Suppression of apolipoprotein C-II amyloid formation by the extracellular chaperone, clusterin Danny M. Hatters 1 , Mark R. Wilson 2 , Simon B. Easterbrook-Smith 3 and Geoffrey J. Howlett 1 1 Department of Biochemistry and Molecular Biology, The University of Melbourne, Victoria, Australia; 2 Department of Biological Sciences, University of Wollongong, NSW, Australia; 3 School of Molecular and Microbial Biosciences, University of Sydney, NSW, Australia The effect of the extracellular chaperone, clusterin, on amyloid fibril formation by lipid-free human apolipoprotein C-II (apoC-II) was investigated. Sub-stoichiometric levels of clusterin, derived from either plasma or semen, potently inhibit amyloid formation by apoC-II. Inhibition is dependent on apoC-II concentration, with more effective inhibition by clusterin observed at lower concentrations of apoC-II. The average sedimentation coefficient of apoC-II fibrils formed from apoC-II (0.3 mgÆmL )1 ) is reduced by coincubation with clusterin (10 lgÆmL )1 ). In contrast, addition of clusterin (0.1 mgÆmL )1 ) to preformed apoC-II amyloid fibrils (0.3 mgÆmL )1 ) does not affect the size distribution after 2 days. This sedimentation velocity data suggests that clusterin inhibits fibril growth but does not promote fibril dissociation. Electron micrographs indicate similar morphologies for amyloid fibrils formed in the presence or absence of clusterin. The substoichiometric nature of the inhibition suggests that clusterin interacts with transient amyloid nuclei leading to dissociation of the monomeric subunits. We propose a general role for clusterin in suppressing the growth of extracellular amyloid. Keywords: analytical ultracentrifugation; amyloid forma- tion; nucleation; substoichiometric inhibition; protein folding. Human apoC-II (apoC-II) is mostly associated with plasma very-low-density lipoproteins where it plays an important physiological role as an activator of lipoprotein lipase [1,2]. When associated with polar lipids (e.g. phospholipids or SDS) apoC-II adopts a primarily a helical conformation [3–5]. However, under lipid-free conditions, apoC-II self- associates in a time- and concentration-dependent manner to form twisted ribbon-like fibrils with all the hallmarks of amyloid [6]. These characteristics include the binding to Congo Red with red-green birefringence under cross- polarized light, binding to thioflavin T and increased b structure [6]. The limited conformational stability of apoC-II in the absence of lipid is typical of lipid-free apolipoproteins and may underlie the general propensity of apolipoproteins to form amyloid in vivo [7]. The ability of apoC-II to form amyloid in vitro can be compared to in vivo amyloid formation involving many of the exchangeable apolipoproteins, including apoA-I [8,9], apoA-II [10,11], apoA-IV [12], apoE [13], and apolipoprotein–like proteins, a-synuclein [14] and serum amyloid A [15]. Amyloid formation by apoC-II provides a convenient model to explore physiological parameters that modulate amyloid formation in vivo. Previous studies suggest that the chaperones Hsp27 and aB-crystallin suppress Ab amyloid formation in vitro [16,17]. Our recent studies indicate that a-crystallin inhibits amyloid formation at the nucleation phase of amyloid formation, with no evidence of binding to the apoC-II monomer or suppression of fibril elongation [18]. In view of the extracellular location of amyloid deposits, we investigated whether amyloid formation by apoC-II is affected by the extracellular chaperone, clusterin [19,20]. This protein is a disulfide-linked heterodimer, with an average molecular mass of 75–80 kDa [19], and is present in many biological tissues, with particularly high concentrations present in plasma (35–105 lgÆmL )1 )and seminal fluid ( 400 lgÆmL )1 ) [21,22]. Clusterin is also found in many amyloid-related lesions associated with diseases including Alzheimer’s and atherosclerosis [19,23,24]. Clusterin binds to the Alzheimer’s related peptide Ab with nanomolar affinity, and inhibits its aggregation [25]. Qualitative studies also point to an inhibitory effect of clusterin on amyloid formation by Ab, and a fragment of the prion protein [26,27]. In this investigation, we explore the effect of clusterin on amyloid formation by apoC-II and provide evidence that clusterin acts at substoichiometric levels to inhibit amyloid nuclea- tion and fibril extension. MATERIALS AND METHODS Materials ApoC-II was expressed [28] and purified as described previously [6,28]. ApoC-II was stored as a stock in 5 M Correspondence to G. J. Howlett, Department of Biochemistry and Molecular Biology, The University of Melbourne, Victoria 3010, Australia. Fax: + 61 39347 7730. Tel.: + 61 3 8344 7632, E-mail: ghowlett@unimelb.edu.au Abbreviations: apo, apolipoprotein; GdnHCl, guanidine hydrochloride; rmsd, root mean squared deviation. Note: a website is available at http://www.biochemistry.unimelb.edu.au/bch/research/ ghowlett_rp.htm. (Received 13 December 2001, revised 19 April 2002, accepted 25 April 2002) Eur. J. Biochem. 269, 2789–2794 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02957.x guanidine hydrochloride (GdnHCl) at a concentration of 40 mgÆmL )1 . Clusterin was purified from human serum or from human seminal fluid by immunoaffinity chromatog- raphy as described previously [29]. Clusterin was stored at a concentration of 1–1.4 mgÆmL )1 in refolding buffer (10 m M sodium phosphate, 150 m M NaCl, 0.1% sodium azide, pH 7.4) and kept at )20 °C until required. Experiments were performed at 20 °C unless otherwise indicated. Amyloid formation followed by thioflavin T reactivity To follow time-dependent amyloid formation, apoC-II was refolded by direct dilution into refolding buffer (0.3– 1mgÆmL )1 ) containing various concentrations of clusterin or BSA. Control samples containing clusterin or BSA alone were prepared in refolding buffer. GdnHCl was added to provide the same concentration in all samples (160 m M ). In a microplate, aliquots containing 15 lg apoC-II, or 5 lg protein for samples of BSA and clusterin alone, were added to final solution volumes of 300 lL containing refolding buffer and 5 l M thioflavin T. The fluorescence was monitored using an f max fluorescence plate reader with a 444/485 nm excitation/emission filter set. Analytical ultracentrifugation Sample volumes of 300–400 lL were analyzed using the XL-A analytical ultracentrifuge. Radial scans were taken in continuous scanning mode and 0.002 cm radial increments. For samples containing apoC-II aggregates, the sedimenta- tion boundaries at different time points were analyzed to obtain ls ) g*(s) sedimentation coefficient distributions. The method is based on direct linear least-squares boundary modeling by a superposition of sedimentation profiles of ideal nondiffusing particles [30]. A regularization parameter of p ¼ 0.95, and a buffer density of 1.01 gÆmL )1 was used. The nonsedimenting baselines at a rotor speed of 6000 r.p.m. (2600 g), attributed to the presence of monomeric apoC-II, was subtracted from the data by fitting a time-independent absorbance background. A partial specific volume of 0.73 mLÆg )1 was assumed based on the amino-acid compo- sition of apoC-II. For samples of freshly prepared apoC-II and for the slow moving boundary observed for incubated apoC-II samples and mixtures of apoC-II and clusterin, data were fitted to a model assuming a continuous size distribu- tion [31]. For this analysis, the frictional ratio (f/f 0 )was varied to give the lowest root mean squared deviation (rmsd) for the sample containing freshly prepared apoC-II alone. This best-fit value (f/f 0 ¼ 1.75) was constrained for the analyses of samples containing clusterin. A regularization parameter of p ¼ 0.68 was used. Electron microscopy Solutions of apoC-II (0.3 mgÆmL )1 ) or clusterin (0.1 mgÆmL )1 ) or mixtures containing both proteins, were dilutedthreefoldinwaterandappliedtofreshlyglow- discharged carbon-coated copper grids. After 1 min, excess material was removed and the grids were washed twice with 20 lL water before negatively staining with 2% (w/v) potassium phosphotungstate. The samples were imaged using a JEOL 2000 transmission electron microscope (Peabody, MA, USA) operating at 120 kV. RESULTS Sub-stoichiometric concentrations of clusterin inhibit amyloid formation Thioflavin T has low fluorescence in the presence of monomeric apoC-II that increases proportionally to the amount of apoC-II amyloid present [6,18]. This change in fluorescence was used to monitor amyloid formation by apoC-II in the presence of various concentrations of clusterin (Fig. 1). For apoC-II alone, amyloid formation occurs slowly over 9 days at room temperature. The presence of increasing concentrations of serum clusterin systematically reduces the accumulation of thioflavin T reactive material, with near complete suppression of amy- loid formation by 30 lgÆmL )1 clusterin (Fig. 1A). Clusterin alone (0.1 mgÆmL )1 ) shows no change in thioflavin T reactivity over the same time course (data not shown). Assuming a molecular mass of 8900 Da for apoC-II and 80 000 Da for clusterin, 0.3 mgÆmL )1 apoC-II in the presence of 30 lgÆmL )1 clusterin represents a 90 : 1 molar excess of apoC-II to clusterin. This stoichiometry of inhibition is similar to the concentrations of a-crystallin required to inhibit amyloid formation by apoC-II at a Fig. 1. Time-dependent changes in amyloid formation by apoC-II (0.3 mgÆmL )1 ). Amyloid formation was monitored using thioflavin T fluorescence measurements for samples of apoC-II alone and in the presence of serum clusterin (A) or seminal clusterin (B). ApoC-II alone (d) and apoC-II in the presence of 1 lgÆmL )1 (,), 10 lgÆmL )1 (j), 30 lgÆmL )1 (e), and 100 lgÆmL )1 (m) clusterin. 2790 D. M. Hatters et al. (Eur. J. Biochem. 269) Ó FEBS 2002 concentration of 0.3 mgÆmL )1 [18]. Similar levels of inhibition were observed using seminal clusterin, with 0.1 mgÆmL )1 providing complete inhibition of apoC-II amyloid formation, equivalent to a 30 : 1 molar excess of apoC-II to clusterin (Fig. 1B). The specificity of the inhibition of amyloid formation by clusterin was assessed in control experiments using BSA. The presence of BSA (100 lgÆmL )1 ) had negligible effect on the time-dependent formation of thioflavin T reactivity of 0.3 mgÆmL )1 apoC-II while BSA alone produced no changes in thioflavin T reactivity over the same time (data not shown). The substoichiometric concentrations of clusterin required to suppress amyloid formation suggests that the process is inhibited at the nucleation phase of fibril growth. To investigate the kinetics in more detail, we performed thioflavin T binding assays using a fixed molar ratio of apoC-II to serum clusterin (90 : 1) but varied the total protein concentration. Figure 2 shows the effects of varying the concentration of apoC-II (0.3, 0.6 and 1.0 mgÆmL )1 ). ApoC-II alone shows an increase in the rate of amyloid formation as the apoC-II concentration is increased from 0.3 to 0.6 and 1 mgÆmL )1 (Fig. 2). This accelerated rate of amyloid formation is consistent with previous turbidity studies [6]. Inhibition of apoC-II amyloid formation using a fixed molar ratio of clusterin (1 : 90 ratio of clusterin: apoC-II) is more complete at low apoC-II concentrations (0.3 mgÆmL )1 ) compared to the inhibition observed at the higher concentrations of apoC-II. Amyloid fibril size is reduced by the presence of clusterin The effect of clusterin on the aggregation-state of apoC-II amyloid was investigated by sedimentation velocity analysis. Sedimentation velocity data for apoC-II alone (0.3 mgÆmL )1 , incubated 9 days) reveals a fast-sedimenting boundary comprising  65% of the absorbance (Fig. 3A). This fast moving boundary is attributed to the presence of high molecular mass amyloid fibrils [4,6]. Increasing the rotor speed to 40 000 r.p.m. (116 200 g), revealed a slower moving boundary. Analysis of this boundary yielded good fits to a model describing a single sedimenting species with a sedimentation coefficient of  1 S, and a molecular mass 10 000 Da. This is close to the expected mobility and molecular mass of monomeric apoC-II [4,6]. The presence of a distribution of large sedimenting species (centred at  400 S) and monomeric apoC-II is consistent with a bimodal population of high molecular mass amyloid and monomers as previously observed [4,6,18]. The sedimentation velocity profile of 0.3 mgÆmL )1 apoC-II incubated for 9 days in the presence of 10 lgÆmL )1 serum clusterin is shown in Fig. 3B. The sedimentation data shows a fast sedimenting boundary comprising  45% of the absorbance. The absorbance contribution of clusterin is negligible at these concentrations. At higher angular velo- cities (40 000 r.p.m./116 200 g) the fast moving boundary sedimented to the bottom of the cell revealing a slower moving boundary indicative of predominately monomeric apoC-II. It is noteworthy that the proportion of the fast moving material relative to monomeric apoC-II is reduced by the presence of clusterin (Fig. 3B compared to Fig. 3A), consistent with the conclusion drawn from the data in Figs 1 and 2 that clusterin inhibits apoC-II aggregation. Fig. 2. Concentration dependence of amyloid formation by apoC-II in the absence and presence of a fixed molar ratio of serum clusterin (90 : 1 molar ratio, apoC-II/clusterin). Data for apoC-II alone at 0.3 mgÆmL )1 (d), 0.6 mgÆmL )1 (r)and1.0mgÆmL )1 (j). Data for apoC-II in the presence of clusterin at apoC-II and clusterin concentrations of 0.3 mgÆmL )1 and 0.03 mgÆmL )1 (s), 0.6 mgÆmL )1 and 0.06 mgÆmL )1 (e)and1.0mgÆmL )1 and 0.1 mgÆmL )1 (h), respectively. Fig. 3. Sedimentation velocity behaviour of apoC-II (0.3 mgÆmL )1 ) after incubation for 9 days. (A) ApoC-II alone. (B) ApoC-II in the presence of 10 lgÆmL )1 serum clusterin. Radial scans are shown at 15 min intervals and a rotor speed of 6000 r.p.m. (2600 g)(thinblack lines). Data were fitted to ls ) g*(s) analysis, with the fits shown as thick grey lines. The optical density contribution due to clusterin is negligible at these concentrations. Ó FEBS 2002 Inhibition of apoC-II amyloidosis by clusterin (Eur. J. Biochem. 269) 2791 We fitted the data in Fig. 3A to a model describing the sedimentation of a distribution of nondiffusing particles [ls ) g*(s)] and obtained good fits (grey lines in Fig. 3A) to a distribution of sedimentation coefficients averaging 400 S (Fig. 4). ls ) g*(s) analysis of the data in Fig. 3B produced good fits (Fig. 3B, grey lines) to a sedimentation coefficient distribution averaging  180 S (Fig. 4). This is significantly lower than the average for apoC-II alone, suggesting that clusterin reduces the overall size of the fibrils formed even though amyloid formation remains highly cooperative. While incubation of apoC-II in the presence of clusterin reduces the size distribution of the amyloid fibrils (Fig. 4), we wondered whether clusterin could alter the size of preformed fibrils. Sedimentation velocity analysis was used to compare the size distribution of apoC-II (0.3 mgÆmL )1 ) incubated for 7 days at room temperature, followed by the addition of seminal clusterin (0.1 mgÆmL )1 ) or an equivalent volume of buffer and further incubation for 2 days. ls ) g*(s) analysis of the sedimentation boundaries of apoC-II alone produced a size distribution with a modal sedimentation coefficient of 420 S, similar to that shown in Fig. 4. Analysis of the incubated apoC-II sample containing added clusterin produced a size distribution similar to apoC-II alone, with a modal sedimentation coefficient of 450 S. Within the limits of experimental error, these results indicate that clusterin does not affect the size of preformed apoC-II amyloid fibrils. Electron microscopy was used to characterize the mor- phology of amyloid fibrils formed in the presence of clusterin (Fig. 5). A sample of apoC-II (0.3 mgÆmL )1 ) incubated in the presence of serum clusterin (0.1 mgÆmL )1 ) for 10 days was analysed by negative staining with potas- sium phosphotungstate (Fig. 5). The morphology of the fibrils appeared indistinguishable to that of fibrils formed in the absence of clusterin [6]. Distinctive features of clusterin alone were not visible at the nominal magnification (·40 000) used to visualize apoC-II amyloid. However, single particles, attributed to clusterin, were observed at higher magnifications (·100 000; data not shown). These results suggest that despite the overall reduction in the mass of amyloid formed, the fibril morphology was not altered by the presence of clusterin. Clusterin does not bind to apoC-II monomer Clusterin has been reported to bind strongly to other proteins, in particular, apoA-I and Ab [25,32]. Strong binding could conceivably directly alter the kinetics of amyloid formation. We used sedimentation velocity analysis to monitor the state of association of freshly prepared apoC-II and serum clusterin. Figure 6A shows the sedi- mentation behaviour of serum clusterin alone at a rotor speed of 40 000 r.p.m. 116 200 g. The scans (shown at 20 min intervals) reveal broad boundaries indicating a heterogeneous population. Continuous sedimentation dis- tribution analysis [31] resolved species ranging from 4 to 18 S, suggesting multimeric complexes of clusterin, consis- tent with previous studies [19]. It must be emphasized, however, that this analysis assumes noninteracting species over the time period of the experiment, a condition that may not strictly apply as previous studies using gel-filtration chromatography indicate that oligomers of clusterin are in rapid equilibrium [33]. The sedimentation of apoC-II alone (freshly diluted from denaturant) produced scans that fitted with random residuals to a single sharp peak (maximum s ¼ 1 S), indicating the presence of close to 100% mono- meric apoC-II (Fig. 6B). Sedimentation of the mixture of apoC-II and clusterin produced boundaries that superim- posed with a summation of the boundaries for the sedimentation of apoC-II alone and clusterin alone (Fig. 6C). This suggests that the sedimentation behaviour of apoC-II and clusterin remain independent and that there is negligible interaction between monomeric apoC-II and clusterin. Fig. 4. Sedimentation coefficient distributions for apoC-II (0.3 mgÆmL )1 ) incubated in the presence and absence of serum clusterin for 9 days. The ls ) g*(s) distributions correspond to the best fits of the sedimentation velocity data presented in Fig. 3 [30]. ApoC-II alone (solid line). ApoC-II and 10 lgÆmL )1 serum clusterin (dashed line). Fig. 5. Negatively stained transmission electron micrograph of apoC-II fibrils (0.3 mgÆmL )1 ) formed in the presence of serum clusterin (0.1 mgÆmL )1 ). Scale bar represents 200 nm. 2792 D. M. Hatters et al. (Eur. J. Biochem. 269) Ó FEBS 2002 DISCUSSION Clusterin has a multitude of putative physiological functions including controlling cell–cell and cell–substratum interac- tions, regulating apoptosis and transporting lipids [20]. Recent evidence suggests that serum clusterin has a chaperone activity similar to the small-heat-shock protein family [19]. These studies show that clusterin inhibits the aggregation of a variety of different proteins that are partially denatured by reducing agents or heat-shock [19]. Suppression of aggregation of these denatured proteins occurs at slightly substoichiometric molar ratios. Clusterin has also been shown to inhibit the aggregation of amyloid- ogenic peptides. A qualitative study, using transmission electron microscopy, found that clusterin inhibited aggre- gation of an amyloid forming peptide derived from the prion protein at substoichiometric concentrations [26]. The aggregation of the Ab amyloid peptide after a fixed time point with different concentrations of clusterin, revealed that clusterin also inhibits Ab amyloid formation [27]. Our results suggest that clusterin, purified from two different tissue sources, suppresses amyloid formation at substoichiometric concentrations (1 : 30–100, clusterin/ apoC-II). This behaviour is similar to our recent results fortheeffectsofa-crystallin upon amyloid formation by apoC-II [18]. Although the serum and seminal forms of clusterin are encoded by the same structural gene, they are known to differ in their glycosylation patterns [20]. Thus, our finding that the serum and seminal forms of clusterin exhibit similar chaperone activity suggests that the chaper- one function of clusterin is not greatly affected by different glycosylation patterns. We propose that clusterin inhibits apoC-II amyloid formation via a similar mechanism to a-crystallin [18]. This mechanism postulates that clusterin interacts stoichiomet- rically with amyloidogenic precursors (nuclei), leading to dissociation of the nuclei back to monomer. At low apoC-II concentrations clusterin is therefore capable of suppressing amyloid formation substoichiometrically, because the con- centration of amyloid nuclei is low. At higher apoC-II concentrations, where self-association to form an oligomeric nuclei is favored, clusterin is less effective in inhibiting amyloid formation (Fig. 2). This mechanism would explain the inhibition observed with other amyloid systems [26], where nucleation is also rate-limiting [34]. The observation that clusterin reduces the sedimentation coefficient distribution of apoC-II fibrils is significant. The reduction in fibril size raises the possibility that clusterin also inhibits fibril growth by interacting with the reactive ends of fibrils. This action could either retard fibril extension or cause dissociation of monomer subunits from the growing fibril. Our sedimentation velocity data indicates that clus- terin does not reduce the size of preformed fibrils. As nuclei and fibril ends both recruit monomers, it seems likely that their interfaces are structurally similar and consequently they expose a similar binding site to clusterin. Because fibril extension is much more rapid than nucleation rates, the dominant mode of inhibition by clusterin may be on nucleation, due to its rate limiting role in fibril growth. For this reason, a bimodal population of monomers and large polymers may persist, despite the retardation of fibril growth. Clusterin is localized to many amyloid deposits in vivo and is expressed ubiquitously across many tissues and species [20,23]. Clusterin is mostly located extracellularly and is found localized to sites of injury (e.g. tissue damage associated with ischemia, amyloid plaques, and cell necrosis [20,23,35]). Such a biological distribution, in line with its activity as a potent inhibitor of amyloid nucleation, suggests that clusterin may play an active role in the regulation of amyloid nucleation in both disease states and during normal homeostasis. Fig. 6. Sedimentation velocity analysis of 0.6 mgÆmL )1 serum clusterin (A), 0.3 mgÆmL )1 freshly prepared apoC-II (B) and a mixture of apoC-II and clusterin (C). Radial scans were acquired at 20-min intervals at a rotor speed of 40 000 r.p.m. (116 200 g) [(A,B) solid lines; (C), points (s)]. The solid lines in (C) represent the summation of the sedimen- tation profiles of freshly prepared apoC-II and clusterin (A,B). Ó FEBS 2002 Inhibition of apoC-II amyloidosis by clusterin (Eur. J. Biochem. 269) 2793 ACKNOWLEDGEMENTS This work was supported by grants from the National Health and Medical Research Council of Australia to G. J. H. and S. B. E., and the Australian Research Council to S. B. E., and a Melbourne Research Scholarship to D.M.H. We thank Lynne Lawrence at C.S.I.R.O., Health Sciences and Nutrition, Parkvillle, Australia, for the technical assistance with the electron microscopy. REFERENCES 1. Havel, R.J., Fielding, C.J., Olivecrona, T., Shore, V.G., Fielding, P.E. & Egelrud, T. (1973) Cofactor activity of protein components of human very low density lipoproteins in the hydrolysis of tri- glycerides by lipoproteins lipase from different sources. Biochem- istry 12, 1828–1833. 2. Jackson, R.L. & Holdsworth, G. (1986) Isolation and properties of human apolipoproteins C-I, C-II, and C-III. Methods Enzymol. 128, 288–297. 3. Tajima, S., Yokoyama, S., Kawai, Y. & Yamamoto, A. (1982) Behavior of apolipoprotein C-II in an aqueous solution. J. Bio- chem. 91, 1273–1279. 4. Hatters, D.M., Lawrence, L.J. & Howlett, G.J. (2001) Sub- micellar phospholipid accelerates amyloid formation by apolipo- protein C-II. FEBS Lett. 494, 220–224. 5. MacRaild, C.A., Hatters, D.M., Howlett, G.J. & Gooley, P.R. (2001) NMR structure of human apolipoprotein C-II in the pre- sence of sodium dodecyl sulfate. Biochemistry 40, 5414–5421. 6. Hatters, D.M., MacPhee, C.E., Lawrence, L.J., Sawyer, W.H. & Howlett, G.J. (2000) Human apolipoprotein C-II forms twisted amyloid ribbons and closed loops. Biochemistry 39, 8276–8283. 7. Hatters, D.M. & Howlett, G.J. (2002) The structural basis for amyloid formation by apolipoproteins. Eur. Biophys. J. 31, 2–8. 8. Westermark, P., Mucchiano, G., Marthin, T., Johnson, K.H. & Sletten, K. (1995) Apolipoprotein A1-derived amyloid in human aortic atherosclerotic plaques. Am. J. Path. 147, 1186–1192. 9. Wisniewski, T., Golabek, A.A., Kida, E., Wisniewski, K.E. & Frangione, B. (1995) Conformational mimicry in Alzheimer’s disease. Role of apolipoproteins in amyloidogenesis. Am.J.Path. 147, 238–244. 10. Higuchi, K., Kitagawa, K., Naiki, H., Hanada, K., Hosokawa, M. & Takeda, T. (1991) Polymorphism of apolipoprotein A-II (apoA- II) among inbred strains of mice. Relationship between the molecular type of apoA-II and mouse senile amyloidosis. Biochem. J. 279, 427–433. 11. Benson, M.D., Liepnieks, J.J., Yazaki, M., Yamashita, T., Hamidi Asl, K., Guenther, B. & Kluve-Beckerman, B. (2001) A new human hereditary amyloidosis: the result of a stop-codon muta- tion in the apolipoprotein AII gene. Genomics 72, 272–277. 12. Bergstrom, J., Murphy, C., Eulitz, M., Weiss, D.T., Westermark, G.T., Solomon, A. & Westermark, P. (2001) Codeposition of apolipoprotein A-IV and transthyretin in senile systemic (ATTR) amyloidosis. Biochem. Biophys. Res. Commun. 285, 903–908. 13. Wisniewski, T., Lalowski, M., Golabek, A., Vogel, T. & Frangione, B. (1995) Is Alzheimer’s disease an apolipoprotein E amyloidosis? Lancet 345, 956–958. 14. El-Agnaf, O.M. & Irvine, G.B. (2000) Review: formation and properties of amyloid-like fibrils derived from alpha-synuclein and related proteins. J. Struct. Biol. 130, 300–309. 15. Husebekk, A., Skogen, B., Husby, G. & Marhaug, G. (1985) Transformation of amyloid precursor SAA to protein AA and incorporation in amyloid fibrils in vivo. Scand. J. Immunol. 21, 283–287. 16. Kudva, Y.C., Hiddinga, H.J., Butler, P.C., Mueske, C.S. & Eberhardt, N.L. (1997) Small heat shock proteins inhibit in vitro Ab (1–42) amyloidogenesis. FEBS Lett. 416, 117–121. 17. Stege, G.J., Renkawek, K., Overkamp, P.S., Verschuure, P., van Rijk, A.F., Reijnen-Aalbers, A., Boelens, W.C., Bosman, G.J. & de Jong, W.W. (1999) The molecular chaperone aB-crystallin enhances amyloid b neurotoxicity. Biochem. Biophys. Res. Com- mun. 262, 152–156. 18. Hatters, D.M., Lindner, R.A., Carver, J.A. & Howlett, G.J. (2001) The molecular chaperone, a-crystallin, inhibits amyloid formation by apolipoprotein C-II. J. Biol. Chem. 276, 33755–33761. 19. Humphreys,D.T.,Carver,J.A.,Easterbrook-Smith,S.B.&Wilson, M.R. (1999)Clusterin has chaperone-like activity similar to that of small heat shock proteins. J. Biol. Chem. 274, 6875–6881. 20. Wilson, M.R. & Easterbrook-Smith, S.B. (2000) Clusterin is a secreted mammalian chaperone. Trends Biochem. Sci. 25, 95–98. 21. O’Bryan, M.K., Baker, H.W., Saunders, J.R., Kirszbaum, L., Walker, I.D., Hudson, P., Liu, D.Y., Glew, M.D., d’Apice, A.J. & Murphy, B.F. (1990) Human seminal clusterin (SP-40,40). Isola- tion and characterization. J. Clin. Invest. 85, 1477–1486. 22. Murphy, B.F., Kirszbaum, L., Walker, I.D. & d’Apice, A.J. (1988) SP-40,40, a newly identified normal human serum protein found in the SC5b-9 complex of complement and in the immune deposits in glomerulonephritis. J. Clin. Invest. 81, 1858–1864. 23. May, P.C. & Finch, C.E. (1992) Sulfated glycoprotein 2: new relationships of this multifunctional protein to neurodegeneration. Trends Neurosci. 15, 391–396. 24. Calero, M., Rostagno, A., Matsubara, E., Zlokovic, B., Frangi- one, B. & Ghiso, J. (2000) Apolipoprotein J (clusterin) and Alzheimer’s disease. Microsc. Res. Techn 50, 305–315. 25. Matsubara, E., Soto, C., Governale, S., Frangione, B. & Ghiso, J. (1996) Apolipoprotein J and Alzheimer’s amyloid b solubility. Biochem. J. 316, 671–679. 26. McHattie, S. & Edington, N. (1999) Clusterin prevents aggrega- tion of neuropeptide 106–126 in vitro. Biochem. Biophys. Res. Commun. 259, 336–340. 27. Oda, T., Wals, P., Osterburg, H.H., Johnson, S.A., Pasinetti, G.M., Morgan, T.E., Rozovsky, I., Stine, W.B., Snyder, S.W. et al. (1995) Clusterin (apoJ) alters the aggregation of amyloid beta- peptide (Ab 1–42) and forms slowly sedimenting Ab complexes that cause oxidative stress. Exp. Neurol. 136, 22–31. 28.Wang,C.S.,Downs,D.,Dashti,A.&Jackson,K.W.(1996) Isolation and characterization of recombinant human apolipo- protein C-II expressed in Escherichia coli. Biochim. Biophys. Acta. 1302, 224–230. 29. Wilson, M.R. & Easterbrook-Smith, S.B. (1992) Clusterin binds by a multivalent mechanism to the Fc and Fab regions of IgG. Biochim. Biophys. Acta 1159, 319–326. 30. Schuck, P. & Rossmanith, P. (2000) Determination of the sedi- mentation coefficient distribution by least–squares boundary modeling. Biopolymers 54, 328–341. 31. Schuck, P. (2000) Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and lamm equation modeling. Biophys. J. 78, 1606–1619. 32. Jenne, D.E., Lowin, B., Peitsch, M.C., Bottcher, A., Schmitz, G. & Tschopp, J. (1991) Clusterin (complement lysis inhibitor) forms a high density lipoprotein complex with apolipoprotein A-I in human plasma. J. Biol. Chem. 266, 11030–11036. 33. Hochgrebe, T., Pankhurst, G.J., Wilce, J. & Easterbrook-Smith, S.B. (2000) pH-dependent changes in the in vitro ligand-binding properties and structure of human clusterin. Biochemistry 39, 1411–1419. 34. Ferrone, F. (1999) Analysis of protein aggregation kinetics. Methods Enzymol. 309, 256–274. 35. Kida, E., Pluta, R., Lossinsky, A.S., Golabek, A.A., Choi-Miura, N.H., Wisniewski, H.M. & Mossakowski, M.J. (1995) Complete cerebral ischemia with short-term survival in rat induced by car- diac arrest. II. Extracellular and intracellular accumulation of apolipoproteins E and J in the brain. Brain Res. 674, 341–346. 2794 D. M. Hatters et al. (Eur. J. Biochem. 269) Ó FEBS 2002 . extracellular location of amyloid deposits, we investigated whether amyloid formation by apoC-II is affected by the extracellular chaperone, clusterin [19,20] propensity of apolipoproteins to form amyloid in vivo [7]. The ability of apoC-II to form amyloid in vitro can be compared to in vivo amyloid formation