Based on the paper by Prusiner[1], please describe in one page or less how one of the techniques in the paper (HPLC, proteolysis, Edman degradation, amino acid analysis, SDS page or what not) was util

Cell, Vol. 36, 127-134, August 1964, Copyright Q 1964 by MIT Purification and Structural Studies of a Major Scrapie Prion Protein 0092.6674/&I/060127-06 $0200/O Stanley B. Prusiner,* Darlene F. Groth,* David C. Bolton,* Stephen B. Kent,+ and Leroy E. Hood+ *Department of Neurology and Department of Biochemistry and Biophysics University of California San Francisco, California 94143 +Division of Biology California Institute of Technology Pasadena, California 91125 Summary Scrapie is a degenerative, neurological disorder caused by a slow infectious agent or prion. Exten- sively purified preparations of prions were dena- tured by boiling in sodium dodecyl sulfate and the major protein component (PrP 27-30) was isolated by preparative HPLC size exclusion chromatography after proteinase K digestion. The purified PrP 27-30 molecules were not infectious. Ultraviolet absorption spectra of purified PrP 27-30 demonstrated the ab- sence of covalently linked polynucleotides. Amino acid composition studies showed that PrP 27-30 contains at least 17 naturally occurring amino acids.

A single N-terminal amino acid sequence for PrP 27- 30 was obtained; the sequence is N-Gly-Gln-Gly Gly-Gly-Thr-His-Asn-Gln-Trp-Asn-Lys-Pro-Ser-Lys and it does not share homology with any known proteins. The same amino acid sequence was found when an extensively purified preparation of prions aggregated into rods and containing -10s5 IDso U/ ml was sequenced directly. Knowledge of the amino acid sequence should permit determination of the genetic origin and replication mechanism of prions.

Introduction Scrapie is a degenerative neurological disease of sheep and goats (Hadlow et al., 1979). The disease occurs after a prolonged incubation period of months to years followed by a relatively short, progressive clinical course ending in death. Scrapie seems to be similar to several neurological disorders of humans: kuru, Creutzfeldt-Jakob disease, and Gerstmann-Straussler syndrome (Gajdusek, 1977; Masters et al., 1981). The infectious agent causing scrapie has many unusual molecular characteristics, which have led to introduction of the term prion to denote this class of slow infectious pathogens (Prusiner, 1982). Although prions clearly replicate, attempts to identify a nucleic acid within them have been unsuccessful to date.

The unusual properties of the scrapie prion and the prolonged bioassays required for its measurement have complicated earlier attempts at purification (Hunter, 1972; Kimberlin, 1976; Millson et al., 1976; Siakotos et al., 1976; Prusiner et al., 1978, 1980, 1982a). Recently, purification of the infectious scrapie prion to a nearly homogeneous state has been accomplished (Prusiner et al., 1983). To date, one major macromolecule has been identified within the scrapie prion isolated from infected hamster brain, This molecule is a protein designated PrP 27-30 and it has a molecular weight of 27,000-30,000 as determined by sodium dodecyl sulfate (SDS) gel electrophoresis (Bolton et al., 1982; McKinley et al., 1983).

Recent studies with rabbit antisera produced against PrP 27-30 have demonstrated the presence of four cross- reactive proteins of lower M,, as well as one protein of higher M,, in extensively purified preparations of prions (P.

E. Bendheim, R. A. Barry, S. J. DeArmond, D. P. Stites, and S. B. Prusiner, submitted). These proteins appear to be present in low concentrations compared to PrP 27-30 based on radioiodination and silver-staining data. Whether the lower M, proteins are degradation products of PrP 27- 30 generated during the purification of prions by proteinase K digestion or they are distinct prion molecules with related amino acid sequences remains to be determined. It is possible that the higher M, protein is a precursor of PrP 27-30.

Development of a protocol for extensive purification (Prusiner et al., 1983) of scrapie prions allowed us to demonstrate that rod-shaped structures previously found in less purified preparations (Prusiner et al., 1982a) are polymers of PrP 27-30 molecules. By electron microscopy, the rods measure 1 O-20 nm in diameter and 100-200 nm in length when negatively stained (Prusiner et al., 1983).

Antibodies raised against electrophoretically purified PrP have been shown by immunoelectron microscopy to bind specifically to the rods (Ft. A. Barry, M. P. McKinley, M. B.

Braunfeld, and S. B. Prusiner, unpublished data). Sonica- tion of the rods had no effect on the titer, but reduced their length to

The ultrastructural appearance of the prion rods is similar to that of amyloid and when arrays of rods were stained with Congo red dye, they exhibited green birefringence identical with that of amyloid (Prusiner et al., 1983). Im- munoperoxidase staining with antibodies to PrP has dem- onstrated that amyloid plaques found in the brains of scrapieinfected hamsters are composed of prion rods (P.

E. Bendheim, R. A. Barry, S. J. DeArmond, D. P. Stites, and S. B. Prusiner, submittted).

To extend these observations, homogeneous prepara- tions of the scrapie prion protein were prepared for amino acid composition and sequence determination. Extensively purified preparations of prions were digested with protein- ase K, disaggregated by boiling in SDS, and further purified by HPLC size exclusion chromatography. Fractions con- taining PrP 27-30 judged to be homogeneous by SDS gel electrophoresis were then subjected to amino acid analysis Cell 128 Hamster braln homogenate A Pl $1 4% Trltan X-100 2% No deoxYCholate 8% PEG-8000 A P2 S2 Micrococcal nuclease, 16 h, 4’ Protelnose K, 8 h, 4’ 0,1x SarkosYl 2% No chalate A 30% (NH4j2S04 P3 $3 2% Trlton X-100 0.8% No dadecY1 SO4 Zonal rotor sucrose gradient froctlans Protelnase K, 2 h, 37’ 1% No dOdecY1 SO4 4 min, 100’ HPLC size exCluSlOn chrciTatograahY froctlons Figure 1. Scheme for Preparation of Extensively Purified Scrapie Pnons from Infected Hamster Brain and for isolation of PrP by HPLC Size Exclusion Chromatography P, = pellet, end S,, = supernatant.

and sequence determination. In this communication, we report the N-terminal amino acid sequence of PrP 27-30, as well as its composition and a further description of some of its molecular characteristics. Results Denaturation of Prions and Purification of PrP Purification of the scrapie prion protein (PrP 27-30) was accomplished by disaggregation of extensively purified preparations of scrapie prions followed by HPLC size exclusion chromatography (Figure 1). Purified preparations of prions were concentrated by precipitation with ice-cold 80% methanol, digested with proteinase K (4 pg/ml) at 37°C for 2 hr, and then boiled for 4 min in the presence of 1% SDS and 150 mM sodium phosphate, pH 6.8. This procedure generally diminished the infectivity by a factor of >l 02, disrupted the structure of the rods, and rendered PrP 27-30 sensitive to protease-catalyzed hydrolysis of its peptide bonds (Prusiner et al., 1983). The disaggregated and denatured protein was then chromatographed across a preparative HPLC size exclusion column. The resulting A2,5 profile is shown in Figure 2A. The peak in the void volume presumably contains nucleic acid or protein com- plexes that are excluded from the column. The second peak contains PrP 27-30 and a few contaminating pro- teins. When fractions obtained from the brains of animals inoculated with normal brain were prepared by the same procedure and then chromatographed on the same col- umn, a similar profile was obtained, but the second AZ15 peak was much smaller and contained only contaminating proteins (Figure 26). This second peak in normal prepara- Figure 2. Purification of PrP by HPLC Size Exclusion Chromatography of Scrapie Prion and Control Preparations (A) Soraple prion preparation. PrP is found in the second A215 peak (fractions 85-90). (B) Control preparation from uninfected hamster brain.

tions appears to correspond to the trailing shoulder of the PrP 27-30 peak.

In order to estimate the recovery of PrP 27-30 after HPLC size exclusion chromatography, samples were la- beled with ‘“51-Bolton-Hunter reagent. Approximately 80% of the radioactivity applied to the column could be re- covered in the eluted fractions. The PrP 27-30 peak contained 30%-50% of the recoverable radioactivity.

Aliquots of each of the fractions from scrapie prion preparations were analyzed by polyacrylamide gel electro- phoresis (Laemmli, 1970). The polyacrylamide gels were stained with silver (Merril et al., 1981), as shown in Figure 3. Those fractions (86-88) corresponding to lanes 26-28 and containing only PrP 27-30 were pooled from four or five chromatograms and then subjected to further analysis.

Even after purification by HPLC size exclusion chromatog- raphy, the microheterogeneity of PrP 27-30 with respect to size persisted. The cause of this microheterogeneity demonstrated by SDS polyacrylamide gel electrophoresis is unknown but glycosylation must be considered since recent studies indicate that PrP 27-30 is a sialoglycopro- tein (D. C. Bolton, R. Meyer, and S. B. Prusiner, unpub- lished). Protease digestion during purification may also contribute to the size heterogeneity of PrP 27-30. PrP Does Not Contain Covalently Linked Polynucleotides The ultraviolet absorption spectrum of chromatographically purified PrP 27-30 is shown in Figure 4. The spectral profile is consistent with that of a protein, with a peak at 280 nm. An A280/A260 ratio of 1.41 indicates that there is less than 0.75% nucleic acid contaminating these prepa- rations (Layne, 1957). Assuming a M, of 28,000 for PrP 27-30, we calculate that there must be less than one nucleotide per PrP 27-30 molecule in these preparations. Amino Acid Sequence Studies of PrP 27-30 129 Figure 3. SDS Polyacrylamide Gel Electrophoresis of Fractions Prepared by HPLC Gel Exclusion Chromatography The migration of PrP 27-30 is denoted by the arrow. Lanes l-40 correspond to fractions 61-100 of the chromatogram shown in Figure 2A 0.025 r- 0.020 tfl 0.015 f 2 0 s ( 0.010 0.005 0.0 L 200 240 280 320 360 400 WAVELENGTH (nm) Frgure 4. Ultraviolet Absorption Spectrum of PrP 27-30 Frachons from the HPLC gel filtration column shown to contain only PrP 27-30 by polyacrylamide gel electrophoresis were analyzed by spectro- photometry The ultraviolet spectrum was determined using a Hewfett- Packard 8450 Spectrophotometer equipped with an HP 9872A plotter.

Molecular Weight of PrP A M, of 27,000 to 30,000 for PrP was determined by electrophoresis in three different SDS polyacrylamide gel systems (Prusiner et al., 1982a). By HPLC size exclusion chromatography on a TSK-2000 SW column, PrP 27-30 was found to have a M, of 19,500 in an SDS/sodium phosphate buffer (Figure 5). The elution position of PrP 27-30 was unaltered whether the sodium phosphate buffer contained 0.1% or 1 .O% (w/v) SDS. PrP 27-30 molecules isolated by size exclusion chromatography in the SDS/ sodium phosphate buffer were unaltered in their electro- phoretic migration properties. In other words, chromato- graphically purified PrP 27-30 gave a M, of 27,000 to 30,000 when electrophoresed through SDS polyacrylamide gels. The cause of this discrepancy in M, values between those determined by polyacrylamide gel electrophoresis and those by HPLC size exclusion chromatography re- 2 I I / I I I I 1 150 160 170 IS0 190 200 210 ELUTION VOLUME (ml) Figure 5. Molecular Weight Determination of PrP 27-36 by HPLC Size Exclusion Chromatography TSKX00 SW columns connected rn tandem 60 and 30 cm in length and 2.1 cm In diameter. Chromatograms were developed at 3 ml/min and the effluent monitored at 215 nm. The column buffer contained 150 mM Na phosphate, pH 6.6, and 0.1% SDS. Standard proteins were treated under conditions identtcal wtth those employed for PrP 27-30 except the protein- ase K digestion was omrtted: 1 = collagenase. 105,Mx); 2 = bovine serum albumin, 69,OCO; 3 = ovalbumin, 43,OCQ 4 = carbonic anhydrase, 30,COO; 5 = a-chymotrypsin, 25,CGO; 6 = trypsin inhibitor 21,500; 7 = lysozyme, 14,400, and 8 = RNAase A, 13.700. Arrow Indicates position of PrP 27-30 elution.

mains to be established. The anomalous, delayed elution of asymmetric proteins from size exclusion columns has been reported previously (Nozaki et al., 1976; Meredith and Nathans, 1982). Selective binding of PrP 27-30 to the TSK column matrix resulting in its delayed elution is another possible explanation. Anomalous migration of PrP 27-30 through SDS polyacrylamide gels is still a third possibility for explaining this discrepancy between M, determinations.

Since PrP 27-30 appears to be a sialoglycoprotein, its retarded migration in SDS polyacrylamide gels is certainly a possibility.

Amino Acid Analysis of PrP Samples for amino acid analysis and N-terminal sequence determination were prepared by HPLC size exclusion chro- Cell 130 matography. These samples contained between 300 and 1,000 pmoles of PrP 27-30 assuming a M, of 28,000.

The amino acid composition of PrP 27-30 is given in Table 1. The number of amino acid residues per molecule is based on the M, of 28,ooO for the PrP 27-30 molecule, and assumes 245 amino acid residues; however, this number of amino acids may represent an overestimate due to oligosaccharides, which are covalently bound to PrP 27-30. The most common amino acid residues in PrP 27-30 were glycine, glutamate/glutamine, and aspat-tate/ asparagine. Interestingly, the high content of these amino acids is similar to that reported for proteins in amyloid plaques isolated from the brains of patients dying of Alzheimer’s disease (Allsop et al., 1983). PrP polymers, like amyloid proteins, show green birefringence after stain- ing with Congo red dye when examined by polarization microscopy. This property suggests that the secondary Table 1. Amino Acrd Composrtron of PrP 27-30 Amino Acrd Residues/Molecule’ Asx 27 Thr 17 Ser 11 GIX 27 Pro 6 GIY 27 Ala 16 CYS ND Val 14 Met ia Ile 14 Leu 9 Tyr 15 Phe 6 HIS a LYS 16 Arg 12 TrP N.D.

’ Assumes PrP contarns 245 ammo acids. structure of native PrP probably has extensive regions of P-pleated sheet (Glenner et al., 1974). Our finding that glycine is a common amino acid within PrP 27-30 is consistent with the hypothesis that PrP 27-30 contains substantial amounts of P-pleated sheet secondary struc- ture. Besides amyloids, other proteins with extensive fl- structure also contain high levels of glycine (Earnshaw et al., 1979). N-Terminal Amino Acid Sequence of PrP In order to begin studying the molecular mechanisms responsible for the replication of prions and the biosyn- thesis of PrP, we undertook to determine the amino acid sequence of purified PrP. N-terminal amino acid sequence analysis of PrP 27-30 purified by HPLC gave one predom- inant signal and a number of significant minor signals. The data are summarized in Table 2. Five different preparations of PrP 27-30 were sequenced and all gave essentially the same results. The sequence of the major component is G- Q-G-G-G-T-H-N-Q-W-N-K-P-S-K for the amino-terminal 15 residues.

Analysis of the data led to the interpretation that multiple signals arose from a single amino acid sequence with several N-terminal starting points (Table 3). Presumably, the proteinase K digestions that were used in the purifica- tion generated several N-terminal starting points. These multiple N termini are consistent with the known specificity of proteinase K (Ebeling et al., 1974). The N termini of PrP 27-30 account for more than 90 mole % of the protein applied for sequencing. Thus the protein sequence data unambiguously define a single protein sequence corre- sponding to PrP 27-30. All minor signals can be accounted for by the above interpretation (Table 3).

An extensively purified preparation of scrapie prions containing -log5 IDso U/ml of infectivity was sequenced without further purification of PrP 27-30 (Figure 1). The purified prion fraction from a zonal rotor sucrose gradient was analyzed by SDS polyacrylamide gel electrophoresis as shown in Figure 6. The silver-stained gel shows a single protein. A single sequence was found, identical with the first 15 residues of the major component of PrP 27-30 purified by HPLC size exclusion chromatography (Table 2). The absence of the principal minor component of PrP 27-30 is probably due to the fact that proteinase K diges- Table 2. N-Terminal Amino Acid Sequence of PrP 27-30 and Infectious Scrapie Prions Degradatron Cycle Number 1 2 3 4 5 6 7 a 9 10 11 12 13 14 15 Purified PrP G Q G G G T H N cl w N K P S K P (0.3)b T (0.4) H (0.4) N (0.4) w (0.4) K (0.4) P (0.4) cl (0.2) w (0.2) w (0.2) 0 (0.2) T (0.2) H (0.2) Infectious Prions G Q G G G T H N Q W N K P S K w (0.2) Q (0.2) w (0.1) “The three sequences shown for PrP account for at least 90% of the protein applred to the gas-phase sequencer as determined by amino acid analysrs.

b Numbers m parentheses are relative molar amounts reflectrng the strengths of weak signals in relation to the marn signal, whrch was defined as 1 .O. Am;no Acid Sequence Studres of PrP 27-30 131 tion was used only once in the preparation of this purified prion fraction (Figure 1). A weak signal corresponding to the other minor component in PrP 27-30 was observed in two cycles. These results confirm that the major protein sequence in the infectious prion particle has the N-terminal amino acid sequence given in Tables 2 and 3.

Discussion Scrapie Prions Contain a Single Major Protein The purification of scrapie prions to a nearly homogeneous state has shown that the infectious particle contains one major protein, PrP 27-30 (Prusiner et al., 1983). The protein has been shown to be necessary for infectivity, and its concentration has been shown to be directly proportional to the titer of the infectious particles (McKinley et al., 1983).

The native protein was first identified by taking advantage of its protease-resistant properties (Bolton et al., 1982).

Denaturing the protein renders it sensitive to protease digestion. Denaturation also causes a concomitant de- crease in the prion titer. The secondary structure of native PrP presumably contains a considerable amount of p- pleated sheet since PrP polymers, i.e., prion rods, bind Congo red dye and show green birefringence by polari- zation microscopy (Prusiner et al., 1983). Although X-ray crystallographic studies will be needed to confirm the presence of a p structure, polarization microscopy and the high glycine content of PrP 27-30 support this contention.

When we discovered PrP 27-30 in partially purified preparations of scrapie prions, we observed its microhet- erogeneity with respect to size (Bolton et al., 1982; Prusiner et al., 1982a). Subsequent studies using two-dimensional gel electrophoresis have shown that PrP 27-30 is com- posed of a series of charge isomers (D. C. Bolton and S.

B. Prusiner, unpublished data). The molecular basis for these charge isomers appears to be sialic acid residues, at least in part. In our initial attempts to determine the primary structure of the N terminus of PrP 27-30, we found variability among several amino acids, suggesting a family of proteins. At positions 3, 4, 5, 6, 8, IO, 11, 13, and 14, weaker signals indicating other possible amino acids were observed (Table 2). We have been able to interpret all those weak signals as arising from a single polypeptide chain with “ragged ends,” namely, free N termini starting at different amino acids (Table 3).

Although considerable evidence indicates that PrP 27- 30 is a structural component of the scrapie prion, the Table 3. Interpreted N-Terminal Amino Acid Sequence Showing that PrP 27-30 Contains “Ragged Ends” Relative Amount Ammo Acrd Sequence’ 1 G-Q-GG-G-T-H-N-Q-W-N-K-P-S-K 0.4 X-X-X-T-H-N-X-W-X-K-P 0.2 X-X-P-W-X-Q-X-X-X-T-H-X-Q-W “Single-letter amino acid code. X = ammo acid not determrned at that cycle. remote possibility remains that the active scrapie protein is present at very low levels and purifies with PrP 27-30. If this were the case, then the amino acid sequence of PrP 27-30 would not reflect that of a structural component of the prior-r. We have attempted to minimize this possibility by employing a variety of experimental approaches, but we have been unable to exclude it completely.

Several “strains” of scrapie prions have been described; each produced a different incubation time and histopa- thology (Dickinson and Fraser, 1979). It will be of interest to determine whether or not the different biological prop- erties of prions are manifest within the amino acid se- quence of PrP 27-30.

The Amino Acid Sequence of PrP Is Unrelated to Any Other Known Proteins We have queried computerized data bases containing all known amino acid sequences and translated DNA se- quences in search of sequences that are homologous with PrP 27-30. To date, we have failed to find any proteins with sequences that show any homology (more than 7 amino acids out of 15) to PrP 27-30. However, the known amino acid sequence is less than 10% of the PrP 27-30 molecule and further studies will be needed before a conclusive search can be done in this respect. The behav- ior of PrP 27-30 and its polymerization suggests the possibility that it may, in fact, be related to other filamen- tous proteins. The N-terminal sequence of PrP 27-30 is unlike any known sequence for systemic amyloid proteins (Glenner, 1980). The amino acid sequences of brain amy- loid proteins are unknown (Prusiner, 1984). However, we have failed to find PrP 27-30 in preparations from normal hamsters inoculated with normal brains, and we have failed to show scrapie prion infectivity in such preparations, even though our bioassay is capable of detecting one part in 10’0.

The Mechanism of Prion Replication Remains Enigmatic While considerable progress has been made in identifying and characterizing a protein (PrP 27-30) component of Frgure 6. SDS Electrophoresis of an Extensively Purified Scrapie Pnon Preparation Used for N-Terminal Amino Acid Sequence Determination Drrectly From left to right. lane 1 IS the prion preparatron; lanes 2 and 3 are PrP 27- 30 fractrons prepared by HPLC srze exclusron chromatography. Cell 132 the infectious particle, our studies and those of others have failed to demonstrate the dependence of prion infectivity upon a nucleic acid molecule. Attempts to identify a nucleic acid molecule structurally within extensively purified prep- arations have also been unsuccessful to date. While the biologic properties of the scrapie prions suggest that they ought to be viruses, chemical and physical data continue to suggest that prions lie outside the realm of both viroids and viruses (Alper et al., 1966, 1967; Diener et al., 1982; Prusiner, 1982). However, some investigators still contest this point of view (Rohwer, 1984).

The studies presented here eliminate two important models for the infectious particle previously suggested.

The ultraviolet absorption spectra demonstrate the ab- sence of a covalently linked polynucleotide to the prion protein, PrP 27-30. This finding makes the small nucleo- protein model unlikely. Another possible model for prions also has been eliminated by the amino acid sequence data reported here. That model suggested that PrP was com- prised of a repeating oligopeptide containing only a few amino acids. This was a particularly attractive model, because it could have explained the ability of the infectious particle to replicate without a nucleic acid template, using protein-directed, protein-synthetic mechanisms similar to that for the synthesis of some polypeptide antibiotics (Kleinkauf and von Dohren, 1981). Because we find at least 17 naturally occurring amino acids (Table l), we are forced to conclude that such a mechanism cannot be responsible for replication of PrP 27-30.

Several New Approaches to the Study of Prions Arise from This Work The studies outlined here open the way for two new approaches to studies of scrapie. First, the production of synthetic polypeptides that can be used as antigens to produce antibodies to PrP 27-30, and thus the infectious prion. Second, knowledge of the amino acid sequence has allowed us to construct oligonucleotide probes that can be used to identify cloned cDNAs that encode PrP 27-30. This strategy has been used successfully in the identification of cloned cDNAs that have been reverse- transcribed from mRNAs (Agarwal et al., 1981; Roach et al., 1983). The identification of a PrP 27-30 cDNA may allow us to determine the mechanism of PrP biosynthesis.

The interesting analogies between scrapie prions and brain amyloids may eventually have significance for under- standing the molecular basis of other degenerative dis- eases in which amyloids accumulate, most notably Alz- heimer’s disease (Prusiner, 1984). Knowledge of the N- terminal amino acid sequence of PrP 27-30 provides a chemical basis for the first time to begin investigating the origin and replication of prions.

Experimental Procedures Materials All chemicals were of the highest grades commercially avarlable. All re- agents and solvents used for ammo acid analysis or sequencing were purified as previously described (Hunkapiller et al., 1983). Source and Boarsay of Scraple Prions A hamster-adapted isolate of scrapie prttns was kindly provided by Dr.

Richard Marsh (Marsh and Kimbertin. 1975). The passage history of this material has been previously described (Prusiner et al., 1980).

Scrapie prton titers were determined by an incubation time interval assay (Prusiner et al., 1980, 1982b). Fifty microtiiers of samples to be assayed for prton infectivity were injected intracerebratly into random bred hamsters (LVG/LAK). The time intervals from inoculation to onset of illness and from inoculation to death were used to estimate the titers of the injected samples.

Puffficefiom of krepie Prions Scrapie prtons were purified from hamster brain as shown in Figure 1. The procedure has been described previously in detail (Prusiner et al., 1983).

Briefly, 900-loo0 brains were collected from hamsters sacrificed 70-75 days after intracerebral inoculation with ~10~ IDso U of scraple prfons. The IDso is the mean infectious dose at which 50% of the animals develop scrapie. The frozen brains were homogenized in IO I of 0.32 M sucrose using a polytron. The homogenate was clarified by continuous-flow centrif- ugation and the supernatant fraction extracted with 4% (v/v) Triton X-106 and 2% (w/v) sodium deoxycholate. The extract was precipitated with 8% (w/v) polyethylene glycd 8008 and the precipitate collected by continuous- flow centrifugation. The resuspended precipitate was digested with micro- coccal nuclease followed by proteinase K (100 pg/ml) in the presence of sodium dodecyl sarcosinate (Sarkosyl). Phenytmethylsulfonytfluonde dis- solved in n-propanol was added to a final concentration of 0.1 mM in order to terminate the proteinase K digestion. Cholate (2.0% w/v) was then added and the extract fractionated (NH&SC, at 30% saturation. The precipitate was collected by centrtfugation and then extracted with 2% (v/ v) Trtton X-100 and 0.8% (w/v) SDS prior to centrtfugation on a discontin- uous sucrose gradient in a zonal rotor. A peak of protein and scrapie prion infectivity near the outer edge of the gradient containing approximately 45% sucrose was taken for further purification by HPLC size exclusion chromatography. In one experiment the N-terminal sequence determination was performed directly on a fraction from the leading edge of this peak without further purification. The infectious sample from the preparative sucrose gradient was precipitated with 80% ethanol, resuspended in NH,HCOJ buffer, and sequenced directly.

The protocol generally provided an enrichment of pnon rnfectivtty ranging from 3,ooO to 10,000 fold with respect to protein. The recovery of prions as determined by bioassay varied widely from 10% to 60% due to the imprecrsion of the bioassay and possibly to disaggregation. Purified prep arations contained predominantly one protein, PIP 27-30, as judged by SDS polyacryfamide gel electrophoresis and numerous rods when exam- ined by electron microscopy.

Purification of PrP Two preparative TSK-2000 SW columns 60 and 30 cm in length were connected in tandem. Each column was 2.1 cm in diameter. Sucrose gradient fractions containing extensively purified prions were precipitated with 4 vol of ice-cold methanol and allowed to sit on ice for 20 min. The precipitates were cdlected by centrifugation at 1900 x g for 20 mm and resuspended in IO mM sodium phosphate buffer, pH 7.4. containing 0.2% Sarkosyl at a protein concentration of 1.33 mg/ml. Protein was detenined by the method of Lowry et al. (1951). The suspension was then digested with protenase K (4 MS/ml) at 37°C for 2 hr. The digestion was terminated by adding SDS to 1% and immediately heating at 100°C for 4 min. The sodium phosphate concentration was adjusted to 150 mM before applying 150 ~1 of this sample to the HPLC cdumn. The column system was previously equilibrated with 150 mM sodium phosphate buffer, pH 6.8.

containing 0.1% SDS (w/v).The column was developed at a flow rate of 3 ml/min and the pressure maintained at 18 Atm. A Varian 5000 HPLC pump was used. Fractions (2 ml) were collected after the effluent passed through a Varian spectrophotometer equipped with a 215 nm filter.

Pdyacryiemide Gel Electrophoresis Aliquots from HPLC fractions were diluted 4-fold with 0.2% SDS and precipitated with 0.1 M quinine sulfate in 0.1 N HCI. The precipitates were collected by centrifugation, washed in 80% acetone, and resuspended in sample buffer prepared according to the procedure of Laemmli (1970). The denatured samples were then electrophoresed into 15% polyacrylamide gel and stained wtth srlver (Meml et al., 1981). Ammo Actd Sequence Studies of PrP 27-30 133 Amino Acid Anelysio Alrquots of the protein solutions used for amino acid analysis studies were dried down under vacuum in 4 x 50 mm tubes, 25 cl of redistilled 6 N HCI was added to each tube, and the tubes sealed under vacuum in a 40 x 80 mm tube containing 500 pl of the 6 N HCI. Hydrolysis was carried out at 1 IO’C for 25 hr. The amino acid content of the hydrolyzed protein samples was determined on a modified Dunum D500 analyzer.

Amino Acid Sequencing Alrquots of the same protein solutions used for amino acid analysis were also used for N-terminal sequence anatysrs on a gas-Bquid phase protein sequenator (Hewick et al., 1981). The sample sdution was dried onto a polybrene-coated glass fiber disc in 30 pl aliquots under vacuum until the destred amount had been applied. A 30 PI aliquot of trifluoroacetic acid/ HZ0 (1 :I v/v) was applied to the disc and dried down as the final step.

Conversion of the 2-anilrno-5thioazolinone amino acids to phenyfthiohydan toin (PTH) amino acids was carned out with HCI rn methanol at 52°C. The PTH samples were analyzed by HPLC on a cyanopropyl column as described (Hunkapiller and Hood, 1983) using a modified 21 min program.

The column buffer was composed of 5% tetrahydrofuran in 20 mM Na acetate, pH 5.1. A gradient of acetonitrile without methanol was used to elute the PTH amino acrds.

Acknowledgments The authors thank Todd May, Karen Bowman, Elizabeth Hennessey, Eureka Espanol. and S. Patricia Cochran for expert technical assistance, and Lorraine Gallagher, Rebecca Mead, Susan Mailhot, and Frances Elvin for excellent editorial and administrative assistance. This work was supported by research grants from the National Institutes of Health to S. B. P. and L.

E. H.. as well as by gifts from Ft. J. Reynolds Industries, Inc. and the Shenan Farrchild Foundation.

The costs of publication of this article were defrayed in part by the payment of page charges. Thus article must therefore be hereby marked “a&erTisement” in accordance with 18 U.S.C. Section 1734 solely to Indicate this fact.

Received May 18. 1984; revised June 8, 1984 References Agarwat, K. L., Brunstedt, J., and Noyes, B. E. (1981). A general method for detectron and characterizatron of an mRNA usrng an olrgonuclaotide probe. J. Brol. Chem. 256, 1023-1028.

Allsop. D.. Landon, M.. and Kidd, M. (1983). The isolation and amino acid composition of senile plaque core protein, Brain Res. 259, 348-352.

Alper, T., Haig, D. A., and Clarke, M. C. (1966). The exceptionally small size of the scrapie agent. Brochem. Biophys. Res. Commun. 22, 278-284.

Alper, T., Cramp, W. A., Haig, D. A., and Clarke, M. C. (1967). Does the agent of scrapre replicate without nuclerc acid? Nature 274. 764-766.

Bolton, D. C., McKrnley. M. P.. and Prusiner, S. B. (1982). Identification of a protein that purifies with the scrapie prior?. Science 218. 1309-1311.

Dckinson, A. G., and Fraser, H. (1979). An assessment of the genettcs of scrapie in sheep and mace. In Slow Transmissible Diseases of the Nervous System, 7, S. B. Prusiner and W. J. Hadlow, ads. (New York: Academic Press), pp. 367-386.

Drener, T. O., McKmley, M. P., and Prusiner, S. B. (1982). Vrroids and pnons. Proc. Nat. Acad. SCI. USA 79, 5220-5224.

Drringer, H., Gelderblom, H., Hrlmert, H.. Ozel, M., Edelbluth, C., and Krmberlin. R. H. (1983). Scrapie infectivity, fibrils and low molecular weight protein. Nature 306, 476-478.

Earnshaw, W. C., Goldberg, E. B., and Crowther, R. A. (1979). The distal half of the tail fibre of bacteriophage T4. J. Mot. Biol. 132. 101-131.

Ebeling, W., Hennrich, N., Klockow, M., Metz, H., Orth, H. D., and Lang, H.

(1974). Proteinase K from Tririfachium a/bum limber. Eur. J. Biocham. 47, 91-97.

Galdusek. D. C. (1977) Unconventional viruses and the ongrn and disap- pearance of kuru. Science 197, 943-960. Glenner, G. G. (1980). Amyfoid deposits and amyloidosis. New Eng. J.

Med. 302, 1283-1292, 1333-1343.

Glenner. G. G., Eanes, E. D., Bladen, H. A., Llnke, R. P., and Termine, J.

D. (1974). p-pleated sheet fibrils-a comparison of native amyloid with synthetic protein fibrtls. J. Histochem. Cytochem. 22, 1141-l 158.

Hadlow. W. J., Race, R. E., Kennedy, R. C., and Eklund, C. M (1979).

Natural infection of sheep with scrapie virus. In Slow Transmissible Diseases of the Nervous System, 2, S. B Prusiner and W. J. Hadlow, eds. (New York: Academtc Press), pp. 3-12.

l-fawlck. R. M., Hunkapiller. M. W.. Hood, L. E., and Dryer, W. J. (1981). A gas-liquid solid phase peptide and protein sequenatw. J. Bid. Chem. 256, 7990-7997.

Hunkapiller, M. W., and Hood, L. E. (1983). Analysisof phenylthiohydantoins by ultrasensitive gradient high-performance liquid chromatography. Meth.

Enzymol. 91. 486-493.

Hunkapiller, M W., Hewick, R. M.. Dreyer, W. J.. and Hood, L. E. (1983).

High-sensitivity sequencing with a gas-phase sequenator. Meth. Enzymol.

91,399-413.

Hunter, G. D. (1972). Scrapie: a prototype slow infection, J. Infect. Dis.

125.427-440 Kimberlin, R. H. (1976). Scrapie in the Mouse. (Durham, England: Meadow- field Press), pp. l-77.

Kleinkauf, H., and von Dohren, H. (1981). Nuclerc acid independent synthe- sis of peptides. Curr. Topics Microbial. Immunol. 91, 129-177.

Laemmlr. U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T-4. Nature 227, 680-685 Layne. E. (1957). Spectrophotometnc and turbtdimetnc methods for mea- sunng proteins. Meth. Enzymol. 3, 447-454 Lowry, 0. H.. Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951).

Protein measurement with the Folin phenol reagent. J. Biol. Chem. 793, 265-275.

Marsh, R. F., and Kimberlin, R. H. (1975). Comparison of scrapie and transmissible monk encephalopathy in hamsters. II. Clinical signs, pathology and pathogenesis. J. Infect. DIS. 737. 104-l 10.

Masters, C. L.. Galdusek, D. C., and Gibbs, C. J., Jr. (1981). Creutzfeldt- Jakob dtsease virus isolations from the Gerstmann-Straussler syndrome.

Brarn 104, 559-588 McKinley, M. P., Bolton, D. C., and Prustner, S. B. (1983). A protease- resistant protein is a structural component of the scrapie prion. Cell 35, 57-62.

Meredith, S. C., and Nathans, G. R. (1982). Gel permeation chromatography of asymmetric proteins. Anal. Eiochem. 721, 234-243.

Merril. C. R., Goldman, D., Sedman, S. A., and Ebert, M. H. (1981).

Ultrasensrtrve stain for proteins in polyacrylamide gels shows regional variation in cerebrospinal fluid proteins. Science 211, 1437-1438.

Merz, P. A.. Somerville, R. A., Wisniewskr, H. M., and Iqbal, K. (1981).

Abnormal fibrils from scrapie-infected brarn. Acta Neuropathol. (Berlin) 54, 63-74.

Millson, G. C., Hunter, G. D.. and Kimberlin, R. H. (1976). The physico- chemical nature of the scrapie agent. In Slow Virus Drseases of Animals and Man, R. H. Kimberltn, ed. (New York: American Elsevier Publishing), pp. 243-266.

Nozaki, Y., Schechter, N. M., Reynolds, J. A., and Tanford, C. (1976). Use of gel chromatography for the determination of the Stokes radii of proteins In the presence and absence of detergents. A Reexamination. Biochemistry 15.3684-3690.

Prusiner, S. B. (1982). Novel proteinaceous infectious particles cause scrapie. Science 216, 136-144.

Prusiner, S. B. (1984). Some speculations about prions, amyioid and Alzhermer’s disease. New Eng. J. Med. 370, 661-663.

Prusiner, S. B., Hadlow, W. J., Garfin, D. E., Cochran, S. P., Bannger, J. A., Race, R. E., and Eklund, C. M. (1978). Parttal purification and evidence for multiple molecular forms of the scrapie agent, Biochemistry I7,4993-4999.

Prusiner. S. 6.. Groth, D. F., Cochran. S. P., Masiarz, F. R., McKrnley. M.

P., and Martinez, H. M. (1980). Molecular properties, partial purification and Cell 134 assay by incubation period measurements of the hamster scrapie agent Btochemistry 19, 4883-4891.

Prusiner, S. B., Bolton, D. C., Groth, D. F., Bowman, K. A., Cochran, S. P., and McKinley, M. P. (1982a). Further purification and characterization of scrapie pnons. Biochemistry 21, 6942-6950.

Prusrner, S. B., Cochran. S. P., Groth, D. F., Downey, D. E., Bowman, K.

A., and Martinez, H. M. (1962b). Measurement of the scrapie agent using an incubation time Interval assay. Ann. Neural. 7 1, 353-358.

Prusiner, S. B.. McKrnley, M. P., Bowman, K. A., Bolton, D. C., Bendheim, P. E., Groth, D. F., and Glenner, G. G. (1983). Scrapie prions aggregate to form amylord-like birefringent rods, Cell 35, 349-358.

Roach, A., Boylan, K., Horvath. S., Prusiner, S. B., and Hood, L. E. (1983).

Characterization of cloned cDNA rat myelin basic protein: absence of expresston rn brarn of shtverer mutant mice. Cell 34, 799-806.

Rohwer, R. G. (1984). Scrapie infectious agent IS virus-like in size and susceptibility to inacttvatton. Nature 308, 658-662.

Siakotos, A. N., Gafdusek, D. C., Gibbs, C. J., Jr., Traub. Ft. D., and Bucana, C. (1976). Partial purification of the scrapie agent from mouse brain by pressure disruption and zonal centrifugation in sucrose-sodium chloride gradients. Virology 70, 230-237.