic differences in their kinetics of fibril formation. AL-12 fibril formation was accelerated in the presence of heparan sulfate (Fig. 4J). The restorative mutant AL-12 R65S was unable to form JOURNAL OF BIOLOGICAL CHEMISTRY

4957


Values are shown ⫾S.D. (n ⫽ 30). Proteins are shown in order from lower to higher Tm values. HA, highly amyloidogenic; A, amyloidogenic; MA, mildly amyloidogenic; NA, non-amyloidogenic.


FIGURE 4. TEM images of AL-12 at the end point of the reaction. A, pH 7.4; B, heparin; C, chondroitin sulfate A; D, heparan sulfate; E, dermatan sulfate; F, dextran sulfate; G, dextran. Scale bar, 200 nm. Shown are fibril formation kinetics at pH 7.4 (control) (H) in the absence (black) or presence (red) of 1.0 mg/ml heparin (I), chondroitin sulfate A (J), heparan sulfate (K), dermatan sulfate (L), dextran sulfate (M), or dextran (N). Kinetic traces are average of triplicates. Solid lines, fits to the Boltzmann equation. Data were collected for 800 h, but the last data points were discarded due to inner filter effects, as reported previously (38). A.U., arbitrary units. Error bars, S.E.

fibrils in the presence of heparan sulfate (Fig. 5J), suggesting that Arg-65 is essential to allow AL-12 to form fibrils in the presence of Heparan sulfate. Comparisons between AL-103 and AL-103 H92D show that the presence of Asp in position 92 accelerates amyloid formation in the presence of heparin and heparan sulfate (see Figs. 6 and 7 for kinetic traces). H92D is a thermodynamically protective mutation for this protein. A comparison between AL-09, AL-09 H87Y, and the double restorative mutant AL-09 I34N/H87Y (data not shown) shows that the single and double restorative mutation abolish the fibril formation under all conditions tested, whereas AL-09 fibril formation is accelerated in the presence of heparan sulfate and dextran sulfate (Fig. 8). This confirms that Ile-34 and His-87 play a crucial role in facilitating AL-09 amyloid formation under a number of conditions, including in the presence of GAGs.


In contrast, ␬I Y87H is able to form fibrils under all conditions tested (Fig. 9), whereas its parent protein, ␬I O18/O8, does not form fibrils at all (data not shown). It is worth noting that only one of the three wells of ␬I Y87H in the presence of chondroitin sulfate A gave the established 4-fold ThT fluorescence enhancement considered as a positive fibril formation reaction. Additionally, we noted that ␬I Y87H fibril formation kinetics in the presence of GAGs and at pH 7.4 (Fig. 9, H–M) resemble those of AL-09 (see Fig. 8, H–M). This mutation results in an intermediate level of stability between ␬I O18/O8 and AL-09 and appears to have caused enough disruption to promote amyloid under the conditions tested in this study. GAGs Affect Fibril Morphology at Physiological pH—TEM was employed to confirm and characterize the morphology of protein aggregates formed in the absence and in the presence of VOLUME 290 • NUMBER 8 • FEBRUARY 20, 2015


FIGURE 3. Differential effect of GAGs on the t50 value of fibril formation kinetics of AL proteins at physiological pH. All reactions were performed with 20 mM protein and 1.0 mg ml⫺1 GAG. Data are from fibril formation reactions conducted in triplicate. Reaction was considered positive when ThT fluorescence increased 4-fold (⬎200,000 arbitrary units). Arrowheads, fibril formation reaction was negative for the given protein and GAG. AL-09 H87Y, ␬I O18/O8, AL-09 I34N H87Y, and AL-103 delP95aIns did not form fibrils at pH 7.4 alone or in the presence of GAGs. None of the proteins tested were able to form fibrils in the presence of chondroitin sulfate A under physiological solution conditions, and therefore, differential effects on the t50 values were not included for this GAG.


FIGURE 5. TEM images of AL-12 R65S at the end point of the reaction. A, pH 7.4; B, heparin; C, chondroitin sulfate A; D, heparan sulfate; E, dermatan sulfate; F, dextran sulfate; G, dextran. Scale bar, 200 nm (except where indicated). Shown are fibril formation kinetics at pH 7.4 (control) (H) in the absence (black) or presence (red) of 1.0 mg/ml heparin (I), chondroitin sulfate A (J), heparan sulfate (K), dermatan sulfate (L), dextran sulfate (M), or dextran (N). Kinetic traces are the average of triplicates. Solid lines, fits to the Boltzmann equation. Data were collected for 800 h, but the last data points were discarded due to inner filter effects, as reported previously (38). A.U., arbitrary units. Error bars, S.E.

FIGURE 7. TEM images of AL-103 H92D at the end point of the reaction. A, pH 7.4; B, heparin; C, chondroitin sulfate A; D, heparan sulfate; E, dermatan sulfate; F, dextran sulfate; G, dextran. Scale bar, 200 nm. Shown are fibril formation kinetics at pH 7.4 (control) (H) in the absence (black) or presence (red) of 1.0 mg/ml heparin (I), chondroitin sulfate A (J), heparan sulfate (K), dermatan sulfate (L), dextran sulfate (M), or dextran (N). Kinetic traces are the average of triplicates. Solid lines, fits to the Boltzmann equation. Data were collected for 800 h, but the last data points were discarded due to inner filter effects, as reported previously (38). A.U., arbitrary units. Error bars, S.E.

the GAGs screened. As illustrated in Figs. 4 – 8, TEM images of fibrils at the end of the reaction present different morphologies and degrees of clustering as a function of the protein forming the aggregates and the GAG present in the reaction. In the absence of GAGs at pH 7.4, the AL proteins capable of forming fibrils formed amyloid fibrillar aggregates consistent with the length and morphology of AL protein mature fibrils previously characterized in our laboratory (34, 36, 46). (Figs. 4 –10). The most dramatic effect of GAGs on the fibril morphology was observed for AL-12 and AL-12 R65S. While at pH 7.4, AL-12 forms dense networks of small fibrillar clusters (Fig. FEBRUARY 20, 2015 • VOLUME 290 • NUMBER 8

4A); the presence of heparin promotes the formation of classical long straight fibrils (Fig. 4B), consistent with the large acceleration observed for the AL-12 kinetics of fibril formation. In contrast, for AL-12 R65S, no mature fibrils were found at pH 7.4, in the presence of chondroitin sulfate A and heparan sulfate (Fig. 5, A, C, and D) at the end of the fibril formation reaction, and only short amyloid protofibrils were formed in the presence of dermatan sulfate, dextran, and dextran sulfate (Fig. 5, E–G), consistent with the delay in the fibril formation observed. AL-103 fibrils formed short rods at pH 7.4 (Fig. 6A), whereas we observe clusters of short rods laterally stacked together in JOURNAL OF BIOLOGICAL CHEMISTRY

4959


FIGURE 6. TEM images of AL-103 at the end point of the reaction. A, pH 7.4; B, heparin; C, chondroitin sulfate A; D, heparan sulfate; E, dermatan sulfate; F, dextran sulfate; G, dextran. Scale bar, 200 nm. Shown are fibril formation kinetics at pH 7.4 (control) (H) in the absence (black) or presence (red) of 1.0 mg/ml heparin (I), chondroitin sulfate A (J), heparan sulfate (K), dermatan sulfate (L), dextran sulfate (M), and dextran (N). Kinetic traces are the average of triplicates. Solid lines, fits to the Boltzmann equation. Data were collected for 800 h, but the last data points were discarded due to inner filter effects, as reported previously (38). A.U., arbitrary units. Error bars, S.E.


FIGURE 8. TEM images of AL-09 at the end point of the reaction. A, pH 7.4; B, heparin; C, chondroitin sulfate A; D, heparan sulfate; E, dermatan sulfate; F, dextran sulfate; G, dextran. Scale bar, 200 nm. Shown are fibril formation kinetics at pH 7.4 (control) (H) in the absence (black) or presence (red) of 1.0 mg/ml heparin (I), chondroitin sulfate A (J), heparan sulfate (K), dermatan sulfate (L), dextran sulfate (M), or dextran (N). Kinetic traces are the average of triplicates. Solid lines, fits to the Boltzmann equation. Data were collected for 800 h, but the last data points were discarded due to inner filter effects, as reported previously (38). A.U., arbitrary units. Error bars, S.E.

the presence of GAGs and controls (Fig. 6, B and D–G). In contrast, AL-103 H92D formed individual straight fibrils at pH 7.4 (Fig. 7A), whereas in the presence of heparin it resulted in bundled long fibrils (Fig. 7B) consistent with the acceleration of fibril formation. Chondroitin sulfate A induced the formation of AL-103 H92D spherical particles (Fig. 7C). Dermatan sulfate, dextran, and dextran sulfate induced long and curved fibrils (Fig. 7, E–G). AL-09 did not present any significant difference in the fibril morphology in the presence of GAGs (Fig. 8); in all cases, some individual amyloid fibril segments were observed within the dense network of clustered fibrils that is typical of in vitro AL amyloid fibril samples. AL-09 in the presence of chondroitin sulfate A presented a mixture of spherical intermediates, protofibrillar aggregates, and large networks of clustered fibrils all in the same sample (Fig. 8C). A behavior similar to that observed for AL-09 was found for ␬I Y87H; the presence of GAGs did not affect the morphology of fibrils. In the absence of GAGs, ␬I Y87H formed the classical long, straight fibrils that cluster in networks (Fig. 9). Interestingly, in the presence of chondroitin sulfate A, a mixture of spherical intermediates and large networks of clustered fibrils was observed (Fig. 9C), although the fibril formation reaction presented a 4-fold ThT fluorescence enhancement only in one well. For the most stable proteins, AL-09 H87Y, ␬I O18/O8, AL-09 I34N/H87Y, and AL-103 delP95a, the amyloid forma-


tion reactions are considered negative because no ThT fluorescence enhancement was observed after 800 h of incubation, and only amorphous and protofibrillar material was observed by TEM (data not shown). Fig. 10 summarizes the high polymorphism observed in the aggregates formed in the presence of chondroitin sulfate A for all proteins. All aggregates observed match with the description of early intermediates in the fibril formation pathway (often referred to as protofibrils, oligomers, or prefibrillar species) of amyloidogenic proteins (47), which have non-fibrillar morphology. For AL-09 and ␬I Y87H, different aggregate populations coexisted in the reactions with chondroitin sulfate A. AL-09 formed a mixture of spherical, protofibrillar aggregates and dense fibrillar material (Fig. 10C). Strikingly, for ␬I Y87H, we found mature amyloid fibrils coexisting with spherical oligomeric species (Fig. 10E). This polymorphism can result from alternative aggregation paths from monomer to mature fibrils, but we cannot discard the possibility of off-pathway oligomeric aggregates. Structural Effects of Heparan Sulfate on AL-103 H92D and AL-09 as a Function of Temperature—An interesting observation from the analysis of the fibril formation kinetic traces is that, regardless of the magnitude of the acceleration due to the presence of GAGs, the fibril formation rate of any of the proteins tested at best matches that of AL-09 in presence of dextran or dextran sulfate (t50 ⫽ 152.9 h and t50 ⫽ 184.7 h), and only the VOLUME 290 • NUMBER 8 • FEBRUARY 20, 2015


FIGURE 9. TEM images of ␬I Y87H at the end point of the reaction. A, pH 7.4; B, heparin; C, chondroitin sulfate A; D, heparan sulfate; E, dermatan sulfate; F, dextran sulfate; G, dextran. Scale bar, 200 nm. Shown are fibril formation kinetics at pH 7.4 (control) (H) in the absence (black) or presence (red) of 1.0 mg/ml heparin (I), chondroitin sulfate A (J), heparan sulfate (K), dermatan sulfate (L), dextran sulfate (M), and dextran (N). Kinetic traces are the average of triplicates. Solid lines, fits to the Boltzmann equation. Data were collected for 800 h, but the last data points were discarded due to inner filter effects, as reported previously (38). A.U., arbitrary units. Error bars, S.E.


kinetics of AL-103 H92D in the presence of heparin and heparan sulfate (t50 ⫽ 83.1 h and t50 ⫽ 121.8 h; see Table 2 for all t50 values) come close to the values for AL-09 in the presence of heparan sulfate (t50 ⫽ 30.8 h). These results suggest that AL-09 is the most efficient protein to form fibrils in our study. Moreover, we observed that the initial ThT AL-103 H92D fluorescence intensity showed consistently higher values compared with the other AL proteins (i.e. compare Figs. 4 and 7, H–M), which suggests that at 37 °C, AL-103 H92D is sampling from the beginning of the reaction a different conformational or oligomeric state than the rest of the proteins and that this conformational state is able to bind ThT. AL-103 H92D is less stable (Tm ⫽ 33.6 °C) than AL-09 (Tm ⫽ 41.1 °C). We know from our thermal unfolding experiments that at the conditions employed for the fibril formation assay (37 °C and pH 7.4), the fraction folded estimated for AL-103 H92D was 34.0%, whereas for AL-09 it was 84.6%. The thermal denaturation profiles for AL-103 H92D and AL-09 remained unaffected by the presence of GAGs and the polysaccharide controls (data not shown). In an attempt to determine any possible structural differences between AL-09 and AL-103 H92D caused by the presence of heparan sulfate (accelerating amyloid formation) and chondroitin sulfate A (inhibiting amyloid formation) under fibril formation conditions (37 °C, pH 7.4), far-UV CD spectra were collected at 4 and 37 °C (Fig. 11). We hypothesized that we would detect differences between the effect caused by the different GAGs and the different proteins, specifically if GAGs FEBRUARY 20, 2015 • VOLUME 290 • NUMBER 8

helped to kinetically stabilize a partially folded intermediate. We also wanted to determine whether we could see differences between AL-09 and AL-103 H92D in the presence of GAGs that might explain the high ThT fluorescence observed with AL-103 H92D at the beginning of the reactions. At 4 °C, both proteins were folded as was expected and reported (3, 32), and no change in the spectra occurred in the presence of both heparan sulfate and chondroitin sulfate A. At 37 °C, AL-103 H92D alone showed an increment in the molar ellipticity with a loss of the 230 nm minimum (Fig. 11A), features associated with AL-103 H92D unfolding. Concurrently, AL-09 alone at 37 °C showed a slight increase in molar ellipticity (Fig. 11B). For both proteins alone, the unfolding was fully reversible when the samples were cooled back to 4 °C. In the presence of heparan sulfate, the AL-103 H92D far-UV CD spectrum at 37 °C looked very similar to that of AL-103 H92D alone; however, these conformational changes were only partially reversible (Fig. 11C). In contrast, for AL-09 in the presence of heparan sulfate, the unfolding process was reversible (Fig. 11D). In the presence of chondroitin sulfate A, the far-UV CD spectra for both proteins look different from the spectra at 37 °C with protein alone or in the presence of heparan sulfate; the spectra present a deep minimum at 216 nm and a new minimum around 210 nm. After cooling, the conformational changes observed are fully reversible for AL-09, and although AL-103 H92D does not fully return to the original far-UV CD spectrum at 4 °C, the spectrum resembles the original spectrum of a fully folded protein, although the minimum is shifted to ⬃212 nm. JOURNAL OF BIOLOGICAL CHEMISTRY

4961


FIGURE 10. TEM images of aggregates formed in the presence of chondroitin sulfate A at the end point of the reaction. A, AL-103 H92D; B, AL-103; C, AL-09; D, AL-12; E, ␬I Y87H; F, AL-12 R65S; G, AL-09 H87Y; H, ␬I O18/O8; I, AL-09 I34N/H87Y; J, AL-103 delP95aIns. For ␬I Y87H, a mixture of fibrillar and spherical species was observed. For AL-09, a mixture of fibrillar, amorphous, and spherical chain aggregates were observed. Scale bar, 200 nm (except where indicated). Proteins are shown in order of increasing stability as measured by Tm values, as shown in Table 1.

Light Chains and Glycosaminoglycans TABLE 2 Kinetic parameters of AL proteins in the presence of GAGs

a

Heparin t50

Protein

t50 at pH 7.4 h

h

AL-103 H92D AL-103 AL-09 AL-12 ␬I O18/O8 Y87H AL-12 R65S AL-09 H87Y ␬I O18/O8 AL-09 I34N-H87Y AL-103 delP95aIns

332.5 ⫾ 82.2 172.1 ⫾ 36.7 244.1 ⫾ 14.2 430.0 ⫾ 20.0 256.4 ⫾ 25.6 813.9 ⫾ 13.4 NR NR NR NR

83.1 ⫾ 20.9 236.9 ⫾ 42.2 290.3 ⫾ 26.3 143.6 ⫾ 16.5 306.7 ⫾ 72.7 201.7 ⫾ 54.4 NR NR NR NR

Heparan sulfate t50

Dermatan sulfate t50

h

121.8 ⫾ 33.1 382.3 ⫾ 30.3 30.8 ⫾ 0.7 321.9 ⫾ 40.2 157.2 ⫾ 30.1 NRa NR NR NR NR

h

337.9 ⫾ 11.0 251.8 ⫾ 88.1 265.9 ⫾ 18.4 212.5 ⫾ 47.0 388.0 ⫾ 64.2 475.6 ⫾ 190.7 NR NR NR NR

Dextran sulfate t50 h

258.9 ⫾ 79.2 156.0 ⫾ 5.5 152.9 ⫾ 16.3 141.4 ⫾ 32.4 387.5 ⫾ 78.3 99.0 ⫾ 4.7 NR NR NR NR

Dextran t50 h

154.0 ⫾ 40.7 178.7 ⫾ 29.5 184.7 ⫾ 44.4 172.5 ⫾ 17.0 271.7 ⫾ 54.2 NR NR NR NR NR

NR, negative reaction.

These results suggest that in the case of heparan sulfate, this GAG does not affect the stability of the native state of AL-103 H92D per se but rather accelerates fibril formation by interacting with and stabilizing a partially folded conformation populated at 37 °C that is prone to aggregate. AL-09 has an inherent strong predisposition to form amyloid fibrils on its own, and although the presence of heparan sulfate accelerates amyloid formation, it does not significantly affect the structure of the partially folded conformation found at 37 °C. In the case of chondroitin sulfate A, this GAG did not affect the stability of the native state of either AL-103 H92D (Fig. 11E) or AL-09 (Fig. 11F), but populated a different partially folded state at 37 °C, suggesting that this GAG promotes the conformation of a different partially folded state. During our short incubation period, both AL-09 and AL-103 H92D recovered all or most of the folded state properties that they presented when initially analyzed at 4 °C. However, we propose that long term incubations at 37 °C like the ones happening during amyloid formation reactions may cause irreversibility in the reaction for both proteins.


DISCUSSION In this study, we have shown that Heparan sulfate promotes the in vitro fibril formation reaction of AL amyloidogenic light chains, supporting the notion that GAGs interact with the amyloidogenic fibril precursor and enhance their deposition in vivo. To address the role that the multiple negative charges in GAGs play in this process, we compared the effect of GAGs with those of other large non-GAG polysaccharides (charged and neutral). Although the effects of dextran and dextran sulfate have no biological relevance in the context of AL amyloidosis, it has been reported that the physicochemical behavior of GAGs is similar to that of polysaccharides in their role as polyelectrolytes (48). Interestingly, among all GAGs examined, chondroitin sulfate A exhibited a strong inhibitory effect on the fibril formation reaction. As indicated in Fig. 1, the numbers of carboxyl and sulfate groups per disaccharide unit are 1 and 3, 1 and 2, 1 and 1, and 1 and 1, for heparin, heparan sulfate, dermatan sulfate, and VOLUME 290 • NUMBER 8 • FEBRUARY 20, 2015


FIGURE 11. Irreversible changes caused by partial unfolding in the presence of heparan sulfate in AL-103 H92D. AL-103 H92D and AL-09 secondary structure in the absence (A and B) or presence of 1.0 mg ml⫺1 heparan sulfate (C and D) or chondroitin sulfate A (E and F), followed by far-UV CD spectra. 20 ␮M protein was incubated at 4 °C (black open circles) for 10 min and then incubated at 37 °C (black solid circles) for 10 min and finally incubated at 4 °C (red solid circles).


chondroitin sulfate A, respectively. Because GAGs present differences in the sulfate groups while the carboxyl groups remain constant, we propose that the accelerating effect on amyloid formation observed with AL proteins (Fig. 3) may be roughly correlated with the degree of sulfation present in the GAGs. The GAG that accelerates most reactions (heparan sulfate) is one of the most charged GAGs, whereas chondroitin sulfate showed inhibitory effects and is the least charged GAG of these series. The biophysical and amyloidogenic properties of AL-09, AL-12, AL-103, ␬I O18/O8, and their mutants have been extensively studied by our laboratory at different pH values. We have demonstrated that these proteins are able to dimerize in different conformations that interconvert in a dynamic way. Some of these dimer conformations are highly amyloidogenic, as is the case for AL-09 and, to a lesser extent, ␬I Y87H. The AL-09 altered dimer interface is rotated 90° with respect to the ␬I O18/O8 (non-amyloidogenic) canonical dimer interface (34) and facilitates misfolding events that trigger amyloid formation, whereas AL-12 and AL-103 present crystal structures of canonical dimers (10), thus requiring stochastic conformational fluctuations to populate amyloid precursors (37). For AL-103 (amyloidogenic) and its restorative mutants, the unfolding/refolding pathway is kinetically determined by proFEBRUARY 20, 2015 • VOLUME 290 • NUMBER 8

line 95a cis-trans prolyl isomerization, which can enhance the population of a transient kinetic intermediate prone to fibril formation (33). The in vitro AL amyloid fibril formation can be described, in general, by a nucleation-elongation mechanism modulated by the presence of GAGs (Fig. 12). In the presence of GAGs, in vitro fibril formation of AL proteins can be broadly classified into two different mechanisms/ scenarios, which will be referred to as mechanism/scenario A and B. Mechanism/scenario A is followed by AL light chains that favor the canonical dimer interface (AL-103 H92D, AL-103, AL-12, and AL-12R65S, most of them amyloidogenic and mildly amyloidogenic). In this case, AL light chains sample and populate the altered state (A in Fig. 12) via stochastic conformational fluctuations, which eventually interact with the negative charged GAGs, probably via electrostatic interactions (29). We hypothesize that heparan sulfate increases the population of the fibril precursor, depicted here as a protofibril (P), possibly by increasing the stability of the altered state (A), promoting the accumulation of protofibrils (P), which facilitates the extension of the fibril (F) by the addition of AL light chain monomers. In the presence of chondroitin sulfate A, as depicted in scenario B, our amyloid kinetics data suggest that the altered state is somehow trapped in a conformation in which fibril elongaJOURNAL OF BIOLOGICAL CHEMISTRY

4963


FIGURE 12. A–C, suggested model of GAG mode of action on AL amyloid fibril formation. D, representative morphology of aggregated species observed by TEM. The morphological designations/definitions were based on empirical/visual comparison and contrasted with the description of amyloid aggregates and intermediates in the fibril formation pathway, reported for AL proteins as well as other proteins (47).



tions, are required to promote the nucleation process and therefore accelerate the fibril formation reaction. This explanation is supported by our previous findings (31) and by recent findings with the amyloid precursor protein, serum amyloid A protein (50). Our results are very relevant to the study of AL amyloidosis pathophysiology because AL-09, AL-12, and AL-103 amyloid deposition occurred in cardiac tissue. The differential effects of GAGs presented here help us to rationalize why cardiac tissue is the major site for AL amyloid deposition in these amyloidosis patients. To the best of our knowledge, heparan sulfate is the main sulfated GAG present in AL amyloidosis cardiac tissue based on the report by Ohishi et al. (26) that found heparan sulfate in ex vivo cardiac AL amyloid deposits but not in carpal synovium, where chondroitin sulfate A is abundant. We hypothesize that differences in the GAG content of different tissues modulate the localization of AL amyloid aggregates in specific tissues. In conclusion, we propose that highly sulfated GAGs bind to partially folded, aggregation-prone AL protein species, which populate and promote/enhance fibril formation. Low sulfated GAGs kinetically trap AL proteins into prefibrillar states inhibiting amyloid formation. However, a lack of information about the exact chemical nature of these compounds due to their animal origin as well as the heterogeneity in commercial preparation available to our studies hampers the ability to make better comparisons between the in vivo and in vitro amyloidogenic processes. Future studies with model compounds for all of the GAGs would be advantageous. Acknowledgments—We are grateful for the generosity of amyloidosis patients and their families. We thank Alexander Tischer for critical reading of the manuscript.

REFERENCES 1. Merlini, G., and Stone, M. J. (2006) Dangerous small B-cell clones. Blood 108, 2520 –2530 2. Kumar, S. K., Dispenzieri, A., Lacy, M. Q., Hayman, S. R., Buadi, F. K., Zeldenrust, S. R., Tan, T., Sinha, S., Leung, N., Kyle, R. A., Rajkumar, S. V., and Gertz, M. A. (2011) Changes in serum-free light chain rather than intact monoclonal immunoglobulin levels predicts outcome following therapy in primary amyloidosis. Am. J. Hematol. 86, 251–255 3. Blancas-Mejía, L. M., and Ramirez-Alvarado, M. (2013) Systemic amyloidoses. Annu. Rev. Biochem. 82, 745–774 4. Abraham, R. S., Geyer, S. M., Price-Troska, T. L., Allmer, C., Kyle, R. A., Gertz, M. A., and Fonseca, R. (2003) Immunoglobulin light chain variable (V) region genes influence clinical presentation and outcome in light chain-associated amyloidosis (AL). Blood 101, 3801–3808 5. Janeway, C. A. (2001) How the immune system works to protect the host from infection: a personal view. Proc. Natl. Acad. Sci. U.S.A. 98, 7461–7468 6. Dispenzieri, A., Gertz, M. A., and Buadi, F. (2012) What do I need to know about immunoglobulin light chain (AL) amyloidosis? Blood Rev. 26, 137–154 7. Falk, R. H. (2005) Diagnosis and Management of the cardiac amyloidoses. Circulation 112, 2047–2060 8. Dubrey, S. W., Cha, K., Anderson, J., Chamarthi, B., Reisinger, J., Skinner, M., and Falk, R. H. (1998) The clinical features of immunoglobulin lightchain (AL) amyloidosis with heart involvement. QJM 91, 141–157 9. Kyle, R. A., and Gertz, M. A. (1995) Primary systemic amyloidosis: clinical and laboratory features in 474 cases. Semin. Hematol. 32, 45–59 10. Randles, E. G., Thompson, J. R., Martin, D. J., and Ramirez-Alvarado, M.

VOLUME 290 • NUMBER 8 • FEBRUARY 20, 2015


tion is inhibited, resulting in the stabilization of non-productive (P*) spherical, oligomeric, and protofibrillar aggregates. We propose that P* may be an early intermediate on the fibril formation pathway that may be kinetically trapped in this scenario. A certain amount of canonical dimers are populated in equilibrium with the altered dimers (AL-09 and ␬I Y87H, considered highly amyloidogenic), which results in two populations with different rates of fibril formation. This is more evident in the presence of chondroitin sulfate A, where the alternative dimer interface population easily forms protofibrils (P in Fig. 12) following the normal nucleation-elongation pathway, forming mature fibrils (F), whereas the canonical dimer populates non-productive spherical protofibrillar intermediates (P*) with chondroitin sulfate A. Mechanism/scenario C is a special case observed in AL proteins that preferentially populate amyloidogenic altered dimers (AL-09), as represented by the arrows (Fig. 12). The altered dimer facilitates misfolding events that trigger amyloid formation. In this case, the presence of GAGs has a minimal effect on accelerating fibril formation because the equilibrium is already shifted to the altered conformation (A). From an inspection of the morphology observed by TEM images, it is clear that aggregates formed in presence of Chondroitin sulfate A are characterized by a remarkable polymorphism among the different light chain proteins tested in this study, indicating that the morphology of the oligomeric protofibrillar aggregates formed is highly dependent on the protein properties. We propose that all of these species may be part of the amyloid formation pathway, and depending on the protein properties, some may be more populated than others (Fig. 12D). Although there are extensive reports of the toxicity of soluble oligomers for other amyloid precursor protein systems (49), a systematic study comparing the toxicity of soluble proteins, soluble oligomers, and mature fibrils formed by AL proteins has not been conducted to date. It is worth noting that thermodynamically stable light chains that at 37 °C are 100% folded (␬I O18/O8, AL-09 H87Y, AL-09 I34N H87Y, and AL-103 delP95aIns) resist fibril formation even in the presence of GAGs. As we mentioned earlier, neither GAG substantially perturbs the native structure of the AL light chains studied or affects their thermodynamic stability. The exception is AL-12 (mildly amyloidogenic), presenting an increment on the Tm values in the presence of GAGs (heparin ⌬Tm ⫽ 3.2 °C, dermatan sulfate ⌬Tm ⫽ 5.4 °C, chondroitin sulfate ⌬Tm ⫽ 5.6 °C, and heparan sulfate ⌬Tm ⫽ 8.4 °C) but not in the presence of the polysaccharide controls dextran or dextran sulfate (⌬Tm ⫽ 0.1 °C) (data not shown). Interestingly, AL-12 presents one of the most dramatic accelerations of fibril formation due to the presence of GAGs in our studies. However, we found that heparan sulfate affects the refolding reaction for AL-103 H92D (amyloidogenic), probably trapping the partially folded state formed at 37 °C. Chondroitin sulfate A affected the partially folded state at 37 °C but did not affect the refolding of AL-09 (highly amyloidogenic) and AL-103 H92D (amyloidogenic) within our experimental setup (Fig. 11). Given that GAGs did not induce conformational changes in the native state, we conclude that the interactions between GAGs and transient intermediates, probably partially folded conforma-


11.

12.

13.

14.

16.

17. 18.

19. 20. 21. 22. 23. 24. 25.

26.

27.

28.

29.

30.

31.

32.


33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48. 49.

50.

Ramirez-Alvarado, M. (2008) Structural insights into the role of mutations in amyloidogenesis. J. Biol. Chem. 283, 30950 –30956 Blancas-Mejıá, L. M., Tischer, A., Thompson, J. R., Tai, J., Wang, L., Auton, M., and Ramirez-Alvarado, M. (2014) Kinetic control in protein folding for light chain amyloidosis and the differential effects of somatic mutations. J. Mol. Biol. 426, 347–361 Baden, E. M., Owen, B. A., Peterson, F. C., Volkman, B. F., Ramirez-Alvarado, M., and Thompson, J. R. (2008) Altered dimer interface decreases stability in an amyloidogenic protein. J. Biol. Chem. 283, 15853–15860 Sikkink, L. A., and Ramirez-Alvarado, M. (2008) Salts enhance both protein stability and amyloid formation of an immunoglobulin light chain. Biophys. Chem. 135, 25–31 DiCostanzo, A. C., Thompson, J. R., Peterson, F. C., Volkman, B. F., and Ramirez-Alvarado, M. (2012) Tyrosine residues mediate fibril formation in a dynamic light chain dimer interface. J. Biol. Chem. 287, 27997–28006 Peterson, F. C., Baden, E. M., Owen, B. A., Volkman, B. F., and RamirezAlvarado, M. (2010) A single mutation promotes amyloidogenicity through a highly promiscuous dimer interface. Structure 18, 563–570 Martin, D. J., and Ramirez-Alvarado, M. (2010) Comparison of amyloid fibril formation by two closely related immunoglobulin light chain variable domains. Amyloid 17, 129 –136 Sreerama, N., Manning, M. C., Powers, M. E., Zhang, J.-X., Goldenberg, D. P., and Woody, R. W. (1999) Tyrosine, phenylalanine, and disulfide contributions to the circular dichroism of proteins: circular dichroism spectra of wild-type and mutant bovine pancreatic trypsin inhibitor. Biochemistry 38, 10814 –10822 Albinsson, B., and Norden, B. (1992) Excited-state properties of the indole chromophore: electronic transition moment directions from linear dichroism measurements: effect of methyl and methoxy substituents. J. Phys. Chem. 96, 6204 – 6212 Hu, D., Qin, Z., Xue, B., Fink, A. L., and Uversky, V. N. (2008) Effect of methionine oxidation on the structural properties, conformational stability, and aggregation of immunoglobulin light chain LEN. Biochemistry 47, 8665– 8677 Souillac, P. O., Uversky, V. N., Millett, I. S., Khurana, R., Doniach, S., and Fink, A. L. (2002) Effect of association state and conformational stability on the kinetics of immunoglobulin light chain amyloid fibril formation at physiological pH. J. Biol. Chem. 277, 12657–12665 Souillac, P. O., Uversky, V. N., and Fink, A. L. (2003) Structural transformations of oligomeric intermediates in the fibrillation of the immunoglobulin light chain LEN. Biochemistry 42, 8094 – 8104 Khurana, R., Gillespie, J. R., Talapatra, A., Minert, L. J., Ionescu-Zanetti, C., Millett, I., and Fink, A. L. (2001) Partially folded intermediates as critical precursors of light chain amyloid fibrils and amorphous aggregates. Biochemistry 40, 3525–3535 Souillac, P. O., Uversky, V. N., Millett, I. S., Khurana, R., Doniach, S., and Fink, A. L. (2002) Elucidation of the molecular mechanism during the early events in immunoglobulin light chain amyloid fibrillation: evidence for an off-pathway oligomer at acidic pH. J. Biol. Chem. 277, 12666 –12679 Blancas-Mejia, L. M., Tellez, L. A., del Pozo-Yauner, L., Becerril, B., Sanchez-Ruiz, J. M., and Fernandez-Velasco, D. A. (2009) Thermodynamic and kinetic characterization of a germ line human ␭6 light-chain protein: the relation between unfolding and fibrillogenesis. J. Mol. Biol. 386, 1153–1166 Fauerbach, J. A., Yushchenko, D. A., Shahmoradian, S. H., Chiu, W., Jovin, T. M., and Jares-Erijman, E. A. (2012) Supramolecular non-amyloid intermediates in the early stages of ␣-synuclein aggregation. Biophys. J. 102, 1127–1136 Seyrek, E., and Dubin, P. (2010) Glycosaminoglycans as polyelectrolytes. Adv. Colloid Interface Sci. 158, 119 –129 Guerrero-Munõz, M. J., Castillo-Carranza, D. L., and Kayed, R. (2014) Therapeutic approaches against common structural features of toxic oligomers shared by multiple amyloidogenic proteins. Biochem. Pharmacol. 88, 468 – 478 Aguilera, J. J., Zhang, F., Beaudet, J. M., Linhardt, R. J., and Coloń, W. (2014) Divergent effect of glycosaminoglycans on the in vitro aggregation of serum amyloid A. Biochimie 104, 70 – 80

JOURNAL OF BIOLOGICAL CHEMISTRY

4965


15.

(2009) Structural alterations within native amyloidogenic immunoglobulin light chains. J. Mol. Biol. 389, 199 –210 Wall, J. S., Gupta, V., Wilkerson, M., Schell, M., Loris, R., Adams, P., Solomon, A., Stevens, F., and Dealwis, C. (2004) Structural basis of light chain amyloidogenicity: comparison of the thermodynamic properties, fibrillogenic potential and tertiary structural features of four V␭6 proteins. J. Mol. Recognit. 17, 323–331 Pokkuluri, P. R., Gu, M., Cai, X., Raffen, R., Stevens, F. J., and Schiffer, M. (2002) Factors contributing to decreased protein stability when aspartic acid residues are in ␤-sheet regions. Protein Sci. 11, 1687–1694 Hernańdez-Santoyo, A., del Pozo Yauner, L., Fuentes-Silva, D., Ortiz, E., Rudinõ-Pin˜era, E., Sańchez-Lo´pez, R., Horjales, E., Becerril, B., and Rodrı´guez-Romero, A. (2010) A single mutation at the sheet switch region results in conformational changes favoring ␭6 light-chain fibrillogenesis. J. Mol. Biol. 396, 280 –292 Schormann, N., Murrell, J. R., Liepnieks, J. J., and Benson, M. D. (1995) Tertiary structure of an amyloid immunoglobulin light chain protein: a proposed model for amyloid fibril formation. Proc. Natl. Acad. Sci. U.S.A. 92, 9490 –9494 Hurle, M. R., Helms, L. R., Li, L., Chan, W., and Wetzel, R. (1994) A role for destabilizing amino acid replacements in light-chain amyloidosis. Proc. Natl. Acad. Sci. U.S.A. 91, 5446 –5450 Wall, J., Schell, M., Murphy, C., Hrncic, R., Stevens, F. J., and Solomon, A. (1999) Thermodynamic instability of human ␭ 6 light chains: correlation with fibrillogenicity. Biochemistry 38, 14101–14108 Wetzel, R. (1997) Domain stability in immunoglobulin light chain deposition disorders. Adv. Protein Chem. 50, 183–242 Poshusta, T. L., Katoh, N., Gertz, M. A., Dispenzieri, A., and RamirezAlvarado, M. (2013) Thermal stability threshold for amyloid formation in light chain amyloidosis. Int. J. Mol. Sci. 14, 22604 –22617 Buxbaum, J. N., and Linke, R. P. (2012) A molecular history of the amyloidoses. J. Mol. Biol. 421, 142–159 Stevens, F. J., and Kisilevsky, R. (2000) Immunoglobulin light chains, glycosaminoglycans, and amyloid. Cell. Mol. Life Sci. 57, 441– 449 Bishop, J. R., Schuksz, M., and Esko, J. D. (2007) Heparan sulphate proteoglycans fine-tune mammalian physiology. Nature 446, 1030 –1037 Shi, X., and Zaia, J. (2009) Organ-specific heparan sulfate structural phenotypes. J. Biol. Chem. 284, 11806 –11814 Varki, A. (2006) Nothing in glycobiology makes sense, except in the light of evolution. Cell 126, 841– 845 Bosman, F. T., and Stamenkovic, I. (2003) Functional structure and composition of the extracellular matrix. J. Pathol. 200, 423– 428 Nelson, S. R., Lyon, M., Gallagher, J. T., Johnson, E. A., and Pepys, M. B. (1991) Isolation and characterization of the integral glycosaminoglycan constituents of human amyloid A and monoclonal light-chain amyloid fibrils. Biochem. J. 275, 67–73 Ohishi, H., Skinner, M., Sato-Araki, N., Okuyama, T., Gejyo, F., Kimura, A., Cohen, A. S., and Schmid, K. (1990) Glycosaminoglycans of the hemodialysis-associated carpal synovial amyloid and of amyloid-rich tissues and fibrils of heart, liver, and spleen. Clin. Chem. 36, 88 –91 Jiang, X., Myatt, E., Lykos, P., and Stevens, F. J. (1997) Interaction between glycosaminoglycans and immunoglobulin light chains. Biochemistry 36, 13187–13194 Trinkaus-Randall, V., Walsh, M. T., Steeves, S., Monis, G., Connors, L. H., and Skinner, M. (2005) Cellular response of cardiac fibroblasts to amyloidogenic light chains. Am. J. Pathol. 166, 197–208 Martin, D. J., and Ramirez-Alvarado, M. (2011) Glycosaminoglycans promote fibril formation by amyloidogenic immunoglobulin light chains through a transient interaction. Biophys. Chem. 158, 81– 89 Ren, R., Hong, Z., Gong, H., Laporte, K., Skinner, M., Seldin, D. C., Costello, C. E., Connors, L. H., and Trinkaus-Randall, V. (2010) Role of glycosaminoglycan sulfation in the formation of immunoglobulin light chain amyloid oligomers and fibrils. J. Biol. Chem. 285, 37672–37682 McLaughlin, R. W., De Stigter, J. K., Sikkink, L. A., Baden, E. M., and Ramirez-Alvarado, M. (2006) The effects of sodium sulfate, glycosaminoglycans, and Congo red on the structure, stability, and amyloid formation of an immunoglobulin light-chain protein. Protein Sci. 15, 1710 –1722 Baden, E. M., Randles, E. G., Aboagye, A. K., Thompson, J. R., and

Differential effects of glycation on protein aggregation and amyloid formation.

Formation of amyloid fibers by monomeric light chain variable domains.

Thermal stability threshold for amyloid formation in light chain amyloidosis.

The effect of glycosaminoglycans (GAGs) on amyloid aggregation and toxicity.

Effects of Charged Cholesterol Derivatives on Aβ40 Amyloid Formation.

Recurrent mutations of MAPK pathway genes in multiple myeloma but not in amyloid light-chain amyloidosis.

Does ear keloid formation depend on the type of earrings or piercing jewellery?

The burden of amyloid light chain amyloidosis on health-related quality of life.

Differential regulation of amyloid precursor protein sorting with pathological mutations results in a distinct effect on amyloid-β production.

Repetition probability effects depend on prior experiences.

Proteoglycans, glycosaminoglycans, amyloid-enhancing factor, and amyloid deposition.

Update on treatment of light chain amyloidosis.

Light Chain Amyloid Fibrils Cause Metabolic Dysfunction in Human Cardiomyocytes.

Effects of polyamino acids and polyelectrolytes on amyloid β fibril formation.

Effects of acidic oligosaccharide sugar chain on amyloid oligomer-induced impairment of synaptic plasticity in rats.

Inhibition by small-molecule ligands of formation of amyloid fibrils of an immunoglobulin light chain variable domain.

Differential effects of amyloid-β peptide aggregation status on in vivo retinal neurotoxicity.

Effects of high glucose on human umbilical vein endothelial cell permeability and myosin light chain phosphorylation.

Genetic activity of bleomycin: differential effects on mitotic recombination and mutations in yeast.

Beta-amyloid associated differential effects of APOE ε4 on brain metabolism in cognitively normal elderly.

Interpreting Alzheimer's disease clinical trials in light of the effects on amyloid-β.

Differential inhibitory effects of isosorbide dinitrate and its mononitrate metabolites on platelet aggregation and thromboxane formation.

Inhibitory effects of β-ionone on amyloid fibril formation of β-lactoglobulin.

Effects of light and temperature on macrocyst formation in paired mating types of Dictyostelium discoideum.