Methodological variables in the assessment of beta amyloid neurotoxicity

NeurobiologyofAging, Vol. t3, pp. 609-612, 1992

0197-4580/92 $5.00 + .00 Copyright © 1992PergamonPressLtd.

Printed in the U.S.A.All rightsreserved.

Methodological Variables in the Assessment of Beta Amyloid Neurotoxicity J O R G E B U S C I G L I O , A L F R E D O L O R E N Z O A N D B R U C E A. Y A N K N E R l

Department of Neurology, Harvard Medical School and The Children's Hospital, 300 Longwood Ave, Boston, MA 02115

Received 15 M a y 1992; Accepted 15 J u n e 1992

BUSCIGLIO,J., A. LORENZO AND B. A. YANKNER. Methodologicalvariablesin the assessmentof beta amyloidneurotoxicity. NEUROBIOL AGING 13(5) 609-612, 1992.--Cell culture systems for evaluating the biologicaleffects of the/3-amyloid protein are potentiallyimportant tools in the study of the pathogenesisof Alzheimer'sdisease. In this report, methodologicalconsiderations in the assessment of/3-amyloid neurotoxicity are discussed. Chronic incubation of/31-40 in primary human cortical cultures results in progressive neuronal degeneration. The neurodegenerative process occurs in association with localizeddeposition of/3amyloid on the neuronal soma ultimately resulting in the formation of compact/3-amyloid deposits. A/31-40 preparation from another laboratory was tested that did not form neuronal/3-amyloid deposits and was not neurotoxic. Thus, the conformational state of the/31-40 peptide leading to the formation of neuronal amyloid deposits is an important determinant of neurotoxicity. Variables in peptide preparation that influence this property may account for variation in neurotoxic potency. /3-Amyloid

Alzheimer's disease

Neurotoxicity

Peptideconformation

THE establishment of in vitro systems that can recapitulate aspects of the pathogenesis of Alzheimer's disease is of utmost importance in advancing our understanding of the disease mechanism and in facilitating the develoment of therapeutic modalities. The neurotoxicity of/3-amyloid has been observed in several cell culture and animal systems (2,6,8,9,11-13). In this brief report, several methodological aspects of/3-amyloid bioassays developed in our laboratory will be discussed. New data will be presented suggesting that the formation of neuronal /3-amyloid deposits may be necessary for neurotoxicity. Furthermore, a loss of this capacity in some preparations of fl-amyloid peptides can account for a loss of neurotoxic potency.

ronal loss appear after 24 h. At higher cell densities, more prolonged treatment for 36 h to several days is necessary before significant neuronal loss is apparent. Degenerative changes can best be appreciated in fixed cultures after immunocytochemical staining with a variety of antibodies to neuronal antigens including MAP-2 and tubulin. The neurotoxic effects of/3-amyloid can be observed in both serum-containing or serum-free supplemented medium. Amyloid peptide solutions should not be filtered prior to use since hydrophobic peptides can be lost on the filter membrane.

RAT HIPPOCAMPALNEURONALCULTURE

We have examined the effects of/3-amyloid on dissociated cortical neuronal cultures derived from rat embryos and human fetal abortuses. Mixed cortical cultures of neurons and glia are maintained at high cell density with good neuronal viability for several weeks. These high-density cortical cultures require l0 to 14 days of chronic incubation with 1-20 ~tg/ml of i l l - 4 0 peptide before neuronal cell death is clearly apparent (Figs. 1 and 2). Two previous reports concluded that fl-amyloid was not directly toxic to rat or human cortical neurons on the basis of a 4-day incubation period (5,7). We find that this incubation period is too short to cause neuronal loss in high density cortical cultures, although other signs of degeneration appear including diminished neurite outgrowth and altered cytoskeletal proteins (11).

RAT AND HUMAN CORTICALCULTURES

The initial reported observation of the neurotoxicity of/3amyloid was made in dissociated hippocampal cultures from rat embryos (12). The advantage of this CNS culture system is that neurons can be maintained at low cell density facilitating quantitative and morphological analysis (3). We use a range of cell densities obtained by plating cells at 2-8 X l04 cells per 16 mm culture well. As previously reported, fl-amyloid is neurotoxic after neurons have begun to differentiate, usually by day 3 in culture (12). The first morphological change is thickening and retraction of dendrites which can be observed as early as 12 h after addition of the peptide. Perikaryal degeneration and neu-

J To whom reprint requests should be addressed.

609

610

BUSCIGLIO, LORENZO AND YANKNER

Acetonitrile

H20

DMSO

Z

T

Y

e

x

¢.

FIG. 1. Neurotoxic potency of/31-40 in different solvent vehicles• Lyophilized HPLC-purified/31-40peptide was dissolvedat a concentration of 1.0 mg/ml (by amino acid analysis) in either deionized water, 35% acetonitrile and 0.1% trifluoroacetic acid, or 100% dimethyl sulfoxide (DMSO). The/31-40 stock solution or vehicle alone was added directly to culture medium at a 1:100 dilution (final peptide concentration 10 ug/ml) in primary dissociated cultures of human embryonic cortex. After 14 days of chronic incubation, neuronal cell counts were performed as previously described (12). Values represent the mean + SD, n = 10-20 counted fields. Cultures were maintained in DMEM with 10% iron-supplemented calf serum (HyClone) for 8 days prior to use. After addition of peptides, a half-volume change of medium and peptide or diluted vehicle was performed every 4 days. *p < 0.01 relative to vehicle by Student's t-test•

PEPTIDES

The neurotoxic potency of the/31-40 peptide was assessed in human cortical cultures after dissolution of the peptide in different solvent vehicles. Lyophilized peptide was dissolved to a stock concentration of 1 mg/ml in either deionized water, 35% acetonitrile and 0.1% trifluoroacetic acid or 100% dimethyl sulfoxide (DMSO) and then diluted l: 100 directly into tissue culture medium. The diluted water and acetonitrile-based vehicles alone did not significantly affect neuronal viability (Fig. 1). The diluted DMSO vehicle at a final concentration of 1% caused some neuronal loss. The neurotoxic potency of i l l - 4 0 was greatest when initially dissolved in the acetonitrile vehicle, but substantial neurotoxicity was also observed for/31-40 initially dissolved in water (Fig. 1). The lyophilized/31-40 peptide is poorly soluble in salt-containing buffers such as PBS, but can be introduced into saline-containing solutions after being initially dissolved in one of the three vehicles described above. The differential neurotoxic potency of/31-40 after dissolution in different vehicles in which the peptide is soluble suggests that the conformational state of the peptide may be an important determinant of biological activity. Several different/3-amyloid peptides from commercial sources and from other laboratories have been tested in our bioassays. In particular, we carefully examined a synthetic/31-40 peptide provided by another investigator who had used this peptide for intracerebral injections in animals with negative results. We found that this peptide (/31-40D) was considerably less neurotoxic in low density rat hippocampal cultures than our peptide (/31-40Y). When our peptide (/31-40Y) was incubated in primary human cortical cultures, a gradual localized deposition of the

amyloid peptide appeared in association with neurons as determined by immunocytochemical double-labeling for/3-amyloid and tubulin (Fig. 2A). The formation of neuronal /3-amyloid deposits was accompanied by neuronal degeneration and the appearance of dystrophic neurites. After 14 days, compact deposits of/3-amyloid formed in association with the cell bodies of degenerating neurons (Fig. 2B). Incubation of human cortical cultures with the other/31-40D peptide at a similar concentration (as determined by amino acid analysis) did not produce any neuronal deposition of/3-amyloid (Fig. 2C and D). The cultures treated with/31-40D did not show any evidence of neurotoxicity and were indistinguishable from untreated cultures (Fig. 2D). Treatment with the reverse sequence/340-1 control peptide synthesized under the same conditions as our/31-40Y peptide did not produce neuronal deposits at this concentration and did not cause neuronal degeneration (data not shown). Thus, the formation of neuronal amyloid deposits may be a necessary accompaniment of neurotoxicity. Furthermore, the diminished neurotoxic potency of the/31-40D peptide appears to be directly related to its inability to form neuronal amyloid deposits. The/31-40Y and/31-40D peptides were HPLC purified after synthesis and coelute on a reverse phase C18 column. We have performed amino acid sequencing on both peptides and have confirmed that the primary sequences are correct. The peptides were analyzed by laser desorption mass spectroscopy which showed a single predominant peak for both peptides with a mass of 4334 for/31-40Y and 4335 for/31-40D (predicted 4330). DISCUSSION

In this report, we have compared the biological effects of two highly purified synthetic /31-40 peptides. One peptide forms neuronal amyloid deposits and is neurotoxic; the other peptide does not form neuronal amyloid deposits and is not neurotoxic. Thus, two/31-40 peptide preparations with correct primary sequences and very similar mass spectroscopic profiles can have quite different aggregation-related behaviors and neurotoxic potencies. This difference is likely to reflect conformational changes as a consequence of variables in peptide synthesis or purification. Subtle changes in the structure or conformation of/3 amyloid can significantly affect its aggregation state and capacity to form amyloid fibrils ( 1). For example, in the mutation associated with the Dutch variant of hereditary cerebral amyloidosis, a single amino acid change in/3-amyloid markedly affects the formation of amyloid fibrils from the soluble peptide (10). Previously reported results suggest that the neurotoxic activity of/3-amyloid is mediated by the tachykinin-homologous hydrophobic domain spanning residues 25 to 35, a region that would be expected to influence the aggregation state of the peptide (12). A role for/3amyloid aggregation is also suggested by the observation that preaggregation of/3-amyloid can cause neurotoxicity in young, undifferentiated hippocampal neurons that are normally resistant to/3-amyloid neurotoxicity (8). Thus, the biophysical state of/3-amyloid is an important determinant of its biological activity. A growing body of evidence suggests that the/3-amyloid protein is not an inert byproduct of the metabolism of amyloid precursor protein but rather has significant direct neurotoxic activity [reviewed in (14)]. In Alzheimer's disease, the critical and rate-limiting event in the neurodegenerative process may be amyloid fibrillogenesis which converts /3-amyloid from an inert substance to a highly neurotoxic agent. The neuropathological observation that mature fibrillar amyloid plaques are almost always associated with neurodegenerative changes suggests

~

•

-'IG. 2. Neuronal deposition of the 31-40 peptide is necessary for neurotoxicity. Two different/31-40 peptides were incubated with primary human cortical neurons for 14 days as described in he Fig. I legend. The culture was then fixed and double-label immunocytochemistry was performed for/3 amyloid (red) and tubulin (brown). (A) Chronic incubation of cortical neurons with )ur peptide 131-40Y. Note the widespread formation of cell-associated/3-amyloid deposits (red) and the prominent neuronal loss. (B) Higher magnification view of 131-40Y-treated culture shows legenerating neurons (arrows) in association with a/3 amyloid deposit. (C) Chronic incubation of cortical cultures with another peptide/31-40D. Note the complete absence of amyloid deposition Lnd the intact well-developed state of the cortical culture. (D) Higher magnification view of/31-40D-treated culture shows viable neurons with intact neuritic processes. Again, note the absence )fcell-associated/3-amyloid. immunocytochemistry was performed with a/3-amyloid antibody (antibody 1280, 1:800) (4) visualized with an alkaline phosphatase-conjugated secondary antibody Lnd with a monoclonal antibody to a-tubulin (clone DMIA, 1:400" Sigma) visualized with a biotin-conjugated secondary antibody and extravidin-peroxidase (Sigma). Scale bar, 20 pM.

B

612

BUSCIGLIO, L O R E N Z O A N D Y A N K N E R

that once amyloid fibrillogenesis occurs, neuronal degeneration is rapid. This is consistent with observations on ~3-amyoid neurotoxicity in vitro. The chronicity of Alzheimer's disease would then reflect the slow rate o f a m y l o i d fibrillogenesis, which may vary among different neuronal populations and brain regions.

ACKNOWLEDGEMENTS We thank D. Selkoe for providing the 1280 ~-amyloid antibody and Larry Duffy and James Labdon for assistance in peptide characterization. This work was supported by grants NS-01240 and AG-09224 from the National Institutes of Health and by a grant from the Alzheimer's Association (to B. A. Yankner).

REFERENCES 1. Barrow, C. J.; Zagorski, M. G. Solution structure of/3 peptide and its constitutent fragments: Relation to amyloid deposition. Science 253:179-182; 1991. 2. Frautschy, S. A.; Baird, A.; Cole, G. M. Effects of injected Alzheimer 3-amyloid cores in rat brain. Proc. Natl. Acad. Sci. USA 88:83628366; 1991. 3. Goslin, K.; Banker, G. Rat hippocampal neurons in low-density culture. In: Banker, G.; Golsin, K., eds. Culturing nerve cells. Cambridge, MA: MIT Press; 1991:251-282. 4. Joachim, C. L.; Mori, H.; Selkoe, D. J. Amyloid/3-protein deposition in tissues other than brain in Alzheimer's disease. Nature 341:226230; 1989. 5. Koh, J.-Y.; Yang, L. L.; Cotman, C. W./3-Amyloid protein increases the vulnerability of cultured cortical neurons to excitotoxic damage. Brain Res. 533:315-320, 1990. 6. Kowall, N. W.; Beal, M. F.; Busciglio, J.; Duffy, L. K.; Yankner, B. A. An in vivo model for the neurodegenerative effects of/3 amyloid and protection by substance P. Proc. Natl. Acad. Sci. USA 88:72477251; 1991. 7. Mattson, M. P.; Cheng, B.; Davis, D.; Bryant, K.; Lieberburg, I.; Rydel, R. E. ~-Amyloid peptides destabilize calcium homeostasis and render human conical neurons vulnerable to excitotoxicity. J. Neurosci. 12:376-389; 1992.

8. Pike, C. J.; Walencewicz, A. J.; Glabe, C. G.; Cotman, C. W. In vitro aging of~3amyloid protein causes peptide aggregation and neurotoxicity. Brain Res. 563:311-314; 1991. 9. Roher, A. E.; Ball, M. J.; Bhave, S. V.; Wakade, A. R. B-Amyloid from Alzheimer disease brains inhibits sprouting and survival of sympathetic neruons. Biochem. Biophys. Res. Commun. 174:572579; 1991. 10. Wisniewski, T.; Ghiso, J.; Frangione, B. Peptides homologous to the amyloid protein of Alzheimer's disease containing a glutamine for glutamic acid substitution have accelerated amyloid fibril formation. Biochem. Biophys. Res. Commun. 179:1247-1254; 1991. 11. Yankner, B. A.; Busciglio, J. Beta amyloid causes neurofibrillary degeneration in cultured human cortical neurons. Neurology 42(Suppl. 3):304; 1992. 12. Yankner, B. A.; Duffy, L. K.; Kirschner, D. A. Neurotrophic and neurotoxic effects of amyloid/3 protein: Reversal by tachykinin neuropeptides. Science 250:279-282; 1990. 13. Yankner, B. A.; Caceres, A.; Duffy, L. K. Nerve growth factor potentiates the neurotoxicity of/3 amyloid. Proc. Natl. Acad. Sci. USA 87:9020-9023: 1990. 14. Yankner, B. A.; Mesulam, M.-M./3-Amyloid and the pathogenesis of Alzheimer's disease. N. Engl. J. Med. 325:1849-1857; 1991.