evidence for synthesis of scrapie prion proteins in the ... · that the scrapie isoform of the...

THE JOURNAL OF BIOLOGICAL CHEMISTRY (0 1992 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 267, No. 23, Issue of August 15, pp. 16188-16199,1992 Printed in U.S.A.

Evidence for Synthesis of Scrapie Prion Proteins Endocytic Pathway*

in the

(Received for publication, March 9, 1992)

David R. Borcheltlj, Albert TaraboulosS, and Stanley B. PrusinerSllll From the Departments of $Neurology and YBiochemistry and Biophysics, University of California, S a n Francisco, California 94143

Infectious scrapie prions are composed largely, if not entirely, of an abnormal isoform of the prion protein (PrP) which is designated PrPS'. A chromosomal gene encodes both the cellular prion protein (PrPC) as well as PrPS'. Pulse-chase experiments with scrapie-infected cultured cells indicate that PrPSc is formed by a post-translational process. PrP is translated in the endoplasmic reticulum, modified as it passes through the Golgi, and is transported to the cell surface. Release of nascent PrP from the cell surface by phosphatidylinositol-specific phospholipase C or hydrolysis with dispase prevented PrPS' synthesis. At 18 "C, the synthesis of PrPSc was inhibited under conditions that other investigators report a blockage of endosomal fusion with lysosomes. Our results suggest that PrP" synthesis occurs after PrP transits from the cell surface. Whether all of the PrP molecules have an equal likelihood to be converted into PrPSc or only a distinct subset is eligible for conversion remains to be established. Identifying the subcellular compartment(s) of PrPSc synthesis should be of considerable importance in de- fining the molecular changes that distinguish PrPSc from PrPC.

Many lines of investigation have converged to demonstrate that the scrapie isoform of the prion protein (PrPS")' mediates the transmission and pathogenesis of transmissible neurode- generative diseases including scrapie, bovine spongiform en- cephalopathy, kuru, Gerstmann-Straussler-Scheinker syn- drome, and Creutzfeldt-Jakob disease (for reviews see Refs. 1 and 2). Both PrPS' and the cellular prion protein (PrP') are

* This work was funded by Grants AG02132, NS14069, AG08967, and NS22786 from the National Institutes of Health and the Ameri- can Health Assistance Foundation as well as by a gift from the Sherman Fairchild Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

§ Supported by a grant from the Alzheimer's Disease and Related Disorders Association.

11 To whom correspondence should be addressed Dept. of Neurol- ogy, HSE-781, University of California, San Francisco, CA 94143- 0518.

The abbreviations used are: PrP"', scrapie isoform of the prion protein; PrP, prion protein; PrPC, cellular isoform of the prion protein; GPI, glycoinositol phospholipid; PIPLC, phosphatidylinositol- specific phospholipase C; FITC, fluorescein isothiocyanate; WGA, wheat germ agglutinin; MEM, modified Eagle's medium; DME, Dul- hecco's modified Eagle's medium; HEPES, 4-(2-hydroxyethyl)-l-pi- perazineethanesulfonic acid; PBS, phosphate-buffered saline; DLPC, detergent-lipid-protein complexes; GdnSCN, guanidinium thiocya- nate; HF, hydrofluoric acid, PMSF, phenylmethylsulfonyl fluoride; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electropho- resis.

encoded by a chromosomal, single copy gene (3). The two PrP isoforms are sialoglycoproteins which possess glycoinositol phospholipid (GPI) anchors (4-6). In contrast to PrPC, PrPS' is resistant to digestion by proteases (4) and aggregates into amyloid rods after limited digestion in the presence of detergents (7). PrP" is the only macromolecule, identified to date, that is consistently found in purified preparations of scrapie prions (1).

Although studies with transgenic mice expressing either foreign or mutant PrP genes have yielded important incre- ments of knowledge about prions (8-11), scrapie-infected cultured cells have provided valuable insights into the synthesis of PrPSc (12-16). Many investigators have described cultured cells which are chronically infected with the scrapie agent (17-19). The finding that clones of scrapie-infected cells in culture produce both infectious prions and PrPS' (12, 14, 15) has stimulated research on the cell biology of scrapie. While these scrapie-infected cloned cells produce relatively low levels of prions as measured by bioassays, the levels of PrPS' are sufficient for detection in radioimmunoassays (13).

Earlier studies established that PrP' synthesis begins with translation of the PrP mRNA in the endoplasmic reticulum, followed by modification of the N-linked carbohydrates in the Golgi and transport to the cell surface by the secretory pathway (6, 13, 20-22). At the cell surface, PrP' is attached by a GPI anchor which can be cleaved by phosphatidylinositol- specific phospholipase C (PIPLC) ( 5 , 23).

Although transgenic mouse studies indicate that PrP' and PrPSc may form a complex during the synthesis of nascent PrPS' molecules (1, 8), the subcellular site(s) at which this interaction could occur is unknown. We report here that the recycling of PrP through the endosomal pathway is likely to be important for PrP" synthesis. The release of nascent PrPC molecules from the surface of scrapie-infected cells in culture by PIPLC digestion, or the hydrolysis of PrP' catalyzed by dispase, prevented PrPSc synthesis. While our studies were in progress, other investigators reported the blockade of PrPSc synthesis in cultured cells by digestion with trypsin or PIPLC (24). Although our conclusions are similar to those stated by others (24), our experimental results contrast sharply. For example, we were unable to prevent PrPSc synthesis using PIPLC a t 37 "C. Only when the studies were performed a t lower temperatures could we demonstrate an effect of PIPLC. Similarly, trypsin digestion gave uninterpretable results with respect to PrP" synthesis. Only when we found more gentle conditions for hydrolyzing cell surface proteins using dispase were we able to obtain meaningful evidence. The reasons for these marked differences in experimental results remain to be established.

The results reported here argue that PrP transits to the cell surface before becoming PrPS' and are consistent with recent work that shows that brefeldin A reversibly inhibits

16188

Scrapie Prion Protein Synthesis 16189

PrP" transport to the plasma membrane and PrPS" synthesk2 Our observation that the conversion of cell-surface PrP into PrPsc was reversibjy arrested when cells were shifted to 18 "C implicates endosomal sorting in PrP& synthesis. Cooling cultured cells to 18 "C has been used by many investigators to inhibit the transport of membrane glycoproteins through the endosomal pathway. Endosomes are further implicated by recent work showing that PrPsr accumulates in lysosomes (25, 26).' However, the synthesis of PrPSc appears to be completed before reaching lysosomes.2

MATERIALS AND METHODS

Reagents, Antibodies, and Cell Lines-Cell culture reagents were ohtained from the Cell Culture Facility of the University of California, San Francisco. Methionine-free MEM culture was made from reagents in the Select-amine kit from GIBCO. ['~~SIMethionine (55.1015) was obtained from Amershanl Corp. Detergents for the extraction of cells and the solubilization of P r P were obtained from Sigma. L-@-Lecithin was purchased from Avanti Polar Lipids (Pel- ham, AL). Dispase was purchased from Boehringer Mannheim. FITC- labeled WGA and t.rypsin were purchased from Sigma.

Antibodies used in the immunopurification of P r P were produced in this laboratory. Rabbit P rP antiserum R017 has been shown to react specifically with PrP' and PrP"' molecules (12, 13, 15, 16, 27). Rabbit P rP antisera R066 and R073 are relatively new antisera that specifically recognize 33-35-, 30-, and 26-kDa protease-sensit.ive P r P ~nolecules as well as 27-30., 25-, and 19-kDa protease-resistant PrP'I" molecules (see Fig. 12). Preimmune sera from these rabbits does not react with these P r P molecules (15, 16, 271." Rahbit. antisera R013, R002, and ROO9 were raised against synthetic P r P peptides P1 (amino acids 90-1021, P2 (amino acids 15-40), and P3 (amino acids 220- 2331, respectively. These antisera have been described elsewhere and react, specifically with both PrPC and PrP% molecules (13, 15, 16,28, 29).

The cell lines ScNla and ScHaB are mouse neuroblastoma Nla and Syrian hamster HaB cells that are chronically infected with scrapie and have been described elsewhere (12, 15). We have recently determined that the Nza cells possess Prn-pa PrP genes4 (see (30) for a description of Prn-pa). The scrapie isolate used to infect the N,a cell line is called RML (31), while the isolate used to infect HaB cells is called Sc237 (32). The Sc237 isolate is similar to an isolate called 263K (33). Both RML and Sc237 give short incubation times in their respective hosts (mice 120 days, hamsters 75 days) (9). Both cell lines were maintained in DME containing 10% fetal bovine serum, 100 units/ml penicillin, and 100 mCg/ml streptomycin. ScN,a cells were passaged 1 t,o 10 every 7 days and fed with fresh medium every 2 or 3 days. ScHaB cells were passaged 1 to 5 every 10 days and fed with fresh medium every 3-5 days. When these cultures were passaged in t,his manner then they tended to be more stahle in their production of prpvC.n

Metabolic Radiolabeling of Cultured Cells-Cultures of ScNza and ScHaB cells were allowed to reach the limits of confluence before radiolabeling. Radiolabeling experiments were performed in 25-cm' flasks. Cultures were incubated in labeling medium (methionine-free MEM with 25 mM HEPES) for 1 h, then incubated in fresh labeling medium with 300 pCi/ml [3'SS]methionine as noted in the figure legends. After labeling, the cells were washed in PBS, then either exaracted in detergent or placed in complete medium (DME, 10% fetal bovine serum with 25 mM HEPES) containing unlabeled methionine and allowed to incubate (chase) for the desired time interval.

T o extract radiolabeled cellular proteins, the cells were washed in I'BS and ext,racted at. 4 'C in T N E (50 mM Tris-HC1, pH 7.5, 150 mM NaCl, 5 mM EDTA) containing 0.5% Triton X-100 and 0.5% deoxycholate. The extracts were centrifuged a t 3000 X g for 10 min, then the supernatants were precipitated with a t least 4 volumes of ice-cold methanol. The precipitated proteins were recovered by centrifugation a t 3000 X g for 20 min. In cases where cell extracts were to be digested with proteinase K, the pellets were resuspended in TNE with 0.2% Sarkosyl and digested with 40 pg/ml proteinase K

'Tarahoulos, A., Raeber, A., Borchelt, D. R., Serban, D., and

:i D. Serban, unpublished observations. Y. Zebarjadian and D. Borchelt, unpublished observations. ' A. Tarahoulos and D. Borchelt, unpublished observations.

Prusiner, S. B. (1992) Mol. Biol. Cell, in press.

for 1 h a t 37 "C. The digestion was stopped by the addition of PMSF (5 mM) and precipitat.ion with a t feast 4 volumes of methanol.

~ ~ ~ u ~ ~ ~ u r ~ f ~ c a t ~ o n of PrP-The immunopurification of radiolabeled P r P used previously described methods (13) wit.h the following modifications. To prepare for immunopurification of PrP", the methanol-precipitated proteins from protease-digested extracts were pelleted by centrifugation (3000 X g, 20 min) and resuspended in 3 M GdnSCN containing 20 mM Tris-HC1, pH 7.5. After 10 min, the proteins were once again precipitated with methanol (at, least 10 volumes). This step was added because we determined that GdnSCN denaturation was required for a more quantitative dispersion of PrPS' into detergent-lipid-protein complexes (DLPC) (see Fig. 1) and because GdnSCN-denatured PrPsc is much more immunoreactive (27). We noted that the immunopurification of radiolabeled P r P ' by antisera R073 and RO66 was enhanced when the protease-digested cell extracts were first, denatured in GdnSCN (see Fig. 12, lanes 9- 12).

Both the undigested and protease-digested (GdnSC~-denatured~ cellular proteins were precipitated in methanoi and recovered by centrifugation before dispersion in 0.5 ml of DLPC buffer (20 mM Tris-HCI, pH 8.2, 150 mM NaCI, 2% Sarkosyl, 0.4% L-cu-lecithin). After 2 min of sonication (Lab Supplies Inc., Hicksville, NY) in a glass tube, 2 pl of P rP antiserum (R073) was added, and the extracts were incubated overnight a t 4 "C. The resulting immunocomplexes were then hound to protein A-Sepharose (Pierce Chemical Co. or Pharmacia LKB Biotechnology Inc.) for 1 h at 4 "C, washed for 30 min in DLPC buffer, transferred to a clean Eppendorf tuhe, and washed three more times in 50 mM Tris-HCI, pH 8.2,500 mM NaCI, and 1% Sarkosyl. The immunopurified P r P proteins were then dis- sociated from immunocomplexes by boiling in SDS-PAGE loading buffer and electrophoresed on polyacrylamide gels (34).

Q u ~ n t ~ t a ~ i o n of R a ~ ~ o l ~ b e ~ e d PrPs'--To quantify the production of' PrPsc in our cell lines, we measured the amount of radiolabeled protein that could be immunopuri~ed from metaholicaiiy radio~abeied cells. Immunopur~fied proteins were separated on polyacrylamide gels and quantitated in one of two ways. In initial experiments described in Fig. 2, we determined the optical densit.y of autoradiographic bands created by radiolabeled PrP':' molecules with a densitometer (Bio- Rad) (films were not preflashed before exposure). In later experiments (Figs. 2-12), we used a Molecular Dynamics PhosphoImager device (Sunnyvale, CA) to measure the amount of radiolabeled PrP"c produced by the cell lines. Quantitation was performed as suggested by the manufacturer and employed Imagequant software.

Estimation of the Kinetics of PrP" Biosynthesis-The raw quantitative data (generated by the above procedures) was used to calculate the kinetics of PrPS' biosynt,hesis in ScHaB and ScN'a cells (see Fig. 2). Although three M, species of PrP"' were seen (see Fig. a), we chose to sum the quantitative values for these three molecules and generated a single value for YrP"' at. each time point in the chase. To normalize for variations in different experiments, we determined the maximum value for PrPsc in each experiment, then converted all the values for a given experiment to a percentage of that maximum value.

The percentage values for time points between 0 and 24 h of chase were entered into a computer program that calculat,es a best fit curve on the assumpt,ion that the production of PrP" follows single expo- nential kinetics. The program also generates a k constant for the kinetics of PrP" production that was used to calculate a t l i2 for PrP" biosynthesis with the formula t l p 2 = In 2/k .

Quantitation of PrP" Dispersion into DLPC-We examined the sedimentation properties of PrP"' molecules in high speed centrifugation to measure the efficiency of PrPSc solubilization in DLPC. Proteinase K-digested extracts were either dispersed directly into DLPC or first denatured in 3 M GdnSCN with 20 mM Tris-HCI, pH 7.5 [see lmmunopurification of PrP). After 2 min of sonication in a bath sonicator, t,he mixtures were divided. One part of the sample was redigested with proteinase K, 20 pg/ml for 1 h a t 37 "C, while the other was maintained on ice. Each sample was then centrifuged a t 40,000 rpm in a Beckman TLA 100.3 rotor for 1 h a t 4 "C. The supernatant was recovered and PMSF (5 mM) was added before precipitation with 10 volumes of ice-cold methanol/chloroform (21). The supernatant proteins were recovered by centrifugation (3,000 X g ) and denatured in GdnSCN before PAGE. The pellets from the high speed centrifugation were resuspended in 3 M GdnSCN with 20 mM Tris-HCI, pH 7.5, and 5 mM PMSF, then precipitated with 10 volumes of methanol. Each supernatant and pellet fraction was examined by PAGE and immunohlot analysis with P rP antiserum ROT3 as previously described (13).

16190 Scrapie Prion Protein Synthesis

RESULTS

Studies of PrP"' biosynthesis require quantitative immunopurification of radiolabeled PrP"'. Because PrP"' resists solubilization in detergents, the immunopurification of PrP"" has proven to be difficult (35, 36). However, PrP"' can be functionally dispersed into DLPC (36), and this method was used in a previous study of PrP"' biosynthesis in ScN,a cells (13). By denaturing PrP"' in GdnSCN prior to dispersion in DLPC, we found that the solubilization of PrP"' could be increased considerably (Fig. 1).

To examine the solubility of PrP"", we determined the sedimentation properties of the protein in DLPC during high speed centrifugation (60,000 X g). We found that most or all of the protease-digested PrP"' from cell extracts that were denatured in 3 M GdnSCN remained in the supernatant fraction of DLPC after centrifugation (Fig. 1, A and B, lanes .5 and 6). In contrast, most of the PrP"' found in cell extracts that were dispersed directly in DLPC pelleted during centrif-

A 1 2 3 4 5 6 7 8

- 66

- 43

- 31

- 21

B 1 2 3 4 5 6 7 8

- 66 - 43

- 31

- 21

FIG. 1. Denaturation in GdnSCN enhances the solubility of PrPSc i n DLPC. ScHaB and ScN,a cultures were detergent extracted and digested with proteinase K as described under "Materials and Methods." The digested extracts were then either dispersed directly in DLPC or first denatured in 3 M GdnSCN containing 20 mM Tris- HCI, pH 7.5, hefore dispersion in DLPC. After 2 min of sonication in a bath sonicator, the mixtures were divided. One part of the sample was redigested with proteinase K, 20 pg/ml for 1 h a t 37 "C, while the other was maintained on ice. Each sample was then centrifuged a t 40,000 rpm in a Beckman TLA 100.3 rotor for 1 h a t 4 "C. The supernatant was recovered and adjusted to 5 mM PMSF before precipitation with 10 volumes of ice-cold methanol/chloroform (2:l). The high speed pellets were resuspended in 3 M GdnSCN with 20 mM Tris-HCI, pH 7.5, and 5 mM PMSF, then precipitated with 10 volumes of' methanol. Each fraction was examined by PAGE and immunohlot analysis with P rP antiserum R073. Panel A, ScHaB cells. Panel H , ScN,a cells. Lanes 1 and 2, the supernatant and pellet ot'extracts that were dispersed directly in DLPC. Lanes 3 and 4 , same as lanes 1 and 2 except the DLPC mixtures were digested with proteinase K prior to centrifugation. Lanes 5 and 6, the supernatant and pellet of cell extracts that were denatured in GdnSCN before dispersion in DLPC. Lanes 7 and 8, same as lanes 5 and 6 except the DLPC mixtures were digested with proteinase K prior to centrifugation. The 19-, 2%. and 27-30-kDa PrP"' bands are not recognized by preimmune serum from rabbit R073 (not shown).

ugation (lanes 1 and 2) , indicating aggregation. PrP"' in GdnSCN-treated extracts lost its resistance to proteases as expected (27) ( A and B, lanes 7 and 8), indicating complete denaturation (37). In contrast, most of the PrP"' that was dispersed directly into DLPC remained protease resistant ( A and B, lanes 3 and 4). Based on these results, we chose to denature PrP"' in GdnSCN before dispersion in DLPC to achieve a more consistent and quantitative solubilization of PrP"' for subsequent immunopurification.

Kinetics of PrP"' Formation in ScN,a and ScHaB Cells- Previous studies in scrapie-infected cell cultures have shown that ScNsa and ScHaB cells produce 27-30-, 25-, and 19-kDa protease-resistant PrPSC molecules that are not found in their uninfected progenitors (N2a and HaB cells) (12, 13, 15). We have also reported that PrP"' molecules are derived from protease-sensitive prion proteins that slowly acquire protease resistance (13,16). Using the improved methodology for PrP"' solubilization described above, we measured the kinetics of PrP"" formation in both ScNpa and ScHaB cells by pulse- chase radiolabeling experiments with [:"S]methionine. To identify radiolabeled PrP"', cells were extracted in detergent, digested with proteinase K (40 pg/ml), and then denatured in 3 M GdnSCN, 20 mM Tris-HC1, pH 7.5, before dispersion into DLPC and immunopurification.

In both ScNfa and ScHaB cells, no protease-resistant PrP"' molecules were recovered when cells were harvested immediately after a 1-h labeling period instead only protease sensitive PrP was recovered (Fig. 2, A and B, lanes 1 and 2 ) as expected (13, 15). Radiolabeled protease-resistant PrP"' was detected in extracts of ScHaB cells by 1 h of chase in unlabeled medium (Fig. 2 A , lane 3) , with greater amounts detected after 2 and 4 h (lanes 4 and 5 ) . No further increase in the amount of radiolabeled PrP'" was seen in cultures that were chased for 4,6, 10, and 24 h (lanes 5-8). In the ScN2a cells, we found that increasing amounts of radiolabeled PrP"' were immunopurified a t each chase time point between 1 and 6 h (Fig. 2B, lanes 3-6), while stable levels were seen in cultures chased for 6, 10, and 24 h (lanes 6-8).

Autoradiographs shown in Fig. 2 suggested that the production of PrPSr reaches a maximum in 2-4 h in ScHaB cells and 6-8 h in ScN2a cells. To quantify the kinetics of PrP"' formation in the ScNna cells and ScHaB cells, we measured the amount of radiolabeled PrP"' that was immunopurified after various intervals of chase with a Bio-Rad densitometer or a Molecular Dynamics PhosphoImager (see "Materials and Methods"). The quantitative values for four experiments with each cell line were fit to a single curve, generating a tllS for acquisition of protease resistance by P rP molecules of about 1 h for ScHaB cells (Fig. 2C) and 3 h for ScNPa cells (Fig. 20) . Of note, these values describe the behavior of a pool of radiolabeled P rP molecules that were translated during the 1- h pulse. We believe that the estimated tIl2 = 3 h value for ScNpa cells that we report here is more accurate than our initial estimate of tIl2 = 15 h (13) because of the improved methods used for solubization (Fig. 1) and analysis of a larger data set.

In both cell lines, the amount of radiolabeled P r p ' recovered after a prolonged chase represented -5-10% of the amount of nascent P rP arguing that most of the nascent PrP behaves as PrP" and is degraded (also see Fig. 3 and Refs. 13 and 16). Since both the ScNPa and ScHaB cell lines are clonal, and most of the cells in these cultures produce both PrPC and PrPSC (13, 16), the foregoing results do not seem to be due to only a fraction of the cells in the culture producing PrP*. Thus, only a minority of the PrP is converted into PrP". Whether all of the PrP molecules have an equal likelihood to


A I 2 3 4 5 6 7 8 B i 2 3 4 5 6 7 6

1 FR

66 -

43- ii 66 -

43 -

31 - 31 -

21 - 21 - - """._

C S M a a Cells

+ Hours of Chase

D ScN2a Cells

25

FIG. 2. The kinetics of PrP"' biosynthesis in ScNaa cells and ScHaB cells. ScN2a and ScHaB cells were metabolically radiolabeled for 1 h as described under "Materials and Methods." Excess [''"Slmethionine was removed by washing the cells in PRS before a chase in unlabeled medium. Radiolabeled P rP molecules were identified by immunopurification with antiserum ROT3 as described under "Materials and Methods." To identify radiolabeled PrP"' molecules, cell extracts were digested with proteinase K and denatured in 3 M GdnSCN before dispersion in DLPC and immunopurification with P r P antiserum R073. This antiserum recognizes the same radiolabeled molecules as other P r P antisera (see Fig. 12), and preimmune serum R073 does not recognize any protease-resistant molecules (not shown). Panrl A, ScHaH cells. Panel R, ScNsa cells. Lane I is a control in which cells were labeled then P rP was extract.ed without protease digestion. Lanrs 2-8 contain protease-resistant PrP" molecules recovered from one confluent 25-cm2 flask of cells after a 1-h labeling and increasing chase times. Lane 2, no chase; lane 3, 1 h chase; lane 4 , 2 h chase; lane 5 , 4 h chase; lane 6,6 h chase; lane 7, 10 h chase; lane 8, 24 h chase. Panel C , ScHaR cells. Panel D, ScN2a cells. We quantitated the amount of radioactive P r P in each lane of the experiments shown in panels A and B with a molecular Dynamics PhosphoImager device (W). We also included in the analysis three experiments with both cell lines that were quantitated by measuring the optical density of autoradiographic bands (0, 0 , O ) . The values for the three PrPS' specific bands (27-30, 25, and 19 kDa) were summed, then plotted as a percentage of the maximum attainable value for a given experiment. The values for the four experiments were combined and fit to a single curve with a previously described computer program (13) that calculates a best fit to a first order reaction. The calculated tIr2 for the acquisition of protease resistance by PrPS' was 1 h for ScHaB cells and 3 h for ScN2a cells.

be converted into PrP"' or only a distinct subset is eligible for conversion remains to be established.

The Conversion of PrP into PrP" Is a Temperature-dependent Process-PrPc is sialoglycoprotein that is bound to the cell surface by a GPI anchor (5, 6), and its transport through the secretory pathway is relatively rapid (13,22). To determine if PrPSc biosynthesis involves membrane traffck- ing events, we studied the synthesis of PrP"' a t 18 "C. This temperature has been used by many investigators to examine the recycling and degradation of membrane proteins in cultured cells (38-45).

In both cell lines, the rate of P rP degradation was slowed at 18 "C (Fig. 3, lanes 3 and 4 ) . However, the two cell lines showed pronounced differences in their ability to synthesize PrP" at 18 "C. ScHaB cells metabolically radiolabeled a t 37 "C for 30 min produced as much PrP"' during a chase a t 18 "C as they did a t 37 "C (Fig. 3A, lanes 6 and 7), but ScNza cells chased a t 18 "C could not synthesize PrP"' (Fig. 3B, lane 7) .

Since the formation of PrP"' in ScHaB cells was unaffected at 18 "C, we asked whether ScHaB cells produce PrP"' via a pathway that completely lacks a 37 "C-dependent event. We found that ScHaB cells that were both labeled and chased at 18 "C (Fig. 4) could not produce PrP"' (lane 6). Control ScHaB cultures that were labeled a t 18 "C but chased a t 37 "C did produce PrP"' (lane 4 ) . One explanation for the lack of inhibition of PrP"' synthesis in ScHaB cells a t 18 "C after labeling a t 37 "C could be that PrP molecules undergo some 37 "C-dependent processing before the end of the labeling period. This explanation is consistent with our observations

that the kinetics of PrP" biosynthesis in ScHaB cells are more rapid than those of the ScNza cells (Fig. 2).

Characterization of the 18 "C Block in PrP" Formation in ScNza Cells-To characterize further the 18 "C block in PrP"' synthesis, confluent cultures of ScNza cells were metabolically radiolabeled at 37 "C for 30 min then chased for varied intervals (0, 1, 2, or 3 h) a t 37 "C before a shift to 18 "C and an additional chase for 20 h. Cultures that were held at 37 "C for 1 h after the initial labeling before a chase for 20 h a t 18 "C, produced twice as much PrP"' (Fig. 5A, lane 5 ) as control cultures which were chased for only 1 h at 37 "C after the initial labeling (lane 6) or were directly placed at 18 "C (lane 4 ) . Similarly, cultures that were held at 37 "C for 2 h then chased a t 18 "C were found to produce three to four times more PrP"' (lane 7 ) than cells that were chased for only 2 h a t 37 "C (lane 8 ) or chased a t 18 "C only (lane 4 ) . These results, in conjunction with those from studies of ScHaB cells, indicate that the biosynthesis of PrP" involves a temperature- dependent event that occurs shortly after the translation of P rP polypeptide chains, and suggests the possibility that two distinct events feature in the synthesis of PrP"'. The first event represents a commitment to the PrP"' pathway while the second involves the acquisition of protease resistance.

The inhibition of PrP"' production that occurred when ScNPa cells were chased a t 18 "C was reversible (Fig. 5B) . Confluent cultures were metabolically radiolabeled a t 37 "C for 30 min then chased a t 18 "C for 20 h before a second chase a t 37 "C for varied intervals (1, 4, and 10 h). The kinetics of PrP"' formation after a shift from 18 to 37 "C (lanes 7-9) were similar to those of control cells that were placed at 37 "C


- 66

- 43

-31

-21

- ~ -

- 66

- 43

-31

-21

- + " + + + Proteinase K 10 d Exposure 2 d

FIG. 3. ScHaB and ScNPa cells differ in their ability to produce protease-resistant PrP"' molecules when labeled at 37 "C and chased at 18 "C. Upperpanel, ScHaR cells. Lowerpanel, ScN'a cells. Confluent 2S-cm' flasks of cells were metabolically radiolabeled for 30 min and prepared for chase as described in Fig. 2. Lane I, no chase incubation. Lane 2, no chase incubation after digestion with proteinase K. Lane 3, chased a t 37 "C for 20 h. Lane 4 , chased at 18 "C for 20 h. To prevent overexposure of some lanes, the autoradiograph shown for lanes 1-4 are from a 2-day exposure, while lanes 5-7 were exposed for 10 days. Lane 5 is a 10-day exposure of lane 2. Lane 6, chased a t 37 "C for 20 h and digested with proteinase K. Lane 7, chased a t 18 "C for 20 h and digested with proteinase K.

immediately following the labeling (lanes 10-12). In both instances, there was little formation of PrP"' during the first hour at 37 "C (lanes 7 and I O ) with greater amounts of PrP"' found after 4 h (lanes 8 and 11) . While little or no synthesis of PrP"' occurred during the 20-h chase period a t 18 "C, the pool of radiolabeled PrP did decrease by -50% (lanes 1 and Z ) , demonstrating the turnover of PrP". Under these conditions, a concomitant reduction in PrP" synthesis was observed (lanes 7-9) as compared to controls (lanes 10-12). Since this 50% reduction was highly reproducible, these findings suggest that PrP destined to become PrP"' can be degraded if the acquisition of protease resistance is prevented.

To determine whether 18 "C slows the endocytosis and vesicular compartmentalization of membrane glycoproteins (38, 39) in our ScNoa cells, we incubated cells with FITC- labeled WGA, which binds to glycoproteins on the plasma membrane. When FITC-WGA was incubated with cells a t 37 "C, well-defined vesicles of varying sizes appeared in the cytoplasm (Fig. 6, panel 1 ). In contrast, cells incubated with FITC-WGA a t 18 "C displayed a diffuse staining in the cytoplasm as well as staining on the cell surface (panel 2). Control cells that were held a t 4 "C with FITC-WGA were diffusely stained over the entire cell surface with no apparent inter- nalization (panel 3 ) . These results indicate that the endocytosis and transport of membrane glycoproteins is slowed a t 18 "C, but not completely inhibited.

PrP Transport to the Cell Surface Occurs Prior to Acquisi- tion of Protease Resistance-To determine whether P rP is

1 2 3 4 5 6

- 66 - 43

- 31

- 21

FIG. 4. The production of PrP" in ScHaB cells is blocked when cells are labeled at 18 "C and chased at 18 "C. Confluent 25-cm' flasks of ScHaR cells were metabolically radiolaheled for 2 h a t 18 "C as descrihed in Fig. 2. Excess [:'.S]methionine was removed by washing the cells in PRS before a chase at 37 or 18 "C (the chase medium was held a t 4 "C before addition to the cells and incubation at the specified temperature). Radiolabeled PrP and PrP'" were identified as described in Fig. 2. Lanes I and 2, PrP immunopurified from cells that were labeled a t 18 "C for 2 h without. and with proteinase K digestion, respectively. Lanes 3 and 4 , PrP immunopurified from cells that were labeled a t 18 "C and chased a t 37 "C for 6 h without and with proteinase K digestion, respectively. I ~ n e s 5 and 6, PrP immunopurified from cells that were labeled a t 18 "C and chased a t 18 "C for 6 h without and with proteinase K digestion, respectively.

transported to the cell surface before it is converted to PrP" and acquires protease resistance, we took advantage of earlier observations which showed 1) nascent P rP could be released from ScNPa cells by PIPLC (13), and 2) the formation of PrP"' is slowed a t 18 "C (see Fig. 3). ScN,a cells were labeled a t 37 "C and chased for 20 h at 18 "C in the presence of PIPLC, and then shifted to 37 "C for 10 h. Under these conditions almost no PrP"' was produced (Fig. 7A, lanes 6 and 8) in contrast to experiments without PIPLC (Fig. 5B). As shown, the 18 "C block of PrP"' production was reversible in the cultures not exposed to PIPLC (Fig. 7A, lanes 5 and 7). PIPLC was active a t 18 "C since -80% of the radiolabeled nascent PrP could be released into the medium (Fig. 7R, lanes 6-8; Fig. 7C). ScNaa cells that were preincubated at 18 "C for 20 h with PIPLC then labeled a t 37 "C and chased at 37 "C produced as much PrPS' as control cells that were not preincubated with the enzyme (Fig. 70 ) . To determine if PrP remains accessible to PIPLC throughout a long incubation at 18 "C, ScN,a cells were labeled a t 37 "C, then chased with or without PIPLC for 16 h a t 18 "C, followed by a second chase for 5 h a t 18 "C with or without PIPLC as noted (Fig. 8). We observed that 90% of the PrP" in cells held for 16 h a t 18 "C could still be released by PIPLC during the second 6-h chase (Fig. 8B, lane 3 ) . We conclude that PrPC remains in a PIPLC accessible compartment a t 18 "C and that incubation of ScNpa cells with PIPLC at 18 "C (Fig. 7) prevents PrP" synthesis. These results argue that the precursor of PrP"' is accessible to PIPLC digestion on the cell surface prior to the acquisition of protease resistance.

If similar experiments with PIPLC were performed at 37 "C, we found that the precursor of PrP" was less accessible to PIPLC (Fig. 9A). Confluent cultures of ScNla cells were radiolabeled with [""Slmethionine for 30 min at 37 "C in the presence of PIPLC, then chased for 1 h at 37 "C with PIPLC before an additional chase of 16 h at 37 "C. ScN,a cells that were labeled and chased in this manner produced substantial


8 1 2

2 3 4 5 6 7 8 9 1 0 - " . . " . .

-66 -43

-31

-21

3 4 5 6 7 8 9 1 0 1 1 1 2 ~

-66 -43

-31

-21

FIG. 5. Characterization of the 18 "C block in PrPS' formation in ScN2a cells. Confluent 25-cm' flasks of ScNla cells were metabolically radiolabeled for 30 min a t 37 "C, prepared for chase, and radiolabeled P rP and PrP" molecules identified as in Fig. 2. All lanes contain immunopurified P rP from extracts of cells that were manipulated as described below. Panel A, lane 1 is a 2-day autoradiograph of an extract of cells that were neither chased nor digested with proteinase K. Lane 2 is a 10-day exposure of lane 1. The remaining lanes 3-10 were exposed for 10 days and contain PrP"' molecules immunopurified from proteinase K-digested extracts of cells that were chased as follows: lane 3, none; lane 4 , 20 h a t 18 "C; lane 5, 1 h at 37 "C then 20 h a t 18 "C; lane 6, 1 h a t 37 "C; lane 7,2 h at 37 "C then 20 h a t 18 "C; lane 8 , 2 h a t 37 "C; lane 9 , 3 h a t 37 "C then 20 h at 18 "C; lane 10, 3 h a t 37 "C. Panel R, lane I is a 2-day autoradiograph of PrP immunopurified from cells that were neither chased nor digested with proteinase K. Lane 2 is a 2-day autoradiograph of PrP immunopurified from cells that were chased for 20 h at 18 "C but not digested with proteinase K. Lanes 3-12 are all 10- day autoradiographs of P rP immunopurified from cells treated as described below: lane 3 is a 10-day autoradiograph of lane 1. Lane 4 , I'rP immunopurified from cells not chased but digested with proteinase K. Lane 5 is a 10-day autoradiograph of lane 2. Lane 6 , PrP immunopurified from cells chased 20 h a t 18 "C and digested with proteinase K. Lanes 7-9, P rP immunopurified from cells chased for 20 h at 18 "C then shifted to 37 "C for 1, 4, and 10 h, respectively, followed by digestion with proteinase K. Lanes 10-12, PrP immunopurified from cells chased for 1, 4, and 10 h a t 37 "C, respectively, followed by digestion with proteinase K.

amounts of PrP"' (lane 8 ) as compared to controls (lanes 4 and 9). No PrPS' was produced during the first 1.5 h of labeling and chase without or with PIPLC (lanes 3 and 7) as expected. Controls showed that radiolabeled nascent P rP could be released by PIPLC digestion into the culture medium at 37 "C (lane 6) while the medium from control cells that were not exposed to PIPLC contained no P rP (lane 2) . These results indicate that the P rP which is destined to become PrP"' is largely insensitive to PIPLC digestion at 37 "C, suggesting that it may be rapidly endocytosed and protected from PIPLC-catalyzed hydrolysis. Our results contrast with those of other investigators who report that ScN,a cells incubated with PIPLC at 37 "C blocks the formation of PrP"' (24).

To determine whether all species of mature PrPS' possess

1

37°C

18" C

3 1 1

I 4OC

FIG. 6. Endocytosis of WGA in into ScNza cells is slowed at 18 "C. ScN'a cells were plated onto 8-well microscope culture slides and allowed to adhere overnight. The cella were washed with PBS at 4 "C twice then immersed in Opti-MEM a t 4 "C with FITC-labeled WGA and incubated a t 4, 18, or 37 "C. After 1 h, the cells were washed with PBS a t 4 "C twice then immersed in Opti-MEM a t 4 "C and incubated for 16 h a t their prior incubation temperature of 4, 18, or 37 "C. After two washes in PBS at 4 "C, the cells were fixed in PBS a t 4 "C with 10% formalin for 1 h and photographed in ultraviolet light. Frame I, cells incubated a t 37 "C. Frame 2, cells incubated at 18 "C. Frame 3, cells incubated a t 4 "C.

a GPI anchor modification (Fig. 9B), radiolabeled PrP"' molecules were treated with aqueous HF, which hydrolyzes the phosphodiester bonds within GPI anchors leaving glycosidic bonds intact (46). Aqueous HF treatment increased the migration of all three species of PrP"' in SDS-PAGE compared to controls (Fig. 9B, lanes 1 and 2) . This shift in SDS-PAGE migration indicates a loss of mass and suggests that all three PrPS' species possess GPI anchors. Unglycosylated PrP"' that was synthesized in cells in the presence of tunicamycin (16) showed a similar shift in SDS-PAGE migration after treatment with aqueous HF (lanes 3 and 4 ) .

PrP Is Endocytosed Before Acquiring Protease Resistance- Our observation that the PrP destined to become PrP"' is less accessible to PIPLC at 37 "C than at 18 "C suggests that PrP may be endocytosed at 37 "C before acquiring protease resistance. Consistent with this postulate, nonspecific endocytosis in our ScN,a cells was found to be slower a t 18 "C (Fig. 6), and the formation of PrP"' was inhibited (Figs. 3-5). TO provide additional evidence that PrP is transported to the cell surface and then endocytosed before acquiring protease resistance, we exposed labeled ScN,a cells to proteases at varied intervals after the initial labeling period. While others have recently demonstrated that trypsin digestion of nascent PrP" reduces PrP" synthesis (24), we found that trypsin digestion did not readily digest PrP" on the surface of living cells.6 In addition, we could not determine whether PrPS' synthesis was ~~

' D. Borchelt and A. Taraboulos, unpublished observations.

16194 Scrapie Pr ion Protein Synthesis

.................... m 18'c 18'C 18°C 18'c

' L a n e 1 2 3 4 5 6 7 8

A $ m rf ._ rn D

5 6 7 8

Cells (C) . .

-66

-43

-31

-21

- 66 - 43 -31 -21

-31

-21

Pulse 30' .' (37°C)

D Lane

......

1 I "I" 2 3 4 5 1 6 - 66

- 43

-31

-21

Proteinase K - + + - t + .

FIG. 7 . The precursor of PrP"' is transported to the cell surface before acquiring protease resistance. Confluent cultures of ScNpa cells were metabolically radiolabeled for 30 min. After the labeling, cells were washed with PBS then immersed in Opti-MEM with or without PIPLC (200 milliunits/ml) as noted below, then chased at 18 "C for 20 h. The 18 "C-chase medium was harvested and soluble proteins were precipitated with a t least 4 volumes of methanol (-20 "C) after the media was adjusted to 1 X T N E from a 10 X stock (see "Materials and Methods"). P r P molecules in the medium were then identified by immunopurification. The cells were washed with I'RS then immersed in DME base medium with HEPES and chased for 10 h at 35 "C. Radiolabeled P rP or PrP"' molecules were identified as described in Fig. 2. Panel A, cell extracts were digested with proteinase K to identify radiolabeled PrP"'. Lane I, extract from cells not chased. Lane 2, extract from cells chased for 20 h a t 18 "C. Lane 3 , extract from cells chased for 20 h at 18 "C with PIPLC. Lane 4, extract from cells chased for 10 h a t 37 "C. Lane 5 , extract from cells that were chased for 20 h a t 18 "C t.hen shifted to 37 "C for 10 h. Lane 6, extract from cells chased for 20 h at 18 "C with PIPLC and then shifted to 37 "C for 10 h. Lane 7, repeat of experiment in lane 5. Lane 8, repeat of experiment in lane 6. Panel B, all lanes are P r P immunopurified from cell culture medium of cells shown in panel A, lanes 5-8. Lanes 5 and 7, culture medium from cells after 20 h at 18 "C without PIPLC. Lanes 6 and 8, culture medium from cells after 20 h a t 18 "C with PIPLC. Gel was autoradiographed for 5 days. Panel C, lane 1, extract of cells not chased. Lane 2, extract of cells chased for

1 2 3

t -

- +

0

c 1 2 3

-43 -31

-21

-43 -31 -21

-43

-3 1

-21

FIG. 8. PrPC remains accessible to P I P L C at 18 "C. T o determine whether PrP" remains on the cell surface during a long (16 h) chase at 18 "C, we metabolically radiolabeled ScN,a cells for 30 min a t 37 "C as described in Fig. 2, then chased cells a t 18 "C in Opti- MEM with or without PIPLC as noted. After the first chase, the medium was harvested, and cells were chased a second time at 18 "C for 5 h with or without PIPLC as noted. P rP molecules from the medium and cell extracts were immunopurified with R053 in DLPC as described under "Materials and Methods." Panel A, medium from the first chase a t 18 "C for 16 h. Lane I, without PIPLC. Lane 2, with PIPLC. Lane 3, without PIPLC. Panel H, medium from the second chase a t 18 "C for 5 h. Lane I , without PIPLC. Lane 2, without PIPLC. Lane 3, with PIPLC. Panel C, extracts of cells. Lane C, control cells that were not chased. Lane I , chased without PIPLC. Lane 2, chased with PIPLC in the first chase only. Lane 3, chased with PIPLC in the second chase only. The sharp hands seen a t 34 kDa in lanes 2 and 3 are not P rP specific and appear to hind immunoglobulin.

inhibited as a result of trypsin digestion of cell surface protein or as a result of nonspecific effects such as dislodging and reattachment of cells. In contrast, dispase digested PrP molecules on the cell surface, and it did not dislodge the ScNna cells from the surface of cell culture flasks. By using dispase which allowed the ScNza cells to remain attached to the surface of the cell culture flasks, we avoided the potential problems attendant with cell attachment, which might influ- ence the synthesis of PrPS'. A further of advantage of dispase is that it is easily inhibited by EDTA.

When ScNna cells were labeled at 37 "C and chased for 30 min in the presence of 10 mg/ml dispase a t 37 "C, PrP was degraded (Fig. 10, lune 5) . Cells exposed to dispase under these conditions did not synthesize PrP" after removal of the

20 h a t 37 "C. Lane 3, culture medium of cells from lane 2. Lane 4 , extract of cells chased for 20 h at 18 "C. Lane 5, culture medium from cells in lane 4 . Lane 6, extract of cells chased for 20 h at 18 "C in medium with PIPLC. Lane 7, culture medium from cells in lane 6. Gel was autoradiographed for 5 days. Panel D, control experiment demonstrating that incubation of cells at 18 "C in the presence of PIPLC does not prevent PrP" synthesis in ScNta cells. Cells were preincuhated a t 18 "C for 20 h with PIPLC (lanes 4-6) and then labeled as described in Fig. 2. Lanes 1-3, no preincubation. Lanes 1 and 4, extracts of cells not chased. Lanes 2 and 5, extracts of cells not chased hut digested with proteinase K. Lane 3 and 6, cells chased a t 37 "C for 10 h followed by proteinase K digestion.


-66 -43

-31

-21

-43 "31

FIG. 9. PrP"' and its precursor PrP molecules possess GPI anchors. Panel A, the precursor of PrP" rapidly becomes inaccessi- ble to PIPLC at 37 "C. ScN2a cells were metabolically radiolabeled lor 30 min at 37 "C in the presence of PIPLC (200 milliunits/ml). The labeling medium was harvested and prepared for immunopurification as described in Fig. 7 before the cells were washed in PBS then chased for 1 h at 37 "C in Opti-MEM with PIPLC (200 milliunits/ ml). The chase medium was harvested, prepared for immunopurification, and combined with the labeling medium. Some cultures were washed with PBS after the first 1-h chase then chased for an additional 16 h at 37 "C. Radiolabeled PrP and PrP" molecules were identified as described in Fig. 2. Lanes I, 2, 5 , and 6, which contain total radiolabeled nascent PrP, were autoradiographed for 2 days. Lanes 3 and 4 and 7-9, which contain radiolabeled PrP"', were autoradiographed for 10 days. Lane I , extract of cells chased for 1 h. Imze 2, culture medium from cells in lane 1. Lane 3, extract of cells chased for 1 h followed by proteinase K digestion. Lane 4 , extract of cells chased for 16 h followed by proteinse K digestion. Lane 5, extract of cells labeled for 30 min and chased for 1 h with PIPLC present. Imne 6, culture medium from cells in lane 5 . Lane 7, extract of cells treated as in lane 5 with proteinase K. Lane 8, extract of cells labeled for 30 min and chased for 1 h with PIPLC present followed by an additional chase for 16 h; extracts digested with proteinase K. Lane 9 is a repeat of lane 4 . Panel R, all species of PrP" that are produced by ScN,a cells possess a GPI anchor modification. Confluent cultures of' ScNsa cells were preincubated in methionine-free medium in the absence or presence of 30 pg/ml of tunicamycin for 1 h, then metabolically radiolabeled in the absence or presence of 30 pg/ml of tunicamycin with 300 pCi/ml [""Sjmethionine for an additional hour. The cells were then chased in unlabeled medium overnight, detergent extracted, digested with proteinase K, and precipitated with 10 volumes of methanol. After centrifugation (3,000 X g for 25 rnin), the pellets were resuspended in DLPC buffer (no sonication) and centrifuged a t 60,000 X g for 1 h. The pellet, which was highly enriched for I'rP'', was denatured in GdnSCN and reprecipitated in methanol as described under "Materials and Methods." This protocol partially purifies the PrPsc molecules (not shown) and is required to reduce the formation of insoluble aggregates after treatment in aqueous HF. The partially purified PrPsc was resuspended in aqueous HF (48% v/ v) and divided into 2 aliquots. One aliquot served as a control and was immediately diluted with 4 volumes of H20 and desiccated in a Savant speed vacuum until dry. The other aliquot was incubated on ice for at least 36 h but not more than 48 h, then diluted, and dried. PrP"' molecules were then immunopurified in DLPC as described under "Materials and Methods." Lane I , control extract. Lane 2, extract incubated in aqueous HF for 36 h. Lane 3, extract of cells labeled in the presence of tunicamycin. Lane 4 , same as lane 3 after treatment with aqueous HF for 36 h.

Dispase 30

..,....,.... .... ..... 1 2 3 4 5 6 , 8 9 1 0 1 1

?

::

- 66

-43 v)

u n 0

-31

-21

- + - + - f t - t t f Proteinase K

FIG. 10. The PrP that will become PrP'" becomes inacces- sible to media-soluble dispase before acquiring resistance to detergent-solubilized proteinase K. Confluent cultures of ScN2a cells were metabolically radiolahed for 30 min and prepared for chase as described in Fig. 2. The labeled cells were then immediately exposed to dispase (10 mg/ml in Opti-MEM) for 30 min a t 37 "C, or allowed to chase for 2 h a t 37 "C before exposure to the protease. After dispase digestion, the cells were washed twice in PBS then harvested immediately, or allowed to chase for 16 h a t 37 "C in complete medium. P rP molecules were immunopurified from detergent extracts of cells as described in Fig. 2. The top pane/ is from a 3-dav autoradiograph while the lower panel was exposed for 30 days. Lane 1, extract of cells not chased. Lane 2, extract of cells not chased followed by proteinase K digestion before immunopurification. Iane 3 , extract of cells chased for 16 h. Lane 4 , extract of cells chased for 16 h followed by proteinase K digestion before immunopurification. Lane 5 , extract of cells that were digested with dispase immediately following the labeling period. Lane 6, same as lane 5 except extracts were digested with proteinase K before immunopurification of PrP. Lane 7 is from cells that were digested with dispase immediately following the labeling period then chased for 16 h at 37 "C before harvest and digestion with proteinase K. Lanes 8 and 9 are from cells that were held a t 37 "C for 2 h then digested with dispase without and with proteinase K digestion before immunopurification. Lane IO is from cells that were held at 37 "C for 2 h then digested with dispase then chased for 16 h a t 37 "C before harvest and digestion with proteinase K. Lane I 2 is from cells that were chased for 16 h then digested with dispase before harvesting and proteinase K digestion.

dispase and a further chase a t 37 "C for 16 h in complete medium (lune 7). Cells chased a t 37 "C for 2 h (lanes 8-10) prior to digestion with dispase for 30 min did produce PrP"' which was measured after a further chase a t 37 "C for 16 h (lune 10). Of note, PrP was largely degraded after the dispase treatment (lane 8), and little PrP" formation was detected at the end of the dispase treatment (lune 9). Dispase exposure did not degrade the PrP"' formed after a long chase (16 h) at 37 "C (lune 11 ) and was inactive against newly radiolabeled P rP in lysis buffer containing detergent and EDTA (not shown). These observations suggest that some PrP may become protected from the dispase prior to the acquisition of protease resistance during PrP"' formation and that PrP may be endocytosed before acquiring protease resistance.

Synthesis of Unglycosylated PrP"'-The extent of glycosylation of membrane proteins serves as a marker for protein transport through the endoplasmic reticulum and Golgi. Both ScHaB and ScNPa cells produce PrPS' molecules with a molecular mass of 27-30, 25, and 19 kDa. The 19-kDa form of PrP"" comprises -25% of the total amount of PrP"' produced in ScNPa cells and is unglycosylated (16, 47). We found that all glycoforms of mature PrP"" appear to be derived from molecules that are accessible to PIPLC and dispase, and thus probably reach the cell surface. To investigate the origin of the various PrP"' glycoforms, we compared the rate of PrP synthesis and glycosylation in infected cell lines (ScN2a and ScHaB) to that in uninfected controls (N2a and HaB) (Fig.


11). PrP molecules were rapidly synthesized and glycosylated to yield molecules of 30 and 33 kDa, with a broad smear up to 40 kDa in both the infected and uninfected cell lines (unglycosylated full-length PrP has a mass of 26 kDa, see (16, 21, 22)). These results argue that unglycosylated PrP"' is not produced as a result of altered glycosylation in scrapie-infected cells.

That prion infection does not alter the glycosylation of P rP was also demonstrated using antibody binding of newly radiolabeled PrP and mature protease-digested PrP"' (Fig. 12). We consistently recovered glycosylated 33-35-kDa species of newly radiolabeled PrP from ScN2a cells with antisera that were raised against any one of three PrP peptides (Pl, residues 90102; P2, residues 15-40; P3, residues 220-233) (29) (Fig. 12A, lunes 1-6), and with three polyclonal P rP antisera de- noted R017, R073, and R066 that were raised against the protease-resistant core fragment of Syrian hamster PrPSC designated PrP 27-30 (lunes 7-12). All of the antisera described above, except P2 antiserum, immunopurified all three species of mature protease-digested PrP"' (Fig. 12B). This was expected, since the P2 epitope is located at the NH, terminus of full-length PrP and is not part of the protease- resistant core of PrP"' (3, 29, 48, 49). We conclude that the 27-30-, 25-, and 19-kDa species of PrP"' share similar NH2 and COOH termini but differ in their degree of N-linked glycosylation, with the 19-kDa species being unglycosylated (16, 21).

In summary, our results show that the predominant species of newly radiolabeled PrP in ScN2a cells have masses of 33- 35 and 30 kDa and appear to be glycosylated, but the mature forms of radiolabeled PrP"' tend to be less glycosylated. Although PrPS' could be deglycosylated as it accumulates, we do not see any evidence of this in the pulse-chase experiments shown in Fig. 2. All three species of PrP"' maintain a relatively constant ratio even after a prolonged chase of 48 h (not shown). That unglycosylated P rP can acquire protease resistance is in accord with earlier studies demonstrating PrP"' in ScN,a cells treated with tunicamycin or expressing recombinant PrP with mutated N-linked glycosylation sites (16).

DISCUSSION

Radiolabeled PrP" molecules do not appear until hours after synthesis of the full-length polypeptide, which is com-

A B 1 2 3 4 1 2 3 4

~ . . . -

66 - 43 -

31 -

21 -

.66

.43

4 -31

- 21

FIG. 11. Comparison of the early events in PrP biogenesis in infected cell cultures to that of uninfected cultures. Con- fluent cultures of N2a, ScN,a, HaB, and ScHaB cells were incubated in methionine-free medium for 1 h then metabolically radiolabeled for 5 or 10 min with 300 pCi/ml of [:%]methionine. The cultures were cooled to 4 "C immediately after the labeling period then extracted as described under "Materials and Methods." Panel A , N,a and ScN?a cells. Lanes I and 2, N2a cells that were metabolically radiolabeled for 5 or 10 min, respectively. Lanes 3 and 4 , ScNaa cells that were metabolically radiolabeled for 5 or 10 min. Panel I?, HaB and ScHaB cells. Lanes I and 2, HaB cells labeled for 5 or 10 min, respectively. Lanes 3 and 4, ScHaB cells labeled for 5 or 10 min.

1 2 3 4

1 2 3 4

5 6

5 6

7 8 9 10 11 12

-66

-43

- 31

- 21

-66

-43

- 31

, -21

FIG. 12. Undermodified species of PrP preferentially acquire protease resistance. One set of ScN,a cells were metabolically radiolabeled for 1 h and extracted immediately. A second set of cultures was allowed to chase (40 h) after a 1-h metabolic radiolabeling, then detergent extracted, and digested with proteinase K (see "Materials and Methods"). Each of the extracts was then divided. One part of the sample was denatured in 3 M GdnSCN with 20 mM Tris-HCI, pH 7.5, then precipitated with 10 volumes of methanol. Both the untreated and the GdnSCN-denatured extracts were dispersed in DLPC and further aliquoted for immunoaffinity purification with different PrP antisera. Lanes 1-6 contain PrP molecules recovered from nondenatured (lanes I , 3, and 5 ) and GdnSCN-denatured (lanes 2, 4 , and 6) extracts by immunopurification with antiserum raised against synthetic PrP peptide P1 (amino acids 90-102) (lanes 1 and 2 ) , PrP peptide P2 (amino acids 15-40) (lanes 3 and 4 ) , and PrP peptide P3 (amino acids 220-223) (lanes 5 and 6 ) . Lanes 7- 12 show untreated (lanes 7,9, and 11) and GdnSCN-denatured (lanes 8, IO, and 12) PrP molecules immunopurified by polyclonal PrP antiserum from rabbit R017 (lanes 7 and R ) , rabbit R073 (lanes 9 and IO), and rabbit R066 (lanes I I and 12). Upper panel, extracts of ScN2a cells harvested immediately after the labeling period. Lower panel, proteinase K-digested extracts of ScN,a cells harvested after an overnight chase.

plete within a few minutes (Table I). These results argue that PrP'" molecules acquire their protease resistance during post- translational processing events. P rP chains transit from the endoplasmic reticulum through the Golgi and are transported to the cell surface before PrPSC is formed. Those PrP molecules that are destined to become PrP"' appear transiently on the cell surface and can be released with PIPLC (Fig. 7) (24) or hydrolyzed by dispase (Fig. 10). Consistent with these results are other studies showing that brefeldin A reversibly inhibits PrP"' synthesis., Brefeldin A has been found to inhibit the transport of proteins out of the endoplasmic reticulum and Golgi apparatus as well as modifying the endosomal pathway (50-54). Clearly, PrP"' synthesis requires nascent P rP chains to transit from the endoplasmic reticulum through the Golgi and onto the cell surface.

Most of the studies described here involved immunopurification of radioactive PrP"' from metabolically labeled ScNPa cells. Quantitative comparisons of PrPS' synthesis were made

Scrapie Pr ion Protein Synthesis 16197

TABLE I Summary of tth values of PrP biosynthesis and degradation in normal

and scrapie-infected cell lines

Parameter measured

Normal cells Scrapie-infected cells

N2a HaB ScN2a ScHaB

Synthesis of full <2.5 min" <2.5 min" <2.5 min" <2.5 min" length 26- kDa PrP

Acquisition of None None' 3 h I h PrP protease resistance

Degradation of 6 hb 6 h' >24 h >24 h PrPd "Since radiolabeled PrP can be detected in these cells in 5-min

pulse radiolabelings, the t, for the synthesis of PrP must be no more than 2.5 min. We cannot be certain that the precursor to PrP" is synthesized at the same rate since so little of the radiolabeled pool of PrP molecules present after 1 h can eventually acquire protease resistance.

Ref. 13. Refs. 15, 16. These values pertain to PrPC in uninfected cells and PrPS' in the

infected cell lines. The degradation rate of PrPC in scrapie-infected cells is unknown but previous studies indicate that the rate does not differ radically from that of uninfected cells (13).

e A. Taraboulos and D. R. Borchelt, unpublished observations.

between replicate cultures that were separately radiolabeled, extracted, protease digested, and immunopurified. Some differences between samples probably arose from variability in the efficacy of one or more of these procedures. For example, we sometimes observed 2-fold differences in the amount of radiolabeled PrP" that was recovered from replicate experimental cultures. The variability in PrPSc radiolabeling is illustrated in Fig. 2 where data on the kinetics of PrPSc synthesis in four separate experiments are plotted. In general, 4-5-fold changes in the amount of PrPSc synthesis were con- vincing and judged as meaningful. We found that the synthesis of PrP" was dramatically reduced when nascent PrP" was digested with PIPLC at 18 "C, but digestion of nascent PrPSc with PIPLC at 37 "C did not reproducibly diminish the synthesis of PrP'" by more than a factor of 2. We also observed that digestion of nascent PrPC with dispase significantly reduced the synthesis of PrPSc and that this phenomenon could be partially prevented by delaying the exposure of cells t o dispase. While trypsin digestion (2 mg/ml for 20 min at 37 "C) of nascent PrPC did on occasion cause a large decrease in PrPsc synthesis (not shown), these conditions were so harsh that we could not determine if the reduction in PrPS' synthesis was due to digestion of cell surface protein or to nonspecific effects of the treatment, such as the dislodging and reattachment of the cells. Dispase proved to be a more gentle protease that did not dislodge cells from culture vessels.

While there seems to be agreement, in general, on the sequence of events that occur during the synthesis of both PrPC and PrPSc, we remain concerned about substantial differences between our studies employing PIPLC and protease digestions to block PrPS" synthesis and those reported by Caughey and Raymond (24). These investigators also reported experiments in which they used Iodogen-coated coverslips to catalyze the radioiodination of PrPC on the surface of cells and followed its conversion into PrPSc during a 24-h chase. However, approximately 50% of the total amount of iodinated PrPSc was found immediately after labeling for 5 min indicating that some PrPS' was accessible to the labeling procedures (24). We previously reported that -10% of steady-state PrPS' in ScNpa cells was labeled with sulfo-NHS-biotin, a reagent

thought to label only cell surface proteins (55). Whether PrP" was labeled with these reagents because they enter cells through free-flow endocytosis, because some PrPSc is exposed to the cell surface, or because these cultures contain a signif- icant number of dead permeable cells, is unknown. The reasons for the discrepancies between the experimental results reported by us here and those of others (24) are unclear. Whether these differences can be attributed to variations in ScNza subclones remains to be established. Is P r P Derived from a Specific Precursor Pool?-Whether

PrPS' is synthesized from a subset of P rP molecules or all PrP" molecules are eligible for conversion remains to be established. Less than 10% of the radiolabeled nascent PrP molecules that have been synthesized by the end of a 1-h metabolic radiolabeling pulse are converted into PrPSc (Fig. 2). At present, we have not been able to identify any differences between those P rP molecules which are destined to become PrP" and the pool of PrPC molecules. Both the PrPSc precursor molecules and PrP" can be released from the surface of cells by PIPLC digestion or hydrolyzed by dispase (Figs. 7-10). Furthermore, both PrP" and those PrP molecules destined to become PrPS' appear to be susceptible to cellular degradation (Fig. 5).

Synthesis of P r p from Unglycosylnted PrP-ScN2a cells produce three species of PrP" that differ in their degree of N-linked glycosylation (Fig. 12, Ref. 16). In our experiments, we have been unable to uncouple the formation of one partic- ular glycoform of PrPSc from another. These results suggest that all three species of PrPSc traverse the same biosynthetic pathway.

Earlier studies showed that unglycosylated P rP produced in the presence of tunicamycin was converted to unglycosylated PrP" (16), but unglycosylated PrPC is not readily detected at the cell surface Recombinant P rP molecules lacking consensus sites for N-linked glycosylation were not detected on the cell surface (56) but could be converted to unglycosylated PrPSc in ScNza cells (16). Although these results suggest that unglycosylated P rP need not transit to the cell surface before conversion into PrP", we cannot eliminate the possibility that a small fraction of PrP was transported to the cell surface in these experiments and was converted to PrPS' upon re-entry into the cell. Alternatively, in the presence of tunicamycin, P rP may be transported directly from the endoplasmic reticulum or Golgi to the endocytic pathway where it may be converted to PrPSc prior to entering the lysosomes (16, 57).

Where Does PrP Acquire Protease Resistance?-Since infectious prions are composed largely, if not entirely, of PrP"' molecules (1, 8), it is important to identify the site of PrPS' synthesis and define the molecular events involved in this process. Our observations argue that PrPSc synthesis occurs within the endocytic pathway. Exogenous prions could initiate infection by entry through the endocytic pathway to stimulate the conversion of PrPC molecules from cell surface to PrPSc.

Multiple lines of evidence suggest that PrP may acquire protease resistance in the endocytic pathway. First, when ScNpa cells were exposed to dispase immediately after the labeling pulse, no PrPS' was formed during the chase at 37 "C (Fig. 10). In contrast, when ScNza cells were chased for 2 h at 37 "C prior to dispase exposure, some PrP became inacces- sible to the dispase in the media and subsequently acquired resistance to proteinase K (Fig. 10). Second, when ScNaa cells were chased at 18 but not at 37 "C, PrP was released by PIPLC digestion and PrPSc synthesis was abolished (Figs. 7

D. Borchelt, M. McKinley, and A. Taraboulos, unpublished observations.


and 9A). In our hands, PIPLC catalyzes the release of PrPC from the cell surface relatively slowly; thus, we found that long digestions with PIPLC at 18 "C were required to remove those molecules destined to become PrP". At 37 "C, radiolabeled P rP molecules probably transit to the surface and are endocytosed too rapidly for PIPLC digestion to be completely effective. As noted above, our results with both PIPLC and trypsin differ from those reported by others (24), and the basis for these discrepancies is unknown. Third, the formation of PrP"" in ScN2a cells was inhibited when the chase was performed at 18 "C (Fig. 3), and PrP remained largely accessible to PIPLC throughout a long 18 "C chase (Figs. 7 and 8). As predicted from the work of other investigators, endocytosis and vesicular compartmentalization of membrane glycoproteins was retarded at 18 "C in ScN2a cells (Fig. 6). Fourth, if ScNsa cells were held at 37 "C for 1-2 h of chase prior to shifting to 18 "C, then some PrPS' was formed at 18 "C (Fig. 5A). Presumably, some PrP was endocytosed at 37 "C and acquired protease resistance via a process that is less temperature-dependent.

Many investigators have observed that the endosomal transport of membrane glycoproteins to lysosomes is inhibited at 18 "C (38-45). We confirmed that our ScN2a cells behave in a similar manner by demonstrating that endocytosis of FITC-WGA occurred a t 18 "C but the formation of well- defined cytoplasmic vesicles was greatly inhibited as compared to 37 "C (Fig. 6). We conclude that 18 "C probably inhibits endosome function in ScN2a cells as reported for other types of cells.

Our findings suggest that PrPsc acquires protease resistance in the endocytic pathway. Many studies have shown that most membrane glycoproteins are transported to lysosomes via endosomes, which act as receptacles for plasma membrane proteins that are destined for degradation or recycling (42, 57-59). After endosomes have received membrane glycoproteins from endocytic vesicles they appear to mature as they change from "early" to "late" endosomes before delivery of their contents to lysosomes (57).

Many membrane proteins are degraded after endocytosis and transit through endosomes to lysosomes (40-42, 45, 60). Although PrPsc accumulates in cytoplasmic vesicles, many of which are secondary lysosomes (7), recent studies argue that PrPSc is formed prior to entry into lysosomes. First, while lysosomotropic amines block the digestion of the NH2-termi- nal 90 amino acids of PrPS', they do not interfere with the formation of PrPS' (25, 64). Second, lysosomotropic amines do not alter substantially the degradation of PrPC (64), suggesting that it might be degraded before reaching lysosomes. Third, recent kinetic studies indicate that PrPSc acquires protease resistance approximately 1 h before exposure to lysosomal proteases and digestion of the NH2 terminus (47).

Several caveats force our conclusions to be tentative with respect to the site at which PrP" is converted to PrPS'. First, although it is likely that PIPLC and dispase prevent PrPS' synthesis by digestion of PrPC at the cell surface, we cannot exclude the entry of these enzymes into cells prior to digestion of PrP". Second, we cannot exclude the possibility that PrP" is converted at the cell surface into PrPS'. Perhaps PrP enters the endocytic pathway as a result of an event that initiates PrP"' synthesis at the cell surface to produce a molecule that resists PIPLC and dispase digestion, but has not yet acquired all the properties of PrP". Third, multiple routes of PrPC trafficking during conversion into PrP" might occur (16). Fourth, we do not know how either PrPC or the PrP" precursor enters the endocytic pathway. Some GPI-anchored proteins are internalized via cholesterol-rich membrane patches

called caveolae (61, 62). Alternatively, PrPC could enter the endocytic pathway via free flow endocytosis (63). While each of these issues requires further resolution, the current studies argue that the conversion of PrPC into PrP" is likely to occur in the endocytic pathway.

Acknowledgments-We thank Dan Serhan for maintaining stocks of PrP antisera and for his efforts to document the specificity of these antisera for PrP molecules. We also thank Neil Stahl and Mike Scott for stimulating discussions. The efforts of Lorrie Gallagher and Van Nguyen in preparing this manuscript are also appreciated.

REFERENCES 1. Prusiner, S. B. (1991) Science 2 5 2 , 1515-1522 2. Weissmann, C. (1991) Nature 352,679-683 3. Basler, K., Oesch, B., Scott, M., Westaway, D., Walchli, M., Groth, D. F.,

McKinley, M. P., Prusiner, S. B., and Weissmann, C. (1986) Cell 4 6 ,

4. Oesch, B., Westaway, D., W Ichli, M., McKinley, M. P., Kent, S. B. H., 417-428

Aebersold, R., Barry, R. A,, Tempst, P., Teplow, D. B., Hood, L. E., Prusiner, S. B., and Weissmann, C. (1985) Cell 40,735-746

5. Stahl, N., Borchelt, D. R., Hsiao, K., and Prusiner, S. B. (1987) Cell 5 1 , 229-240

6. Endo, T., Groth, D., Prusiner, S. B., and Kobata, A. (1989) Biochemistry 28.8380-8388

7.

8.

9.

10.

11.

12.

McKiiey,M. P., Meyer, R., Kenaga, L., Rahbar, F., Cotter, R., Serban, A,, and Prusiner, S. B. (1991) J. Virol. 6 5 , 1440-1449

Prusiner, S. B., Scott, M., Foster, D., Pan, K.-M., Groth, D., Mirenda, C., Torchia, M., Yang, S.-L., Serban, D., Carlson, G. A., Hoppe, P. C.,

Scott, M., Foster, D., Mirenda, C., Serban, D., Coufal, F., W Ichli, M., Westaway, D., and DeArmond, S. J. (1990) Cell 63,673-686

Torchia, M., Groth, D., Carlson, G., DeArmond, S. J., Westaway, D., and Prusiner, S. B. (1989) Cell 59,847-857

Westaway, D., Mirenda, C. A., Foster, D., Zeharjadian, Y., Scott, M., Torchia, M., Yang, S. L., Serban, H., DeArmond, S. J., Ebeling, C.,

Hsiao, K. K., Scott, M., Foster, D., Groth, D. F., DeArmond, S. J., and Prusiner, S. B., and Carlson, G. A. (1991) Neuron 7, 59-68

Butler, D. A,, Scott, M. R. D., Bockman, J. M., Borchelt, D. R., Taraboulos, Prusiner, S. B. (1990) Science 2 5 0 , 1587-1590

A., Hsiao, K. K., Kingsbury, D. T., and Prusiner, S. B. (1988) J. Virol.

13. Borchelt, D. R., Scott, M., Taraboulos, A,, Stahl, N., and Prusiner, S. B.

14. Caughey, B., Neary, K., Butler, R., Ernst, D., Perry, L., Chesebro, B., and

15. Taraboulos, A,, Serban, D., and Prusiner, S. B. (1990) J. Cell Biol. 110 ,

62,1558-1564

(1990) J. Cell Biol. 1 1 0 , 743-752

Race, R. E. (1990) J. Virol. 6 4 , 1093-1101 Ol" O l O O

16. Taraboulos, A., Rogers, M., Borchelt, D. R., McKinley, M. P., Scott, M., Serban, D., and Prusiner, S. B. (1990) Proc. Natl. Acad. Sci. U. S. A. 8 7 , 8262-8266

17. Clarke, M. C. (1979) in Slow Transmissibk Diseases of the Neruous System (Prusiner, S. B., and Hadlow, W. J., eds) Vol. 2, pp. 225-234, Academic Press, New York

18. Markovits, P., Dormont, D., Delpech, B., Court, L., and Latarjet, R. (1981)

19. Race, R. E., Fadness, L. H., and Chesebro, B. (1987) J. Gen. Virol. 6 8 , C. R. Acad. Sci. Paris 293,413-417

20. Bolton, D. C., Meyer, R. K., and Prusiner, S. B. (1985) J. Virol. 53,596- 1391-1399

21. Haraguchi, T., Fisher, S., Olofsson, S., Endo, T., Groth, D., Tarantino, A., 606

Borchelt, D. R., Teplow, D., Hood, L., Burlingame, A,, Lycke, E., Kobata, A., and Prusiner, S. B. (1989) Arch. Biochem. Biophys. 2 7 4 , 1-13

22. Caughey, B., Race, R. E., Ernst, D., Buchmeier, M. J., and Chesebro, B. (1989) J. Virol. 6 3 , 175-181

23. Stahl, N., Borchelt, D. R., and Prusiner, S. B. (1990) Biochemistrv 29.

Lll I"L1JL

26.

27.

28.

29.

30.

31. 32. 33. 34. 35.

36.

37.

38. 39.

40.

5405-5412 , ~ . .

24. Caughey, B., and Raymond, G. J. (1991) J. Biol. Chem. 266,18217-18223 25. Caughey, B., Raymond, G. J., Ernst, D., and Race, R. E. (1991) J. Virol.

McKinley, M. P., Taraboulos, A,, Kenaga, L., Serban, D., Stieber, A,, 65,6597-6603

DeArmond, S., Prusiner, S. B., and Gonatas, N. (1991) Lab. Inuest. 6 5 , w u n n

Serhan, D., Taraboulos, A,, DeArmond, S. J., and Prusiner, S. B. (1990)

Barry, R. A., Kent, S. B. H., McKinley, M. P., Meyer, R. K., DeArmond, Neurology 4 0 , 110-117

S. J., Hood, L. E., and Prusiner, S. B. (1986) J. Infect. Dis. 153 , 848- 8.54

"" """

Barry, R. A,, Vincent, M. T., Kent, S. B. H., Hood, L. E., and Prusiner, S. Westaway, D., Goodman, P. A,, Mirenda, C. A., McKinley, M. P., Carlson,

Chandler, R. L. (1961) Lancet 1,1378-1379 Marsh, R. F., and Kimberlin, R. H. (1975) J. Infect. Dis. 1 3 1 , 104-110 Kimberlin, R., and Walker, C. (1977) J. Gen. Virol. 34 , 295-304 Laemmli, U. K. (1970) Nature 227,680-685 Gabizon, R., McKinley, M. P., and Prusiner, S. B. (1987) Proc. Natl. Acad.

Gabizon, R., McKinley, M. P., Groth, D. F., and Prusiner, S. B. (1988)

Bolton. D. C., McKinley, M. P., and Pruslner, S. B. (1984) BiochPmktrv

" ~

B. (1988) J. Immunol. 140 , 1188-1193

G. A,, and Prusiner, S. B. (1987) Cell 51,651-662

Sei. U. S. A. 84,4017-4021

Proc. Natl. Acad. Scl. U. S. A . 8 5 , 6617-6621 . ""

23,5898-5906 Salzman, N. H., and Maxfield, F. R. (1989) J. Cell Bid. 109 , 2097-2104 Griffiths, G., Hoflack, B., Simons, K., Mellman, I., and Kornfeld, S. (1988)

Hare, J. F., and Huston, M. (1984) Arch. Biohem. Biophys. 2 3 3 , 547-555 Cell 52,329-341

41. 42.

43. 44. 45.

46.

47.

48.

49.

50.

51. 52.

53.

Scrapie Prion Protein Synthesis 16199 Hare, J. F. (1988) J. Biol. Chem. 263,8759-8764 Dunn, W. A,, Huhhard, A. L., and Aronson, N. N., Jr. (1980) J. Biol. Chem. 54. Wood, S. A,, Park, J. E., and Brown, W. J. (1991) Cell 67,591-600 255,5971-5978 55. Stahl, N., Borchelt, D. R., and Prusiner, S. B. (1990) in Molecular and Cell

Weigel, P. H., and Oka, J. A. (1981) J. Biol. Chem. 256,2615-2617 Marsh, M., Bolzau, E., and Helenius, A. (1983) Cell 32, 931-940

Biology of Membrane Proteins (Turner, A. J., ed) pp. 189-216, Ellis

Wolkoff, A. W., Klausner, R. D., Ashwell, G., and Harford, J. (1984) J. Cell 56. Rogers, M., Tarahoulos, A,, Scott, M., Groth, D., and Prusiner, S. B. (1990) Horwood Limited, Chichester, United Kingdom

Ferguson, M. A. J., Homans, S. W., Dwek, R. A,, and Rademacher, T. W. 57. Stoorvogel, W., Strous, G. J., Geuze, H. J., Oorschot, V., and Schwartz, A.

Tarahoulos, A,, Raeher, A,, Borchelt, D., McKinley, M. P., and Prusiner, 58. Marsh, M., and Helenius, A. (1980) J. Mol. Biol. 142, 439-454

Prusiner, S. B., Groth, D. F., Bolton, D. C., Kent, S. B., and Hood, L. E. 60. Steinman, R. M., Mellman, I . S., Muller, W. A,, and Cohn, Z. A. (1983) J.

Hope, J., Multhaup, G., Reekie, L. J. D., Kimherlin, R. H., and Beyreuther, 61. Rothherg, K. G., Ying, Y., Kolhouse, J. F., Kamen, B. A,, and Anderson,

Misumi, Y., Miki, K., Takatsuki, A,, Tamura, G., and Ikehara, Y. (1986) J. 62. Anderson, R. G. W., Kamen, B. A,, Rothherg, K. G., and Lacey, S. W.

Doms, R. W., Russ, G., and Yewdell, J. W. (1989) J. Cell. Biol. 109, 61-72 63. Tooze, J., and Hollinshead, M. (1991) J. Cell Biol. 116, 635-653 Lippincott-Schwartz, J., Yuan, L. C., Bonifacino, J. S., and Klausner, R. 64. Taraboulos, A., Borchelt, D., McKinley, M. P., Raeber, A., Serban, D., and

Lippincott-Schwartz, J., Yuan, L., Tipper, C., Amherdt, M., Orci, L., and Prusiner, S. B. (1991) Prion Diseases in Humans and Animals Sympo- sium, September 2-4, 1991, London (abstr.)

Klausner, R. D. (1991) Cell 67,601-616

Biol. 98,375-381 Clycobiology 1, 101-109

(1988) Science 239,753-759

S. B. (1991) FASEB J. 5, A1177

(1984) Cell 38, 127-134

K. (1988) Eur. J . Biochem. 172,271-277 R. G. W. (1990) J. Cell Biol. 110,637-649

Bid. Chem. 261, 11398-11403

L. (1991) Cell 65,417-427

59. Willingham, M. C., and Pastan, I. (1980) Cell 21, 67-77

Cell Biol. 96, 1-27

(1991) Science 255,410-411

D. (1989) Cell 56,801-813

evidence for synthesis of scrapie prion proteins in the ... · that the scrapie isoform of the...

Documents