Jun 12, 1989 - profile, and stability, Dickinson and coworkers (12-14) have reported that more ..... becomes fixed. Bruce and Dickinson (13) have reported a.
Proc. Natl. Acad. Sci. USA Vol. 86, pp. 7475-7479, October 1989 Genetics
Primary structure of prion protein may modify scrapie isolate properties (epigenetics/neurodegenerative disease/mouse genetics/infectious disease/incubation period)
GEORGE A. CARLSON*, DAVID WESTAWAYt, STEPHEN J. DEARMONDt, MARILYN PETERSON-TORCHIAt, AND STANLEY B. PRUSINERt§ *McLaughlin Research Institute, 1625 3rd Avenue North, Great Falls, MT 59401; and Departments of tNeurology, :Pathology, and §Biochemistry and Biophysics, University of California, San Francisco, CA 94143
Communicated by Elizabeth S. Russell, June 12, 1989
and onset of disease (incubation time) provided an additional approach to the study of prions (6). The importance of this approach was corroborated by the finding that the Prn-pa allele of NZW mice, in which the scrapie incubation time is short (referred to as short-incubation time mice) encodes PrP molecules that differ by two amino acids from those found in long-incubation time I/Ln mice (Prn-pb) (7). By using allelespecific oligonucleotide probes encompassing codons 108 and 189, all short-incubation time Prn-pa mice encoded leucine and threonine at these positions, whereas a phenylalanine and valine were substituted in each of the six long-incubation time Prn-pb strains (refs. 7 and 8; G.A.C., unpublished results). Though it has not been proven that control of scrapie incubation time is a pleiotropic effect of Prn-p (8), it is clear that prions from short and long scrapie incubation time mice contain distinct PrPSc allotypes, PrPSc-A and PrPSc-B, respectively. Although many studies, including subtractive hybridization (9-11), have failed to identify a scrapie-specific nucleic acid, these negative results do not exclude the possibility that the scrapie agent has a host-independent genome. Based on a variety of criteria, including incubation period, lesion profile, and stability, Dickinson and coworkers (12-14) have reported that more than 20 distinct "strains" of scrapie agent exist. These investigators contend that their results provide evidence for a functional nucleic acid genome within the scrapie agent. Some of these scrapie isolates were derived in either C57BL or VM/Dk mice and were distinguished by differences in their incubation times. These two mouse strains are now known to have the Prn-pa and Prn-pb alleles, respectively (7, 15). Therefore, it is possible that the distinguishable behaviors of some prion isolates may have been due to an epigenetic effect of PrPSc allotype (16). We prepared inocula by passage of scrapie prion infectivity through mice producing different PrPSc allotypes. These scrapie isolates exhibited distinct incubation time behavior. Although our results can be explained by a hypothetical nucleic acid within the prion, they can equally as well be interpreted as a consequence of PrPSC allotype. We find that transplantation terminology is useful in describing the inoculum donor and recipient host combinations: (i) syngeneic, where the inoculum donor and recipient are genetically identical; (ii) allogeneic, where the donor and recipient are different inbred strains; (iii) semi-syngeneic, where the donor is an inbred parent of the F1 hybrid recipient; and (iv) semi-allogeneic, where the F1 hybrid donor inoculum is injected into an inbred parental strain.
ABSTRACT Scrapie is an infectious neurodegenerative disease caused by unusual pathogens called prions, in which no scrapie-specific nucleic acid has been detected to date. The only known component of the prion is the scrapie isoform of prion protein (PrP&), which is encoded by a host gene (Prn-p). Isolates of scrapie agent were prepared by passage of infectivity through inbred strains of mice that differ in scrapie incubation time and produce PrP& molecules differing by two amino acids. Both the length and variability of the incubation period were increased by inocula containing allogeneic PrPsc. For example, Prntpb I/Ln mice inoculated with scrapie isolate passaged through Prm-pa NZW mice had incubation times of283 ± 21 days compared with a 193 ± 6 day incubation time seen with isolate passaged once through isologous I/Ln mice. No further shortening of incubation time was observed following further isologous passage. NZW incubation times were prolonged by inoculation with prions from I/Ln mice. Results from (NZW x I/Ln)F2 mice and from using inocula from donors isologous for Prn-p but otherwise allogeneic with respect to the recipient suggest that the primary structure of PrPsc is responsible for these incubation time results. Incubation times in (NZW X I/Ln)Fl mice were constant regardless of the passage histories of the scrapie isolates and were equivalent to those of I/Ln mice inoculated with I/Ln prions, contending that prolongation of scrapie incubation time by the prion incubation time gene Prn-i is fully dominant. I/Ln incubation times longer than those in F1 hybrids may reflect a reduced efficiency of allogeneic PrPs' in initiating disease. Although some investigators propose that differences in behavior among scrapie isolates reflect host selection and argue for a nucleic acid genome, we suggest that the variation observed among our scrapie isolates is epigenetic, reflecting host-directed differences in the amino acid sequence of PrPs.
Scrapie is a transmissible neurodegenerative disease that is caused by infectious pathogens called prions (1). The only macromolecule of the scrapie prion that has been identified to date is an abnormal (scrapie) isoform of the prion protein (PrPSc), which is encoded by a host gene (2). Although it has been suggested that PrPSc is a pathologic by-product of infection with the scrapie "virus" (3, 4), several experimental studies (1), including immunoaffinity purification and neutralization (5), argue that PrPSc is a necessary and major component of the prion. That an essential component of the infectious prion particle is host-encoded is a significant feature distinguishing prions from viruses. Prions produced in a given host have PrPSC molecules specified by the host PrP gene and not by the inoculum. The discovery of genetic linkage in mice between the PrP gene (Prn-p) and the single gene (Prn-i) that predominates in controlling the interval between inoculation
MATERIALS AND METHODS Mice. I/LnJ (I/Ln) and NZW/LacJ (NZW) strains and F1 and F2 crosses between these two strains were produced in
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Abbreviations: PrP, prion protein; PrPSC, scrapie isoform of PrP. 7475
7476
Genetics: Carlson et al.
Proc. Natl. Acad. Sci. USA 86 (1989)
our research colony or obtained from animal resources at The Jackson Laboratory. Swiss mice were obtained from Charles River Breeding Laboratories. Preparation of Scrapie Isolates. All isolates in these experiments were derived from the Chandler murine scrapie isolate (17) kindly provided by William Hadlow (Rocky Mountain Laboratory, National Institute of Allergy and Infectious Diseases, Hamilton, MT) in its fourth passage in Swiss mice. Isolates were prepared by homogenizing brains of clinically ill mice in 0.8 M sucrose to yield a 10% (wt/vol) homogenate; homogenates were frozen and diluted just prior to use. Scrapie Incubation Time. All mice were inoculated intracerebrally with 30 p.l of diluted isolate with a 27-gauge hypodermic needle, which was inserted into the right parietal lobe. Eighty days after inoculation, examination of the mice for clinical signs began. Mice were examined every 3 days and scored positive when signs of neurological dysfunction were observed as described below. Days to onset of illness indicate the first positive score, followed by consecutive positive diagnoses or death. Also presented are survival time data, which were collected on all inoculated mice and afford a comparison with onset of illness times; the latter depend on the diagnosis of scrapie, which cannot always be made prior to death. Therefore, there are sometimes fewer incubation times than the corresponding number of survival times. The classical signs for diagnosis of murine scrapie include generalized tremor, ataxia, difficulty righting from a supine position, and rigidity of the tail. I/Ln mice are particularly difficult to diagnose; clinical signs in these mice may be limited to a dull, glazed expression, lack of usual motor activity, and a "hunched-up" stilted gait. The interval between inoculation and death provides an objective assessment of scrapie incubation period with the caveat that
occasional mice dying of other causes might be included. Neuropathologic assessment was performed on selected animals that were killed after they developed signs of scrapie. All animals, regardless of strain and source of isolate showed significant spongiform degeneration of subcortical and cortical grey matter and reactive astrocytic gliosis. Genotyping F2 Mice. (NZW x I/Ln)F2 mice were typed for Prn-p by restriction fragment length polymorphism analysis as previously described (8). Statistical Analysis. Analysis of variance, using the main National Institutes of Health computer with the assistance of David Alling (National Institute of Allergy and Infectious Diseases), was conducted on the sets of survival data from I/Ln, NZW, and F1 mice; isolates were grouped according to the Prn-p genotype of the donor. Each P value was corrected for multiple comparisons (three in number) by Bonferroni's inequality, which increases the P value. RESULTS Scrapie Prion Isolates Modulate Incubation Times in Inbred Mice. To determine whether passage of scrapie infectivity through mice that differ in scrapie incubation times and produce PrPSc molecules differing by two amino acids might influence the subsequent behavior of prions, we prepared inocula by passage of Chandler murine scrapie isolate (17) in Swiss (Prn-pa), NZW (Prn-pa), I/Ln (Prn-pb), and (NZW I/Ln)Fl (Prn-pa/Prn-pb) mice. The lengths of the incubation times measured from inoculation to onset of illness and to death are presented in Table 1. A single passage of Swiss mouse isolate through I/Ln mice caused a -dramatic shortening in incubation and survival times in syngeneic I/Ln mice (Table 1, experiment 3); no further shortening of I/Ln X
Table 1. Scrapie prion isolates modulate incubation times in inbred mice Incubation times Inoculum
Illness Mouse 2
Pm-pse Prn-p
Exp.
Mouse 1 Swiss Swiss Swiss NZW
Mouse 2 Swiss NZW
allele a
Death
RecipientIlnsDet 1:10* 1:10*
Recipie-p Prn-p Mouse
allele b b b b b b b
n
Days
n
1:10,000* Days
n
Days
1:10,000* n
Days
ND 1 11 314 ± 13 ND ND 27 288 ± 12 ND I/Ln 8 283 ± 21 14 313 ± 27 4 2 a 364 ± 32 16 274 ± 17 I/Ln b 10 248 ± 21 29 210 ± 101 15 260 ± 183 16 193 ± 6 I/Ln I/Ln b ND 4 11 211 ± 12 ND ND 11 239 ± 10 ND I/Ln I/Ln ND b 4 183 ± 7 ND ND 5 190 ± 7 ND F1 5 I/Ln I/Ln b 9 203 ± 8 15 6 253 ± 3 18 203 ± 8J 15 265 ± 4J I/Ln I/Ln I/Ln 7 Swiss F1 a/b 13 254 ± 15 10 295 ± 31 20 278 ± 14 23 249 ± 20 I/Ln ND NZW a ND ND 8 Swiss Swiss a 20 113 ± 2 ND 20 120 ± 2 9 Swiss NZW a NZW a 21 114 ± 1 14 130 ± 2 21 120 ± 1 16 134 ± 2 NZW 11 112 ± 3 8 141 ± 4 10 NZW NZW a a 6 132 ± 6 11 118 ± 2 b 11 NZW a 16 129 ± 4 9 167 ± 6] Swiss 9 159 ± 5 22 134 ± 3] I/LN 14 15 141 ± 3 12 F1 a/b NZW a 21 113 ± 2 130 ± 3 22 122 ± 2 Swiss ND F1 9 232 ± 4 ND 13 Swiss Swiss a a/b 13 224 ± 4 ND ND ND a/b F1 a/b 16 216 ± 2 ND ND 14 Swiss F1 ND 16 233 ± 2 ND F1 7 211 ± 4 ND 8 217 ± 5 ND 15 Swiss NZW a a/b ND ND ND 16 NZW NZW a F1 a/b 27 218 ± 2 ND ND 37 216 ± 3 ND b ND 17 F1 a/b 7 206 ± 5 ND ND 7 229 ± 7 Swiss I/Ln ND b 18 F1 a/b 19 208 ± 3 ND ND 21 223 ± 4 ND I/Ln I/Ln Analysis of variance, using the main NIH computer, was conducted on the sets of survival data collected from the I/Ln (experiments 1-7), NZW (experiments 8-12), and F1 (experiments 13-18) recipients; within each set of recipients, isolates were grouped according to the Prn-p genotype of the donor. The number of replicates for each experiment varied from 0 to 4. Among the I/Ln recipients, the mean survival time of animals inoculated with I/Ln isolate (bracketed data) was significantly less (df = 1, F = 12.1, P = 0.03) than those of the remaining two groups of mice, and the same was true of the standard deviations (F = 11.9, P = 0.04). Among the NZW recipients, survival was significantly longer among those inoculated with I/Ln material (bracketed) than among the remaining mice (df = 1, F = 39.0, P < 0.01); the same finding held for the standard deviations (F = 15.6, P = 0.04). In F1 mice there were no significant differences in the three mean incubation times. Note that each P value has been corrected for multiple comparisons (three in number) by Bonferroni's inequality, which increases the P value. For the sake of comparison, by Student's t test, survival times of I/Ln mice receiving Swiss isolate (experiment 1) were less than those receiving I/Ln isolate (experiment 3) at a level of P < 0.0001. ND, not determined. *Dilution of isolate used for inoculation.
Genetics: Carlson et al.
Proc. Natl. Acad. Sci. USA 86 (1989)
incubation period was observed with a second syngeneic (Table 1, experiment 6). The shortening of I/Ln incubation period was observed not only with the I/Ln isolate prepared from mice inoculated with the Swiss mouse isolate (314 13 down to 193 6 days), but also with I/Ln isolate from mice that received isolate from NZW (283 21 down to 211 12 days) or F1 (254 7 down to 183 7 days) mice. Incubation times of I/Ln mice inoculated with isologous isolate were equal to or slightly less than those of F1 mice. Therefore, prolongation of scrapie incubation time by the I/Ln Prn-i allele behaves as a fully dominant trait for PrPSc-B prions in contrast to longer incubation times in I/Ln mice after inoculation with PrPSc-A prions or PrPSc-A/B prions from F1 mice. NZW incubation periods (Table 1, experiments 8-12) were not shortened by syngeneic passage; incubation times of NZW mice inoculated with allogeneic Swiss mouse isolate (113 2 days) were identical to those inoculated with the first passage (114 1 days) or second passage (112 3 days) NZW isolate. Both NZW and Swiss mice are Prn-pa strains. In contrast, inoculation with PrPSc-B prions from I/Ln mice significantly delayed the onset of clinical signs and death (129 4 days). Incubation periods in (NZW x I/Ln)F1 mice, which express both the a and b alleles of Prn-p, were nearly constant, regardless of the passage history of the inoculum (Table 1). Three-way analysis of variance for survival data in Table 1 grouped according to the three Prn-p genotypes of the inocula donors and recipients indicated that the predominant effect (NZW vs. I/Ln and F1 mice, df = 1, F = 123.4, P < 0.001) was due to the prion incubation time gene Prn-i, which is linked to Prn-p (6-8). Overall, there were no significant differences in the survival times associated with the three types of inocula when averaged over the three recipient types. This suggests that the three types of inocula do not differ in infectious titer. Although end-point titration was not performed due to logistical considerations, groups of NZW passage
±
±
±
±
±
±
±
±
7477
and I/Ln mice were inoculated with a 1:10,000 dilution of some isolates (Table 1). The incubation period is proportional to prion titer (18), and it is unlikely that the shortening of I/Ln incubation times seen with isolate from syngeneic I/Ln mice was due to an increase in prion titer, because there was a prolongation of incubation period in NZW mice at both 1:10 and 1:10,000 dilutions following inoculation with I/Ln prions. More importantly, comparison of survival times by analysis of variance in each host separately indicated that the homologous isolate significantly shortened scrapie incubation times in NZW and I/Ln mice but that the three types of inocula did not differ in F1 recipients (see the legend to Table 1 for analysis). Presence of Allogeneic PrPsc in the Inoculum Prolongs Scrapie Incubation Time and Increases Variability among Recipients. The shortening of incubation periods in I/Ln and NZW mice inoculated with isolates from Prn-p identical mice was accompanied by a dramatic reduction in variation among individuals. Table 2 is a presentation of incubation time data showing, in addition to the mean and the standard error of the mean (from Table 1), the shortest and longest incubation time in each experiment, the standard deviation from the mean, and the variance. It is important to emphasize that the mice used in these studies differ at numerous loci in addition to Prn-p. The prolongation of incubation time associated with increased variance following inoculation with allogeneic PrPSc is illustrated in Fig. 1, which shows distributions of onset of illness and survival times for I/Ln and NZW mice given isolates from I/Ln, NZW, or F1 mice. For example, incubation times of I/Ln mice ranged from 182 to 344 days (variance = 3641) following inoculation with prions from allogeneic NZW mice and from 136 to 363 days (variance = 2973) after injection with isolate from semi-allogeneic F1 mice; in contrast, incubation times in I/Ln mice given isolate from syngeneic mice ranged from 149 to 224 days (variance = 610). There also was a significant increase in variation between individual NZW mice inoculated with isolate from
Table 2. Presence of foreign PrP in inoculum prolongs scrapie incubation time and increases variability among recipients Onset of illness, Donor/recipient Foreign PrP combination (group)* Donor -* Recipient n Variance Max. SD Min. days Group 1 610 193 ± 6 224 24.7 16 149 Syngeneic I/Ln - I/Ln 644 203 ± 8 9 143 233 25.4 1488 211 ± 12 11 180 285 38.6 204 183 ± 7 4 174 14.2 204 + NZW I/Ln 283 ± 21 8 182 344 3641 60.3 Allogeneic + Swiss - I/Ln 314 ± 13 11 1929 204 351 43.9 + F1 254 ± 15 13 136 363 54.5 2973 I/Ln Semi-allogeneic Group 2 47 NZW - NZW 114 ± 1.5 21 104 125 6.9 Syngeneic 79 112 ± 3 11 97 125 8.9 82 Swiss - NZW 113 ± 2 20 9.0 96 137 Allogeneic + 193 17 13.9 129 ± 3 115 174 I/Ln - NZW + F1 NZW 113 ± 2.5 21 142 11.5 132 104 Semi-allogeneic Group 3 F1 81 217 ± 4 16 9.0 203 231 F1 Syngeneic 206 ± 5 7 192 12.7 163 224 I/Ln F1 Semi-syngeneic NZW F1 211 ± 4 7 11.5 131 202 230 242 224 ± 4 15.6 Swiss F1 13 182 235 Allogeneic Group 4 155 15 12.4 Swiss - F2 Prn-pa 127 ± 3 107 149 Allogeneic 392 32 Swiss - F2 Prn-pa/b 213 ± 3.5 181 272 19.8 + 2389 Swiss F2 Prn-pb 274 ± 14 12 48.9 193 359 ± The onset of illness data from Table 1 are presented in more detail, showing, in addition to onset of illness (mean SEM) and number of mice (n), the shortest (Min.) and longest (Max.) individual incubation times, the standard deviation (SD), and the variance. Only the proximal donor mouse strain is listed. In addition, data from a population of (NZW x I/Ln)F2 mice, excluding three potential Prn-i-Prn-p recombinants (8), are included. *The data are grouped according to the recipient mouse strain: group 1, I/Ln; group 2, NZW; group 3, F1; group 4, F2.
Genetics: Carlson et al.
7478 N
4
Proc. Natl. Acad. Sci. USA 86 (1989) I /LnJ isolate
4
3
3
2 6)
ID
0
4
4
3
3
NZW/LacJ isolate
-
2
L.
E z
Fl isolate
4
4
3 2
180
260
340
420
.13
180
260
340
420
I/LnJ
4
isolate
4
8
3
24
I)
111_===Lw
~~~NZW/LacJ isolate
l s
a)
10
_
E Z
8 )
10
2 6
Fl ~~~~~isolate
8 6
24 120
140
160
Onset of Illn less
180
(days)
120
140
160
Death
(days)
180
Effects of scrapie isolates from I/Ln, NZW, and (NZW I/Ln)F1 mice on distribution of scrapie incubation times in I/Ln (A) and NZW (B) mice. Times in days from inoculation to onset of illness and to death are shown. Note that the scale in A (100-500 days) is larger than that in B (100-200 days). The distributions illustrated are the data summarized in Table 1, experiments 2, 3, and 7 for A, and experiments 9, 11, and 12 for B. FIG. 1.
x
allogeneic I/Ln mice, compared with isolate from Prn-p identical animals. Isolate from F1 mice did not produce a statistically significant increase in variance among NZW recipients, but some broadening of the incubation time distribution is evident in Fig. 1. Recent analysis of amplified cDNA prepared from F1 mouse brain mRNA indicates that both Prn-p alleles are expressed at comparable levels in heterozygous mice (C. Mirenda and D.W., unpublished results). The obvious candidate to account for the differences between our scrapie isolates is PrPSc itself. Even though the scrapie isolates were brain homogenates, rather than purified prions, and contained numerous molecules that differed between the inbred mouse strains, two lines of evidence suggest that allogeneic PrPSc in the inoculum causes prolongation of the incubation time and increased variation. First, as indicated in Table 2, there was no difference in scrapie incubation period between NZW mice inoculated with isolate from syngeneic NZW mice or with isolate from allogeneic Swiss mice. Swiss mice are identical with NZW at Prn-p but differ at many other loci. Prolongation of the NZW incubation period was seen with isolate from allogeneic I/Ln mice that contained allogeneic PrPSc-B molecules. Second, results from a population of (NZW x I/Ln)F2 mice (8) (summarized in Table 2) that were inoculated with the PrPSc-A isolate from Swiss mice suggest that the prolongation of incubation times
and increased variability seen in I/Ln mice given isolate from allogeneic mice was due to the presence of foreign PrPSc in the inoculum. The F2 cross was segregating for numerous genes in addition to Prn-p, but high variability was only seen among Prn-pb homozygous mice (b/b, range 193-359, variance = 2389; a/b, range = 181-272, variance = 392; a/a, range = 107-149, variance = 155). If the large variation between individual I/Ln mice inoculated with isolate from allogeneic animals was due to products of genes unlinked to Prn-p, high variance would have been seen in all three Prn-p genotypes among the F2 mice. The incubation period was also significantly longer in Prn-pb homozygous F2 mice than in Prn-p heterozygous F2 animals (8); results in Table 1 indicated that the long-incubation time Prn-i allele of I/Ln mice appeared to act codominantly following inoculation with PrPSc-A prions but as a fully dominant trait for PrPSc-B prions. We predict that if F2 mice were inoculated with PrPSc-B prions, variance would increase in Prn-pa homozygotes and decrease in Prn-pb homozygotes; incubation times of Prn-pb homozygotes also would be expected to decrease to those of Prn-pa/Prn-pb heterozygous mice. To date, the only molecule in scrapie brain homogenates that correlates with changes in scrapie infectivity is PrPSc (1, 5).
DISCUSSION Although many studies suggest that scrapie prions may be devoid of a scrapie-specific nucleic acid (19-25), the existence of such a molecule remains unresolved (26). Though chemical evidence for a scrapie-specific nucleic acid is still lacking, the results of studies by Dickinson, Kimberlin, and coworkers on strains of scrapie agent (12-14) are used as a biological argument for the existence of this hypothetical polynucleotide. Conversely, there are insufficient biochemical and fluctuation analysis (27) data to equate the genesis of the numerous strains of scrapie agent with nucleic acid mutation. We emphasize that our observations do not address the basis for distinguishable behavior of some scrapie isolates upon repeated passage in the same host. These "true breeding" prion isolates, which produce divergent incubation times in the same inbred mouse strain, provide compelling evidence for a host-independent informational component in the prion. However, caution should be exercised in concluding that nucleic acid is the only conceivable vehicle for prion-specified information (28). Clearly, the biochemical basis for differential behavior of scrapie prion isolates is an intriguing biological problem. It can be argued that our results reflect selection by NZW and I/Ln mice (but not by long-incubation time F1 mice) for the more rapidly replicating variants present as a mixture in the original Chandler scrapie isolate. In our opinion, a more concise interpretation is that the variation between our scrapie isolates is epigenetic and reflects host-directed differences in the amino acid sequence of PrPSc. Our results suggest the existence of a mouse strain barrier that is Prn-p dependent and that operates, in addition to the effects of Prn-i, in modulating scrapie incubation period. This PrP allotype barrier may be analogous to the species barrier for transmission of scrapie (29, 30); for example, mice inoculated with sheep scrapie isolate have an exceptionally long incubation period, but in subsequent passage the incubation period rapidly shortens and becomes fixed. Bruce and Dickinson (13) have reported a mouse strain barrier between VM and C57BL mice to passage of some isolates of scrapie; VM mice have a prolonged scrapie incubation period following inoculation with common scrapie isolates and are Prn-pb (7, 15). Although Dickinson and coworkers (13, 14) state that there was a gradual shortening after crossing a mouse strain or species barrier, their published data suggest that almost all of the change occurred in a single passage, as shown here. It is clear that isolates from Prnp-disparate hosts will contain PrPsc that differs in primary
Genetics: Carlson et al. structure from the endogenous PrP of recipient mice. We are not aware of any data that mitigate against the hypothesis that differences in PrP amino acid sequence can produce scrapie prion isolates with different properties. Our results argue that the prion gene complex (Prn) can exert two distinct influences on the scrapie incubation period. First, the Prn-i gene (which may be Prn-p itself) has a profound effect on the intrinsic length of the incubation period. Regardless of the source of the inoculum in our series of experiments, Prnb acted to cause a dramatic prolongation of the interval between inoculation and illness, perhaps by controlling the rate of prion accumulation and consequent formation of pathological lesions. Second, the PrPSc allotype in the inoculum may influence the initiation ofthe "infectious process" through binding, entry, or some as yet undefined mechanism. In Prna and Prnb mice, isologous prions were the most efficient in initiating the disease and yielding the shortest incubation time. This should force a careful reevaluation of the criteria used in defining strains of scrapie agent. Many strains of scrapie agent were distinguished from one another by whether incubation times in VM/Dk, C57BL, and their F1 hybrid mice revealed dominant, codominant, or "overdominant" genetic control of the incubation period (12-14). In some cases, codominance might reflect an epigenetic effect of foreign PrPSc allotype in the inoculum, as suggested by the fact that the incubation times of I/Ln mice inoculated with PrPSc-A prions were longer and more variable than those of F1 mice. After inoculation with I/Ln PrPSc-B prions, equivalent incubation times were observed in I/Ln and F1 mice. One possible explanation for the influence of PrPSC allotype in the inoculum is an immune response. The behavior of our prion isolates follows the classical laws of transplantation; the prolonged incubation time and increased variance seen among I/Ln mice inoculated with prion isolate from allogeneic or semi-allogeneic mice could be due to transient immunemediated reduction in the number of infectious units able to initiate disease. Recent data (5) showing neutralization of infectivity in hamsters with anti-PrP antibodies indicate that immune effectors can delay onset of disease. Although absence ofdetectable immunity to the scrapie agent is a hallmark of the disease (31), to our knowledge, no studies of protective immunity against isolates containing foreign PrP allotype have been attempted. After disease is initiated, the PrPSc that accumulates is host-encoded and immunologically inert (2). Alternatively, nonimmunologic mechanisms may explain the influence of PrPSc allotype in the inoculum. For example, the normal isoform of PrP (PrPC), which is found on the external surface of cells anchored by a glycosyl phosphatidylinositol moiety (32) might be involved in the initital recognition of inoculated PrPSC molecules. Whether or not PrPC is complexed with other cellular factors that together act as a receptor for the binding and entry of prions remains to be determined. Prion protein, discovered by enriching for scrapie infectivity (33, 34), has been implicated in numerous aspects of scrapie. Extracellular accumulation of PrPSC is the diagnostic hallmark for prion diseases in all species that have been examined (1), and allelic forms of prion protein gene encoding variant proteins have been genetically linked with susceptibility to prion disease in both mice (6, 7) and humans (35). The results presented here indicate that PrP may have another important role in scrapie-determination of at least some biological properties exhibited by distinct scrapie isolates. We thank Dr. David Alling for assistance with statistical analysis and for his helpful comments and suggestions. Some ofthis work was performed while G.A.C. was on the staff of The Jackson Laboratory. This work was supported by research grants (AG02132 and NS14609) and a Senator Jacob Javits Center of Excellence in Neuroscience award (NS22786) from the National Institutes of Health as well as by gifts from the Sherman Fairchild Foundation and RJR/Nabisco.
Proc. Natl. Acad. Sci. USA 86 (1989)
7479
1. Prusiner, S. B. (1987) N. Engl. J. Med. 317, 1571-1581. 2. Basler, K., Oesch, B., Scott, M., Westaway, D., Wachli, M., Groth, D. F., McKinley, M. P., Prusiner, S. B. & Weissmann, C. (1986) Cell 46, 417-428. 3. Braig, H. R. & Diringer, H. (1985) EMBO J. 4, 2309-2312. 4. Chesebro, B., Race, R., Wehrly, K., Nishio, J., Bloom, M., Lechner, D., Bergstrom, S., Robbins, K., Mayer, L., Keith, J. M., Garon, C. & Haase, A. (1985) Nature (London) 315, 331-333. 5. Gabizon, R., McKinley, M. P., Groth, D. S. & Prusiner, S. B. (1988) Proc. Natl. Acad. Sci. USA 85, 6617-6621. 6. Carlson, G. A., Kingsbury, D. T., Goodman, P. A., Coleman, S., Marshall, S. T., DeArmond, S. J., Westaway, D. & Prusiner, S. B. (1986) Cell 46, 503-511. 7. Westaway, D., Goodman, P. A., Mirenda, C. A., McKinley, M. P., Carlson, G. A. & Prusiner, S. B. (1987) Cell 51, 651-662. 8. Carlson, G. A., Goodman, P. A., Lovett, M., Taylor, B. A., Marshall, S. T., Peterson-Torchia, M., Westaway, D. & Prusiner, S. B. (1988) Mol. Cell. Biol. 8, 5528-5540. 9. Wietgrefe, S., Zupancic, M., Haase, A., Chesebro, B., Race, R., Frey, W., Rustan, T. & Friedman, R. L. (1985) Science 230, 1177-1179. 10. Duguid, J. R., Rohwer, R. G. & Seed, B. (1988) Proc. NatI. Acad. Sci. USA 85, 5738-5742. 11. Diedrich, J., Wietgrefe, S., Zupancic, M., Staskus, K., Rietzel, E., Haase, A. T. & Race, R. (1987) Microb. Pathogen. 2, 435-442. 12. Bruce, M. E. & Dickinson, A. G. (1987) J. Gen. Virol. 68, 79-89. 13. Bruce, M. E. & Dickinson, A. G. (1979) in Slow Transmissible Diseases of the Nervous System, eds. Prusiner, S. B. & Hadlow, W. J. (Academic, New York), pp. 71-86. 14. Dickinson, A. G., Bruce, M. E., Outram, G. W. & Kimberlin, R. H. (1985) in Proceedings of Workshop on Slow Transmissible Diseases, ed. Tateishi, J. (Japan Min. Health Welfare, Tokyo), pp. 105-118. 15. Hunter, N., Hope, J., McConnell, I. & Dickinson, A. G. (1987) J. Gen. Virol. 68, 2711-2716. 16. Carlson, G. A., Westaway, D., Goodman, P. A., Peterson, M., Marshall, S. T. & Prusiner, S. B. (1988) in Novel Infectious Agents and the CNS, Ciba Foundation Symposium (Wiley, Chichester, U.K.), No. 135, pp. 84-99. 17. Chandler, R. L. (1961) Lancet i, 1378-1379. 18. Prusiner, S. B., Groth, D. F., Cochran, S. P., Masiarz, F. R., McKinley, M. P. & Martinez, H. M. (1980) Biochemistry 19,48834891. 19. Alper, T., Haig, D. A. & Clarke, M. C. (1966) Biochem. Biophys. Res. Commun. 22, 278-284. 20. Alper, T., Cramp, W. A., Haig, D. A. & Clarke, M. C. (1967) Nature (London) 214, 764-766. 21. Bellinger-Kawahara, C., Cleaver, J. E., Diener, T. 0. & Prusiner, S. B. (1987) J. Virol. 61, 159-166. 22. Bellinger-Kawahara, C., Diener, T. O., McKinley, M. P., Groth, D., Smith, D. R. & Prusiner, S. B. (1987) Virology 160, 271-274. 23. Beilinger-Kawahara, C., Kempner, E., Groth, D., Gabizon, R. & Prusiner, S. B. (1988) Virology 164, 537-541. 24. Gabizon, R., McKinley, M. P. & Prusiner, S. B. (1987) Proc. Natl. Acad. Sci. USA 84,4017-4021. 25. Gabizon, R., McKinley, M. P., Groth, D. F., Kenaga, L. & Prusiner, S. B. (1988) J. Biol. Chem. 263, 4950-4955. 26. Oesch, B., Groth, D. F., Prusiner, S. B. & Weissman, C. (1988) in Novel InfectiousAgents and the CNS, Ciba Foundation Symposium (Wiley, Chichester, U.K.), No. 135, pp. 209-223. 27. Luria, S. E. & Delbruck, M. (1943) Genetics 28, 491-511. 28. Diener, T. 0. (1987) Cell 49, 719-721. 29. Pattison, I. H. (1966) in Slow, Latent and Temperate Virus Infections, eds. Gajdusek, D. C., Gibbs, C. J. & Alpers, M. (Natl. Inst. Neurol. Disorders Stroke, Washington, DC), Monogr. 2, pp. 249257. 30. Kimberlin, R. H., Cole, S. & Walker, C. A. (1987) J. Gen. Virol. 68, 1875-1881. 31. Kasper, K. C., Bowman, K., Stites, D. P. & Prusiner, S. B. (1981) in Hamster Immune Responses in Infectious and Oncologic Diseases, eds. Streilein, J. W., Hart, D. A., Stein-Streilein, J., Duncan, W. R. & Billingham, R. E. (Plenum, New York), pp. 401-413. 32. Stahl, N., Borchelt, D. R., Hsiao, K. & Prusiner, S. B. (1987) Cell 51, 229-240. 33. Bolton, D. C., McKinley, M. P. & Prusiner, S. B. (1982) Science 218, 1309-1311. 34. Prusiner, S. B., Bolton, D. C., Groth, D. C., Bowman, K.A., Cochran, S.P. & McKinley, M.P. (1982) Biochemistry 21, 69426950. 35. Hsiao, K., Baker, H. F., Crow, T. J., Poulter, M., Owen, F., Terwilliger, J. D., Westaway, D., Ott, J. & Prusiner, S. B. (1989) Nature (London) 338, 342-345.
