Journal of General Virology (1992), 73, 3319 3323. Printed in Great Britain 3319 Detection of proteinase K-resistant prion protein and infectivity in mouse spleen by 2 weeks after scrapie agent inoculation Richard E. Race* and Darwin Ernst National Institutes of Health, National Institute of Allergy and Infectious Diseases, Laboratory of Persistent Viral Diseases, Rocky Mountain Laboratories, Hamilton, Montana 59840, U.S.A. The sequential accumulation of the protease-resistant form of the endogenous prion protein (PrP-res) was compared to levels of scrapie infectivity in the spleen and brain of scrapie-infected mice at various times after inoculation. In mouse spleen PrP-res was detected 1 week after inoculation, and increased 65-fold between 1 and 3 weeks post-inoculation and an additional 15-fold during the next 17 weeks. Infectivity in spleen reached a maximum plateau level by 3 weeks. In contrast, in mouse brain PrP-res was not detected until 8 weeks after inoculation and then increased 200-fold during the next 12 weeks. During this same time, infectivity increased approximately 10000-fold. Therefore, in both spleen and brain of scrapie-infected mice accumulation of PrP-res and infectivity appear to be associated. However, it was not possible to show quantitative correlations between PrP-res detection and infectivity, perhaps owing to the inaccuracy of the infectivity assay. Spongiform encephalopathies of man and animals are caused by unique transmissible agents that are unusually resistant to chemical inactivation and treatments that destroy or modify nucleic acid (Alper et al., 1967; Bellinger-Kawahara et al., 1987; Dees et al., 1985; Latarjet, 1979). In preparations enriched for infectivity, a predominant proteinase K-resistant form (PrP-res) of the endogenous prion protein (PrP) has been identified (Bolton et al., 1982; Diringer et al., 1983; Hope et al., 1986). Several investigators have sought to determine the relationship between PrP-res and the causative agent. McKinley et al. (1983) and Jendroska et al. (1991) have shown a reasonably close association between the detection of PrP-res and infectivity in the brains of scrapie-infected hamsters, supporting the notion that the protein itself could be the scrapie agent. However, their data conflict with those reported by Czub et al. (1986) which show that infectivity reaches maximum levels well before significant PrP-res is detected, a finding which led to the conclusion that PrP-res is not the scrapie agent. Rubenstein et al. (1991) detected PrP-res in the spleens of scrapie-infected mice 11 to 12 weeks post-inoculation (p.i.), which again is well after infectivity levels in spleen plateau. However, Doi et al. (1988) were able to detect PrP-res in mouse spleen by 5 weeks p.i., making the association with infectivity more likely, but they did not determine infectivity. Considered together, these data show wide variation with regard to the time p.i. at which PrP-res is detected in animal tissues. Consequently, interpretations of how PrP-res and the scrapie agent might be associated vary. Since we have developed a sensitive technique for detecting PrP-res in mouse spleen, we wondered whether it might be possible to detect PrP-res much earlier p.i. than has been possible previously. If so, this detection might more accurately define possible associations between PrP-res and infectivity. We were particularly interested in evaluating PrP-res in mouse spleen because scrapie agent titres increase rapidly in the spleens of mice within days of inoculation and, in contrast to brain, do so in the absence of histopathological changes. To determine how closely PrP-res and infectivity are associated in mouse spleen we analysed the spleens from mice killed at various times p.i. Rocky Mountain Laboratories (RML) random bred Swiss mice were inoculated intracerebrally with 50 gl of a 1% suspension of brain from clinically sick scrapie-infected mice (Chandler isolate). Groups of mice were killed at 1 day and at 1, 2, 3, 8 and 20 weeks p.i. Their spleens and brains were harvested aseptically and frozen until they could be analysed. PrP-res preparations were made from pooled spleens using several modifications of a technique described previously (Doi et al., 1988). Briefly, 10% suspensions of spleen in 0.01 M-Tris-HC1, 0.005 M-MgC12 were made by forcing the spleens through a fine mesh stainless steel screen. These suspensions, excluding the fibrous material left on the screen, were then treated with DNase (400 ktg/g original tissue) for 1 h at 37 C. An 0001-0895 1992 SGM
3320 Short communication Weeks p.i. 1 2 3 8 20 I II II I I II I 50K 39K 27K 17K Fig. 1. PrP-res detection in mouse spleen at 1, 2, 3, 8 and 20 weeks p.i. Spleen PrP-res was prepared as described in the text. Final pellets containing PrP-res were suspended by sonication in sample buffer (Laemmli, 1970) containing 5 ~ SDS at a concentration of 20 mg/ixl (original wet weight), boiled for 5 rain and held at -70 C until immunoblots could be completed. PAGE and immunoblotting were done as before using rabbit antipeptide antiserum (Race et al., 1992). Values at the top show the mg equivalents of spleen (original wet weight basis) represented in each lane. Mr markers are shown (lane M) and the relative sizes of the bands are shown to the right. The bands shown were demonstrated to be PrP-res by reacting duplicate filters with the antipeptide antiserum pre-adsorbed with the appropriate synthetic peptide. The bands between 17K and 39K disappeared when the anti-prp antiserum was pre-adsorbed in this way, thus verifying that they were PrP-res (data not shown). Intact and degraded proteins other than PrP-res also survive the purification protocol making it impossible to determine infectious units per given amount of protein in these experiments. equal volume of 20 ~ sarkosyl in 0-01 M-Tris-HC1 ph 7.6 was then added and the suspension was maintained at room temperature for 30 min. The suspension was then centrifuged at 10000 g for 30 min. Supernatants were recovered and centrifuged at 215 000 g for 2 h. The pellets were resuspended in 1.0 ml of sterile glass-distilled H20 by sonication. Proteinase K (10 ~tg/g of original tissue) was added and the suspension kept at 37 C for 30 min. PMSF was added to a final concentration of 0.005 M. This suspension was kept on ice for 30 min then centrifuged at 215 000 g for 1 h. The pellet was suspended in sample buffer (Laemmli, 1970) containing 5 ~ SDS at a concentration of 20 mg/~tl (based on original tissue weight) and frozen. PrP-res was prepared from brain using the same procedure except that the initial 10~ homogenates were made using a Dounce homogenizer. Proteins were separated on polyacrylamide gels and the PrP-res bands detected as described (Race et al., 1992). For ease of understanding, we express PrP-res as units (U)/100 mg of tissue. This value was determined by dividing 100 by the minimum mass (mg) of tissue needed to detect PrP-res by immunoblotting. On this basis, 0-2 U of PrP-res/100 mg was detected in spleen 1 week after inoculation. However, detection of 0.2 U at 1 week required that 500 mg equivalents of spleen or approximately five mouse spleen equivalents be analysed by immunoblotting (Fig. 1, lane 2). By 3 weeks after inoculation, 13 U/100 mg was detected, a 65-fold Table 1. Association between PrP-res accumulation and infectivity in mouse spleen and brain Time p.i. P~'P-res LDso/100 mg Tissue (weeks) U*/100 mg tissue tissue t Spleen 1 0.2 6.0 x 105 2 0.8 1.1 107 3 13'0 6'0 107 8 50"0 1"9 X 107 20 200"0 1"2 107 Brain 1 ND~ ND 2 ND ND 3 <0'2 1-2 x 104 8 13.0 1.8 x 105 20 2500-0 1"9 109 * The minimum amount of spleen or brain equivalents needed to detect PrP-res at the various times p.i. was determined by immunoblotting (Fig. 1 and 3) and was divided into 100 to give the number of PrPres U in a standard weight of tissue (100 mg). t The amount of infectivity in the original 10~ homogenates from which PrP-res was recovered was determined as described (Eklund et al., 1967; Race et al., 1992). ND, Not done. increase over the amount detected at 1 week postinoculation. Another 15-fold increase in the amount of PrP-res occurred between 3 and 20 weeks p.i. (Table 1 and Fig. 1). No PrP-res was detected 1 day p.i. when 2000 mg equivalents of spleen (approximately 20 mouse spleen equivalents) were analysed, suggesting that the
Short communication 3321 PrP-res detected in spleen 1 week p.i. was not derived from the inoculum (data not shown). However, when an amount of brain-derived inoculum equivalent to the amount used to inoculate five mice was purified as the spleens were, and then immunoblotted as before, bands representing PrP-res were detected (Fig. 2, lane 5). Thus, the PrP-res detected at 1 week p.i. could, at least in part, have been derived from inoculum if all of the intracerebrally inoculated PrP-res were sequestered in the spleen by 1 week p.i. (compare Fig. 1, lane 2 with Fig. 2, lane 5). By 2 weeks p.i. the intensity of the bands from an equivalent amount of spleen exceeded that of the inoculum (compare Fig. 2, lane 5 with Fig. 1, lane 5). Therefore, by 2 weeks p.i. de novo synthesis of PrP-res had occurred. When 1000 mg equivalents of spleen derived from uninfected mice was purified and immunoblotted, no PrP-res-specific bands were seen (Fig. 2, lane 2). Thus, the PrP-res bands which were observed from the spleen sample from infected mice were the diseasespecific PrP-res species. In contrast to PrP-res, which increased throughout the course of disease, infectivity in the spleens reached a maximum plateau level by 3 weeks p.i. (Table 1). Therefore, it appears that in mouse spleen the rapid early increase in infectivity is accompanied by a corresponding rapid increase in the amount of PrP-res. However, the 15-fold increase in PrP-res which occurred after infectivity had plateaued suggested that the concentration of PrP-res might increase independently of the infectivity titre. It has been shown that in mouse brain the scrapie agent does not replicate to levels equivalent to those in spleen until about 8 weeks p.i., but titres then increase rapidly and exceed the titre in spleen by approximately 100-fold late in the disease (Eklund et al., 1967). We wondered whether PrP-res accumulation would begin immediately p.i. as it did in spleen or later as the agent titre increased. Therefore, we sought to detect PrP-res in brain tissue from the same animals whose spleens were analysed at 1, 2, 3, 8 and 20 weeks p.i. PrP-res was not detected in mouse brain 3 weeks p.i., when 500 mg equivalents of mouse brain were analysed (Fig. 3). Nor was PrP-res detected 1 day p.i., when 2000 mg equivalents of mouse brain were analysed (data not shown). However, at 8 weeks p.i., PrP-res was detected with as little as 7-8 mg equivalents of brain. Between 8 and 20 weeks p.i. the amount of PrP-res increased approximately 200-fold. However, during this interval the amount of infectivity in brain increased 10000-fold (Fig. 3, Table 1). Thus, there was a 50-fold difference between the increase in PrP-res and the increase in infectivity between 8 and 20 weeks p.i. From the data presented here it seemed reasonable to conclude that PrP-res and infectivity are associated, in that PrP-res was never detected in the absence of 50K 39K 27K 17K M 1 2 3 4 5 Fig. 2. Analysis of PrP-res in brain inoculum and spleen tissue from scrapie-infected and uninfected mice. Lane M, M~ markers from which the relative size of the protein bands was estimated; lane 1, PrP-res bands detected with 10 mg equivalents of spleen (wet weight basis) from mice killed 8 weeks p.i. ; lane 2, banding pattern when 1000 mg equivalents of spleen from uninfected mice were analysed; lane 3, duplicate of lane 1 except that the antipeptide antiserum was preadsorbed with the appropriate synthetic peptide before addition to the filter. Bands that disappeared when the anti-prp antiserum was preadsorbed in this way were regarded as PrP-res, the disease-specific species. Lane 4, duplicate filter of lane 1 but reacted with preimmune serum from the same rabbit used to make the antipeptide antiserum used in these studies. The band at approximately 22K, which was also present in the sample from normal spleen (lane 2), is not a diseasespecific band. Lane 5, amount of brain inoculum equivalent to the amount used to inoculate five mice, processed using the same protocol as was used for brain tissue (see text). Five mice give the approximate mg equivalents of spleen necessary to visualize PrP-res 1 week p.i. The intensity of the PrP-res bands is equivalent to that seen in Fig. 1, lane 2, suggesting that the spleen-derived PrP-res seen at 1 week p.i. could represent inoculum. infectivity. However, directly related quantitative increases were not apparent. We wondered whether the discrepancy between increases in PrP-res and infectivity which was observed after 3 weeks in spleen and between 8 and 20 weeks in brain could be due to the margins of error encountered in either the infectivity titrations or PrP-res analyses. Therefore, we investigated the potential infectivity titration error by determining the infectivity titres of eight different but equivalent spleen homogenates and three brain homogenates. The range in titre was 1.7 x 10 4 to 6.0 x 10 5 LDs0/mg for spleen and 1-8 10 6 to 1-9 x l0 7 LDs0 for brain (data not shown). Thus there appeared to be a 10- to 15-fold range of difference in infectivity titrations on comparable samples. To determine the potential error which one might expect from the immunoblot analysis, we analysed PrP-res preparations derived from aliquots of a single
3322 Short communication Weeks p.i. 3 I 8 20 II 11 I 09 M i~. Fig. 3. PrP-res detection in mouse brain at 3, 8 and 20 weeks after scrapie infection. Values at the top of the figure show the amount of brain tissue (rag, based on original wet weight) represented in each lane. M r markers are shown in lane M and the relative sizes of the bands to the right. PrP-res was prepared from brain as described in the text and the legend to Fig. 1 except that homogenates were made using Dounce homogenizers. PAGE and immunoblotting were done as before. The same specificity controls shown in Fig. 2 were also done for brain and gave a similar result (data not shown). pool of mouse spleens taken from mice killed 3 weeks after inoculation of scrapie agent. PrP-res was prepared on different days and immunoblotted on different days but using the same techniques. All values were within a twofold range of each other. Thus, the immunoblot analysis for PrP-res had a much lower potential for error than the infectivity analyses. Nevertheless, the combined potential errors of the infectivity and PrP-res analyses could possibly account for the 15-fold increase in PrP-res between 3 and 20 weeks p.i. in spleen. However, the 50-fold difference between PrP-res and infectivity increases in brain which occurred between 8 and 20 weeks p.i. appeared to exceed the range of error observed. Reports prior to this have shown that whenever PrP-res is detected, infectivity is also present provided agent and PrP-res represent the same species (Jendroska et al., 1991; Rubenstein et al., 1991; McKinley et al., 1983; Czub et al., 1986). However, in all instances detection of infectivity preceded the demonstration of PrP-res accumulation. In the one report comparing PrP-res and infectivity in mouse spleen, infectivity was detected approximately 9 weeks before PrP-res, leading the authors to conclude that PrP-res and infectivity increase independently (Rubenstein et al., 1991). We detected de novo synthesis of PrP-res by 2 weeks p.i., which is by far the earliest that PrP-res has been detected in mice, an observation which makes the association between PrP-res and infectivity stronger than before. However, after a few weeks the association is less precise and our data comparing PrP-res and infectivity after 3 weeks failed to show a quantitative correlation between PrP-res and infectivity. The authors thank Irene Cook Rodriguez and James Nolz for preparing the manuscript, Bob Evans and Gary Hettrick for photographic assistance, and Dr Bruce Chesebro for helpful discussions regarding the data. References ALPER, T., CRAMP,W. A., HAIG, D. A. & CLARKE,M. C. (1967). Does the agent of scrapie replicate without nucleic acid? Nature, London 214, 764-766. BELLINGER-KAWAHARA, C., CLEAVER, J. E., DIENER, T. O. & PRUSINER, S. B. (1987). Purified scrapie prions resist inactivation by UV irradiation. Journal of Virology 61, 159-166. BOLTON, D. C., MCKINLEY, M. P. & PRUSINER, S. B. (1982). Identification of a protein that purifies with the scrapie prion. Science 218, 1309 1311. CzuB, M., BRAIG, H. R. & DIRINGER, H. (1986). Pathogenesis of scrapie: study of the temporal development of clinical symptoms, of infectivity titres and scrapie-associated fibrils in brains of hamsters infected intraperitoneally. JournalofGeneral Virology67, 2005-2009. DEES, C., WADE, W. F., GERMAN, T. L. & MARSH, R. F. (1985). Inactivation of the scrapie agent by ultraviolet irradiation in the presence of chlorpromazine. Journalof General Virology66, 845-849.
Short communication 3323 DIRINGER, H., GELDERBLOM, H., HILMERT, H., OZEL, M., EDELBLUTH, C. & KIMBERLIN, R. H. (1983). Scrapie infectivity, fibrils and low molecular weight protein. Nature, London 306, 476M78. Doi, S., ITO, M., SHINAGAWA, M., SATO, G., ISOMURA, H. & GOTO, H. (1988). Western blot detection of scrapie-associated fibril protein in tissues outside the central nervous system from preclinical scrapieinfected mice. Journal of General Virology 69, 955-960. EKLUND, C. M., KENNEDY, R. C. & HADLOW, W. J. (1967). Pathogenesis of scrapie virus infection in the mouse. Journal of Infectious Diseases 117, 15 22. HOPE, J., MORTON, L. J. D., FARQUHAR, C. F., MULTHAUP, G., BEYREUTHER, K. & KIMBERLIN, R. H. (1986). The major polypeptide of scrapie-associated fibrils (SAF) has the same size, charge distribution and N-terminal protein sequence as predicted for the normal brain protein (PrP). EMBO Journal 5, 2591 2597. JENDROSKA, K., HEINZEL, F. P., TORCHIA, M., STOWRING, L., KRETZSCHMAR, H. A., KON, A., STERN, A., PRUSINER, S. B. & DEARMOND, S. J. (1991). Proteinase-resistant prion protein accumulation in Syrian hamster brain correlates with regional pathology and scrapie infectivity. Neurology 41, 1482-1490. LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, London 227, 68(~685. LATARJET, R. (1979). Inactivation of the agents of scrapie, Creutzfeld- Jakob disease and kuru by radiations. Slow Transmissible Diseases of the Nervous System, vol. 2, pp. 387-407. Edited by S. B. Prusiner & W. J. Hadlow. New York: Academic Press. MCKINLEY, M. P., BOLTON, D. C. & PRUSINER, S. B. (1983). A protease-resistant protein is a structural component of the scrapie prion. Cell 35, 57-62. RACE, R., ERNST, D., JENNY, A., TAYLOR, W., SUTTON, D. & CAUGHEY, B. (1992). Diagnostic implications of detection of proteinase K-resistant protein in spleen, lymph nodes, and brain of sheep. American Journal of Veterinary Research 53, 883-889. RUBENSTEIN, R., MERZ, P. A., KASCSAK, R. J., SCALICI, C. L., PAPINI, M. C., CARP, R. I. & KIMBERLIN, R. H. (1991). Scrapie-infected spleens: analysis of infectivity, scrapie-associated fibrils, and protease-resistant proteins. Journal oflnfectious Diseases 164, 29-35. (Received 4 February 1992; Accepted 29 July 1992)