SUMMARY. High levels of scrapie infectivity were found in detergent-insoluble residues of hamster brain purified by either repeated pelleting in 10~o NaC1 or ...
J. gen. Virol. (1987), 68, 225-231. Printedin Great Britain
225
Key words: scrapie agent~aggregation/purification
Effects of Different Methods of Purification on Aggregation of Scrapie Infectivity By B. E. C A S T L E , t C. DEES, T. L. G E R M A N AND R. F. M A R S H * Department of Veterinary Science, University of Wisconsin-Madison, Madison, Wisconsin 53706, U.S.A. (Accepted 15 September 1986)
SUMMARY High levels of scrapie infectivity were found in detergent-insoluble residues of hamster brain purified by either repeated pelleting in 10~o NaC1 or by separation in Nycodenz~ gradients. Titres determined by the method of incubation interval assay were 100-fold higher than titres measured by endpoint dilution assay. The protein profiles and end-labelled RNA examined by one-dimensional polyacrylamide gel electrophoresis were not different from samples prepared from uninfected brain. Preparations produced by repeated pelleting were treated with RNase A and/or 7 Murea with no loss of scrapie infectivity. However, the infectivity of samples prepared by gradient centrifugation in Nycodenz® were reduced by 2 to 3 IOglo LDso by treatment with RNase A alone but not in combination with SDS. These results suggest that the scrapie agent may be aggregated by methods of purification employing pelleting in high concentrations of salt, or by adding polycations to disaggregated samples. Scrapie, a natural infection of sheep and goats, is characterized by a long incubation period followed by fatal neurological illness. Pathological features include spongiform encephalopathy accompanied by astrocytosis and neuronal loss. There is no inflammatory response (Chandler, 1959). The intracellular location of the scrapie agent has been the focus of several reports. Initially, differential centrifugation showed that the agent was associated with all eeUular fractions. Further studies showed that infectivity was associated principally with membranecontaining preparations (Hunter & Millson, 1966; Millson et aL, 1971 ; Semancik et aL, 1976), a finding which has been extended by additional studies on brain plasma membrane (Dees et aL, 1985 a, b, c). This subcellular fraction can be cleared of nuclei and myelin to produce a more uniform preparation without sacrificing scrapie infectivity. While most investigators agree to the membrane association of the scrapie agent, there is little agreement as to the nature of the essential protein(s) required for scrapie infectivity. Proteins of 27000 tool. wt. (27K) to 30K (Bolton et aL, 1982; Diringer et al., 1983), 33K to 35K (Oesch et al., 1985; Meyer et at., 1986) and 54K (Bendheim & Bolton, 1986) have all been detected as major components in purified preparations, while studies on subfractionation of the cytoskeleton have shown that scrapie infectivity co-purifies with a major 50K protein having characteristics of glial flbrillary acidic protein (Van Alstyne et at., 1986). Part of the reason for these apparently conflicting findings is the use of various purification procedures by different laboratories. The following studies indicate that the choice of purification method selected can affect the aggregation state of the scrapie agent which in turn influences the results of protease and nuclease treatment. Detergent-insoluble residues were prepared from whole brain tissue using a modification of a procedure reported by H ilmert & Diringer (1984). Five brains from hamsters showing advanced ~"Present address: DNAX Research Institute, 901 California Avenue, Palo Alto, California94303, U.S.A. 0000-7268 © 1987 SGM
226
Short communication
clinical signs of scrapie, or from age-matched controls, were homogenized to 10 ~ (w/v) in 10 ~o N-lauroylsarcosine pH 7.4 (adjusted by adding 1 ml 1 M-NaH2PO, per 100 ml) in a Ten Broeck glass grinder. After 30 min at room temperature, the samples were diluted to 7 ~o (w/v) with 10 N-lauroylsarcosine and centrifuged at 22000 g for 30 min at 19 °C, Supernatants were removed and centrifuged at 215000 g for 2 h at 19 °C. Supernatants were discarded and the pellets resuspended in 6 ml 109/o NaC1, 1 ~ N-lauroylsarcosine (NLS buffer). These samples were sonicated for 3 min at 20 °C and stored at 4 °C overnight after which they were re-sonicated, atiquots removed for protein assay and bioassay, and centrifuged at 215000g for 2-5 h at 19 °C. Supernatants were discarded, and the pellets were resuspended in 1.5 ml NLS buffer and sonicated as previously described. Samples again were removed for protein assay and bioassay. The suspensions were incubated for 1 h at 37 °C, then centrifuged in an Eppendorf 5412 centrifuge for 15 rain. The pellets were resuspended in 1 ml NLS buffer and sonicated. Proteinase K (PK) was added (5 ~tg/ml) and the samples were mixed on a rotating table for 2 h at 37 °C. This protease-treated sample was centrifuged in an Eppendorf centrifuge for 15 min. The pellets were resuspended in 1 ml NLS buffer and sonicated. The protease inhibitor phenylmethylsulphonyl fluoride was added to a final concentration of 0-05 mM before the samples were dialysed against I0 mM-Tris-acetate pH 7.4 and 5 mM-EDTA for 24 h at room temperature. These samples are designated WBpx. Plasma membrane fractions were prepared by homogenizing 15 brains from scrapie-infected and control hamsters to 10~ (w/v) in 10 mu-Tris-acetate pH 7.4 and 150 mu-NaC1. These suspensions were centrifuged at 500g for 10 min at 4 °C. The pellets, containing nuclei and large cell fragments, were discarded and the supernatants centrifuged at 3400 g for 30 min at 4 °C to obtain plasma membrane-enriched pellets which were resuspended to 10~ (w/v) in 10 mM-Trisacetate pH 7-4, 150 mM-blaCl, 5 mM-EDTA and stored overnight at 4 °C. These suspensions were re-centrifuged at 3400 g for 30 min at 4 oC and the pellets were extracted with 10~ Nlauroylsarcosine as described above. These samples are designated PMpk. Similar plasma membrane-enriched fractions from infected and control hamster brains were prepared as described above, then extracted as 10~ suspensions with either 2, 4 or 10~ Nlauroylsarcosine at 4 °C for 2 h. The samples were clarified at 22000g for 30 rain and 5 ml of each supernatant was overlaid onto 7 ml 20 to 40~ continuous Nycodenz® (Accurate Chemical Corp.) gradients containing 1 ~ N-lauroylsarcosine, 10 m/cl-Tris-acetate pH 7.4, 0-4 Mammonium sulphate, 150 mM-NaC1 and 2 mM-EDTA. The gradients were centrifuged at 215 000 g for 16 h at 4 °C, then 1.8 ml fractions were collected from the top and their densities measured before dialysis against 10 mi-Tris-acetate pH 7.5, 5 mM-EDTA, at 4 °C for 48 to 72 h. These samples are designated PMG. The total infectivity associated with the initial whole brain material and with each detergentinsoluble pellet in the four centrifugation steps of the purification scheme are presented in Table 1. Infectivity of the pellets from the first hi,h-speed centrifugation and second low-speed centrifugation were calculated by both incubation interval assay (Prusiner et al., 1981) and by serial endpoint dilution (Dougherty, 1964), the latter procedure yielding values over 100-fold lower. The linear relationship between the inoculum dilution and the length of the incubation period, which is seen in the whole brain sample before detergent extraction, was not maintained upon dilution of insoluble residues (e.g. serial tenfold dilutions often produced the same mean length of incubation period). The PMok samples had an infectivity titre of 101° LDs0/ml when measured by the incubation interval assay, and 108 LDso/ml when measured by endpoint dilution assay. The infectivity in the Nycodenz gradients was highest (109 LDso) in a fraction having a density of 1-17 g/ml and which often contained a 'parachute-like' band in the scrapie-infected sample. Table 2 shows the protein concentrations of the detergent-insoluble residues prepared from whole brain and the percentage recovery in each serial pellet. The initial 10~ N-lauroylsarcosine residues from control brain contained twice the amount of protein compared to the scrapieinfected samples, but these results were reversed in the final protease-treated pellets. The percentage of protein recovered from samples resuspended in 10~ NaC1 before and after PK
Short communication T a b l e 1.
227
Infectivity of detergent-insoluble residues pelleted from scrapie-infected hamster brain
Preparation Five g whole brain
First high speed pellet (in 6 ml)
Dilution 5 x 10-L 5 x 10-2 5 x 10-3 5 x 10-4 5 x 10-5 5 x 10-6 5 x 10-7 5 x 10-s 5
X
5 5 5 5 5 5
x l0 -1 x 10 2 x 10-3 x 10-4 x 10-s x 10-6
10 -9
5 X 10 -7
Second high speed pellet (in 4 ml) First low speed pellet (in 2 ml) Second low speed pellet (in 1 ml)
Mean length of incubation Response* in days 3/3 54 3/3 58 3/3 65 3/3 73 3/3 80 3/3 89 3/3 103 3/3 121 1/3 129 3/3 60 3/3 60 3/3 60 3]3 82 3/3 89 2/3 107 3/3 105 0/3
5
x
10-s
5
X
10 -9
0/3
5 5 5 5 5 5 5 5 5 5
x x x X X x × x
10-2 10-z 10-' 10-2 10-3 10-4 10~s 10-6 x 10-7 × 10-s
3/3 3/3 3/3 3/3 3/3 3/3 3/3 3/3 3/3 0/3
5
X
10 -9
0/3
58 63 57 58 65 80 79 go 102
Total infectivityt 11,2 (IIA) 10.5 (ED)
ll-4 (IIA) 8"5 (ED)
11"5 (IIA) 10"8 (IIA) 10"9 (IIA) 8'2 (ED)
* Number of hamsters developing progressive neurological disease/number inoculated, t loglo LDso scruple infectivity measured by the method of incubation interval assay (IIA) calculated using the 5 x 10-2 dilution groups, or by endpoint dilution (ED) using the Spearman-K/irber method.
T a b l e 2.
Protein concentrations of detergent-insoluble residues pelleted from scrapie-infected and control hamster brains Total protein Preparation 22000 g supernatant First high speed pellet in 6 ml Second high speed pellet in 1.5 ml First low speed pellet in 1-5 ml Second low speed pellet in 1 ml
Control 530 7.2 (1.4%) 0.54 (7%) 0-19 (38%) 0.04 (23%)
in
mg (% recovery)*
A
Scrapie-infected 591 4-2 (0-7%) 0.59 (14%) 0.34 (58%) 0.07 (22%)
* As measured by the dye binding method of Bradford (1976). Percentage recovery of protein from preceding sample is indicated in parentheses.
t r e a t m e n t indicates that the h i g h e r protein c o n c e n t r a t i o n in the final pellet does not result f r o m a greater susceptibility o f the control sample to proteolytic hydrolysis, but r a t h e r m a y be attributed to increased aggregation o f the scrapie-infected sample in high c o n c e n t r a t i o n s o f NaC1. T h e protein profiles o f the insoluble residues f r o m control and scrapie-infected b r a i n are presented in Fig. 1. Paired control and scrapie-infected samples r e m o v e d d u r i n g the purification procedure do not c o n t a i n different proteins.
Short communication
228
M
(a)
(b)
(c)
(d)
(e)
(./)
(g)
(h)
M
66 45 34 29 24
Fig. 1. Silver-stained 15% disc SDS-polyacrylamide slab gel of whole brain samples purified using the method of Hilmert & Diringer (1984). Lanes represent control and scrapie-infected samples alternately: (a, b) 22000g supernatant; (c, d)first 215000g pellet; (e,f) second 215000g pellet; (g, h) 12500g pellet after PK treatment. M, Mol. wt. standards ( × 10-3). Arrows indicate the 27K to 30 K protein. Note that there were no detectable differences between scrapie and control preparations at any stage of the purification.
Table 3. Inactivation studies on PMG fraction Treatment None 7 M-Urea RNase A (200 ~tg/ml) 7 M-Urea and RNase A RNase A and 0.1 ~ SDS RNase A and 0.5~ SDS RNase A and 1~ SDS
Trial 1
Trial 2
8-8* 5.0 6-5 2.0 Not ND ND
9.5 8.0 6.0 6.0 9,0 9-0 8"5
* log.o LDs0 as measured by the incubation interval assay. t ~rD, Not done. R N A from the samples designated WBpk and PMpk was extracted with phenol, purified with CsCI, concentrated by precipitation in ethanol, and quantified by u.v. absorbance at 260 nm. The control sample made from whole brain (WBv0 contained 17 ~Lgof single-stranded R N A and the scrapie-infected sample 4 ~tg. The control samples made from plasma membrane-enriched fractions (PMpk) contained 9 ~tg and the scrapie-infected sample 16 lsg, Autoradiographs of gels containing 5' end-labelled R N A from the PMpk samples revealed numerous bands of low molecular weight heterogeneous R N A . PM6 samples were similarly extracted, labelled, and electrophoresed. The control fraction (1.8 ml) contained 5 ~g of single-stranded R N A and the scrapie-infected sample 9 Ixg. Many bands ~ low molecular weight R N A were visualized. WBo~ and PMpk samples were treated with 7 M-urea and/or RNase A with no detectable reduction in scrapie infectivity, In contrast, Table 3 shows a loss of infectivity of P M o samples
Short communication
229
Table 4. Sensitivity of PM6 R N A to RNase A PMG RNA in lag* (~ recovery) ¢
Untreated RNase A (200 lag/ml) RNase A and 0.1~ SDS RNase A and 0-5~ SDS
Trial 1 20.1 4.7 (23'5~0) 4.8 (23-5~) 13.2 (65.7~)
Trial 2 47.9 5.8 (12.19/o) 12.0 (25'1~) 10.9 (23~)
Trial 3 15-8 7-0 (44~) 6.2 (40~) lqD~"
* Based on the absorbance at 260 nm after ethanol precipitation. t NO, Not done.
after treatment with RNase A alone. However, no loss in infectivity was observed when these samples were treated with RNase A in combination with SDS even though RNase A was demonstrated to be active in SDS against PMc RNA (Table 4). The present study is an extension of our earlier studies on scrapie infectivity associated with membrane vesicles which indicated that the agent is a molecular complex which co-purifies with detergent-insoluble residues (Marsh et al., 1984; Dees et al., 1985 a, b, c). These experiments used a method described by other investigators (Hilmert & Diringer, 1984) to extract whole brains or plasma membrane-enriched fractions with a high concentration of an anionic detergent, followed by removal of insoluble residues by centrifugation. These preparations contained a high level of scrapie infectivity as measured by the incubation interval assay, but lesser amounts using serial endpoint dilution. The purification procedure of repeated pelleting in high concentrations of NaC1 produces large aggregates which are only partially dispersed by sonication. Aggregation is a likely reason for lower titres when infectivity is measured by tenfold serial dilutions, but factors other than the physical state of the inoculum also must be considered. As we have discussed previously (Marsh et al., 1984), aggregation of conventional viruses leads to an underestimation of actual titres by causing several particles to enter a single cell. However, the effect of aggregation on unconventional viruses, such as scrapie, remains to be defined. If the agent aggregates principally with itself, then the host may dissociate these aggregates into individual infectious units capable of infecting separate ceils. Infection of a single cell with multiple particles may accelerate the replication of the agent thereby producing short incubation periods. If the agent is aggregating with host components, then the short incubation periods may not be due to an increase in the number of infectious units, but rather the result of altered pathogenesis due to differential protection, adsorption or intracellular processing of the particle. The present study indicates that proteins present in the scrapie-infected samples are more susceptible to aggregation by NaCI than proteins from control samples (Table 2). The higher percentage recovery of protein in scrapie samples after treatment with 10~ NaCI suggests that any differences seen in protein content of the final preparation may be due to greater aggregation of the scrapie-infected sample rather than to innate resistance to proteolytic hydrolysis. Fig. 1 shows that by loading SDS-PAGE gels with equal protein concentrations a 27K to 30K protein can be demonstrated in both the control and the scrapie-infected lanes. Other studies (Meyer et al., 1986) have reported that uninfected samples are completely digested by protease and no 27K to 30K protein was detected, a finding which may depend on the method of purification used. We detected large amounts of low molecular weight RNA in the WBpk, PMpk and PM6 samples. Similar concentrations and profiles in residues from both control and scrapie-infected brains indicate that RNA may be entrapped preferentially in these insoluble fractions during preparation. High concentrations of salt can cause aggregation of ribonucleoproteins (Zeive & Penman, 1981). We did not observe a unique RNA species in the scrapie-infected samples, but the possibility of this was hampered by extensive breakdown of the RNA due to the length of the extraction procedures and repeated sonication of some samples. Future attempts to characterize low molecular weight RNA should emphasize methods to minimize RNA breakdown.
230
Short communication
Addition of RNase A to the PMG fraction reduced scrapie infectivity. A similar effect has been observed on RNase treatment of cytoskeletal extracts containing either soluble (actin and tubulin) or insoluble (GFAP) proteins (Van Alstyne et al., 1986). However, in this latter instance treatment with poly-L-lysine also reduced infectivity indicating that the effect may be caused by aggregation of the sample due to the polycationic properties of RNase A rather than a hydrolytic effect on an essential RNA component. This possibility is strengthened by the observations in these experiments that RNase did not reduce infectivity in the presence of SDS, and that RNase treatment had no effect on the WBpk and PMpk samples which are already highly aggregated. It is clear that we do not yet understand the precise conditions which interact to influence purification of scrapie infectivity. There appears to be some evidence to suggest that proteinprotein or, perhaps, protein-RNA interactions are modified in diseased brain tissue to enhance aggregation. This may explain the finding of differential susceptibility of infected and uninfected samples to proteolytic digestion. Furthermore, there is additional evidence to indicate that the manner of aggregation, or the time in the purification procedure at which the agent becomes aggregated, are both important considerations. In these studies we observed both reduction and prolongation of incubation periods due to aggregation. This phenomenon may be due in part to the purity of the sample at the time of aggregation. The WBpk sample becomes noticeably aggregated with the first high speed pellet (Tables 1 and 2) when the sample still contains a large amount of host tissue, in contrast to the PM~ fractions which were aggregated after purification by the addition of polycations. Exactly how these different conditions of aggregation influence agent-host interaction remains to be defined. This research was supported by the College of Agricultural and Life Sciences, University of WisconsinMadison, and by grants from the National Institutes of Health (NS 14822) and the United States Department of Agriculture.
REFERENCES BENDHEIM,P. E. & BOLTON,D. C. (1986). A 54-KDa normal cellular protein may be the precursor of the scrapie agent protease-resistant protein. Proceedings of the National Academy of Sciences, U.S.A. 83, 2214-2218. gOLTON, D. C., McKINLEY,M. P. & PRUSlNER, s. B. (1982). Identification of a protein that purifies with the scrapie prion. Science 218, 1309 1310. BRADFORD, M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72, 248-254. CHANDLER, R. L. (1959). Attempts to demonstrate antibodies in scrapie disease. Veterinary Record 71, 58-61. DEES, C., GERMAN,T. L, WADE, W. F. & MARSH, R. F. (1985a). Characterization of proteins in membrane vesicles from scrapie-infected hamster brain. Journal of General Virology 66, 851-859. DEES, C., McMILLAN,B, C., WADE, W. F., GERMAN,T. L. & MARSH,R. F. (1985b). Characterization of nucleic acids in membrane vesicles from scrapie-infected hamster brain. Journal of Virology 55, 126--132. DEES, C., WADE,W. F., GERMAN,T. L. & MARSH,R. F. (1985C). Characterization of lipids in membrane vesicles from scrapie-infected hamster brain. Journal of General Virology 66, 861-870. DIRINGER, H., GELDERBLOM,H., HILMERT,H., OZEL,M., EDELBLUTH,C. & KIMBERLIN,R. H. (1983). Scrap ie infectivity, fibrils and low molecular weight protein. Nature, London 306, 476-478. DOUGHERTY, R. M. (1964). Quantitation of viruses. In Techniques in Experimental Virology, pp. 169-224. Edited by R. J. C. Harris. New York: Academic Press. HILMERT, H. & DIRINGER, H. (1984). A rapid and efficient method to enrich SAF-protein from scrapie brains of hamsters. Bioscience Reports 4, 165-170. nurcr~R, G. D. & MILLSON,G. C. (1966). Distribution and activation of lysosomal enzyme activities in subcellular components of normal and scrapie-affected mouse brain. Journal of Neurochemistry 13, 375-379. MARSH,R. F., CASTLE,B. E., DEES,C. & WADE, W. F. (1984). Equilibrum density centtifugation of the scrapie agent in Nycodenz. Journal of General Virology 65, 1963-1968. MEYER, R. K., McKINLEY,M. P., BOWMAN,K. A., BRAUNFELD,M. B,, BARRY, R. A. & PRUSINER,S. B. (1986). Separation and properties of cellular and scrapie prion proteins. Proceedings of the National Academy of Sciences, U.S.A. 83, 2310--2314. MILLSON,G. C., HUNTER, G. D. & KIMBERLIN,R. H. (1971). An experimental examination o f the scrapie agent in cell membrane mixtures. II. The association of scrapie activity with membrane fractions. Journal of Comparative Pathology 81, 255 265. OESC"H,B., WESTAWAY,D., WALCHLI, M., McKINLEY, M. P., KENT, $. B. H., AEBERSOLD,R., BARRY, R. A., TEMIPST,P., TEPLOW, D. B., HOOD, L. E., PRUSINER,S. B. & WEISSMAN,C. (1985). A cellular gene encodes scrapie PrP 27-30 protein. Cell 40, 735-746.
Short communication
231
PRUSINER, S. B., COCHRAN, S. P., GROTH, D. F., DOWNEY, D. E., BOWMAN,K. A. & MARTINEZ, H. M. (1981). M e a s u r e m e n t
of the scrapie agent using incubation time interval assay. Annals of Neurology 11, 353-358. (1976). Properties of the scrapie agentendomembrane complex from hamster brain. Journal of Virology 18, 693-700. VAN ALSTYNE, D., DECAMILLIS, M., SUONA, P. S. & MARSH, R. F. (1986). Rubella virus associated with cytoskeleton (Rubella VACS) particles - relevant to scrapie ? In Positive Strand RNA Viruses(UCLA Symposia on Molecular and CellularBiology, new series, vol. 54). Edited by M. A. Brinton & R. Rueckert. New York: Alan R. Liss (in press). ZEIVE, t3. & PENMAN,S. (1981). Subnuclear particles containing small nuclear RNA and heterogeneous nuclear RNA. Journal of Molecular Biology 145, 501-506.
SEMANCIK, J. S., MARSH, R. F., GEELEN, J. L. M. C. & HANSON, R. P.
(Received 3 July 1986)
