In Vitro Generation of High-Titer Prions

Abstract

Prions are composed mainly, if not entirely, of PrP^Sc, an infectious misfolded isoform of PrP^C, the normal isoform of the prion protein. Here we show that protein misfolding cyclic amplification (PMCA)-generated hypertransmissible mink encephalopathy (HY TME) PrP^Sc is highly infectious and has a titer that is similar, if not identical, to that associated with brain tissue from animals infected with the HY TME agent that are in the terminal stage of disease. These data demonstrate that PMCA efficiently replicates the prion agent and provide further support for the hypothesis that in vitro-generated prions are bona fide and are not due to contamination.

TEXT

Prion diseases are fatal neurodegenerative disorders that affect mammals, including humans. One of these, chronic wasting disease (CWD), is an emerging prion disease of cervids in North America (34). CWD has a poorly defined host range and poses an unknown risk of zoonotic transmission to humans (3, 19, 24, 29, 31, 38). During the disease process, PrP^C, the normal isoform of the prion protein, is posttranslationally converted to PrP^Sc (8, 11). Several lines of evidence now indicate that PrP^Sc is the etiological agent of prion diseases (10, 12, 22, 27, 35). The neuropathology of the prion diseases is characterized by spongiform degeneration, reactive gliosis, and the deposition of PrP^Sc (17).

Animals inoculated with in vitro-generated PrP^Sc develop prion disease. The ability to produce prions in vitro is compelling evidence that prions are not caused by a virus and are indeed caused by PrP^Sc (13). The incubation periods of disease differ in animals inoculated with in vitro-generated PrP^Sc (9, 10, 12, 21, 22, 33). While the reason for this variation is not known, one possibility is that differences in the titers produced by the various methodologies used to generate PrP^Sc account for the observed differences in incubation periods.

There is an inverse relationship between the incubation period of disease and the titer of the prion inoculum. Inoculation of animals with a decreasing titer of the agent results in a corresponding increase in the incubation period (23). This robust and highly reproducible relationship is the basis of the incubation interval assay (28). However, this relationship is not an absolutely foolproof means to estimate titer, as attempts to inactivate the agent with heat or chemical treatment have resulted in a longer incubation period of disease than would have been predicted from the titer (14, 39). The mechanism responsible for alterations in the relationship between incubation period and titer is not known.

We have previously demonstrated that protein misfolding cyclic amplification (PMCA)-generated hypertransmissible (HY) PrP^Sc caused disease at 83 ± 3 days postinfection compared to 61 ± 3 days for the brain-derived HY transmissible mink encephalopathy (HY TME) agent, suggesting that the PMCA samples contained a high titer of HY TME agent (33). To investigate whether the short incubation period seen with animals inoculated with PMCA-generated HY TME agent corresponded to a high titer, we performed an endpoint dilution analysis of both brain-derived and PMCA-generated HY TME agents. The serial PMCA reaction mixtures initially seeded with HY TME agent or mock brain homogenate were produced as described previously (33). The 10th serial round of HY TME agent- or mock-seeded PMCA reactions was used for bioassay analysis, since the dilution of the initial HY TME agent seed was beyond the limit of detection by bioassay (5, 6). Therefore, the prion titer of these samples should have been solely due to in vitro generation of the HY TME agent. Ten-fold serial dilutions of either brain-derived or PMCA-generated HY TME agent were prepared in Dulbecco's phosphate-buffered saline (DPBS) as described previously (18). Each inoculum was intracerebrally (i.c.) inoculated into groups of Syrian hamsters (n = 5) that were monitored for clinical signs of prion disease 3 days per week until day 400 postinfection, when the experiment was terminated. In animals that developed clinical signs of prion disease, the clinical diagnosis was confirmed by the presence of proteinase K (PK)-resistant PrP (PrP^Sc) in the central nervous system (CNS) as detected by Western blot analysis (Fig. 1). Specifically, brain homogenate was digested with PK at 100 μg/ml for 60 min at 37°C, 250-μg equivalents were size fractionated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and PrP^Sc was detected using 3F4 antibody as previously described (1). We found that the animals that developed clinical signs of prion disease harbored PrP^Sc in the CNS (Table 1; Fig. 1, lanes 3 to 7). Conversely, the animals that did not develop clinical disease did not show detectable levels of PrP^Sc in the CNS (Table 1; Fig. 1, lanes 8 to 10). The neuropathological features of the animals that developed clinical disease were consistent with prion infection and included spongiform degeneration, deposition of PrP^Sc, and reactive astrocytosis (Fig. 2, panels A, B, and C). Animals inoculated with PMCA reaction mixtures from the 10th serial round that were seeded with an uninfected brain homogenate (mock seeded) failed to develop disease by day 400 postinfection, and we did not detect PrP^Sc in the CNS or prion-associated neuropathology (Table 1; Fig. 1, lane 11; Fig. 2, panels D, E, and F). Using the method of Reed and Muench (30), the HY TME agent-infected brain homogenate contained a titer of 10^9.3 i.c. LD₅₀/g, while the PMCA reaction mixtures that were initially seeded with the same HY TME agent-infected brain homogenate contained a titer of 10^8.5 i.c. 50% lethal dose (LD₅₀)/g (Table 1). These data indicate that, per gram, the PMCA-generated HY TME agent is 0.8 log units less infectious than the HY TME agent-infected brain homogenate used to initially seed the PMCA reaction mixtures.

Fig. 1.

Western blot detection of PrP^Sc in brain tissue from hamsters inoculated with serial dilutions of PMCA-generated HY TME agent corresponds to clinical diagnosis of disease. Hamsters were inoculated with the Drowsy (DY) TME agent, the HY TME agent, or serial 10-fold dilutions of PMCA reaction mixtures seeded with HY TME agent- or mock (UN)-infected brain homogenate. Western blot analysis of proteinase K-digested brain homogenate indicated that animals that developed clinical signs of prion infection contained PrP^Sc. The migration of the 19- and 21-kDa unglycosylated polypeptides is indicated on the left of the panel.

Table 1.

Response of hamsters to brain-derived or PMCA-generated HY TME agent

Dilution	Result for indicated inoculuma
	PMCA generated		Brain derived
	Incubation period (days ± SEM)	No. of hamsters affected/total no. inoculated	Incubation period	No. of hamsters affected/total no. inoculated
10−2	83 ± 3b	5/5	61 ± 3	5/5
10−3	93 ± 3	5/5	71 ± 3	5/5
10−4	99 ± 4	5/5	79 ± 9	5/5
10−5	164 ± 111	5/5	89 ± 6	5/5
10−6	186 ± 24	5/5	98 ± 2b	5/5
10−7	214 ± 11	3/5	134 ± 9b	4/5
10−8	>400	0/5	192 ± 54b	3/5
10−9	>400	0/5	>400b	0/5
None (mock inoculation)	>400	0/5	>400	0/5

^aTiter for PMCA-generated HY TME agent, 10^8.5 i.c. LD₅₀/g; titer for brain-derived agent, 10^9.3 i.c. LD₅₀/g.

^bAdapted from references 18 and 33.

Fig. 2.

The neuropathological features of animals inoculated with PMCA-generated HY TME agent are consistent with prion infection. Hamsters were inoculated with PMCA reaction mixtures seeded with brain homogenate from either HY TME agent (A to C)- or mock (D to F)-infected hamsters. Brain sections from the superior colliculus were either stained with hematoxylin and eosin (H&E) (A and D) or subjected to immunoreactions performed with antibodies directed to the prion protein (PrP) (B and E) or to the glial fibrillary acidic protein (GFAP) (C and F). Scale bar, 100 μm.

The PMCA process is highly efficient at generation of HY PrP^Sc and HY TME agents. While animal bioassays are highly sensitive, they are not very accurate, and it has been established that differences in titers between two samples should be greater than 0.8 log units to be considered statistically significant (26). Therefore, it is possible that the titer of the PMCA-generated HY TME agent does not differ from that of the brain-derived HY TME agent. Based on the high titer of the PMCA-generated HY TME agent and the lack of detectable PrP^Sc and infectivity in the mock-seeded negative controls, we conclude that the HY TME agent contained in PMCA reaction mixtures was indeed generated in vitro. These observations suggest that the presence of the HY TME agent in the PMCA reaction mixtures was not due to contamination, since the majority, or entirety, of the PMCA reaction would have to be a contaminant. Additionally, it is unlikely that the presence of the PMCA-generated HY TME agent was due to spontaneous generation of PrP^Sc, since we failed to detect PrP^Sc in the mock-seeded PMCA reaction mixtures and because the reaction mixtures did not result in detectable infectivity. It remains to be determined whether PMCA can generate high titers of agent with other prion strains.

The incubation period of disease does not correspond to the titer of the PMCA-generated HY TME agent. The relationship between the titer of the inoculum and the incubation period is highly reproducible (5, 7, 18, 32, 40). By calculation of the titers of the brain-derived and PMCA-generated HY TME agents (Table 1), the relationship between inoculum titer and incubation period was established (Table 2). Compared to that seen with the brain-derived HY TME agent, the incubation period of the PMCA-generated HY TME agent was, in general, significantly longer per titer unit (Table 2). At low titers, the higher variance of the incubation period may obscure actual differences in incubation periods. The mechanism responsible for this observation is not known; however, several possibilities exist. Exposure of the prion agent to high temperatures can alter the relationship between incubation period and titer (14, 25, 39). Sonication can result in a localized increase in temperature; therefore, it is possible that sonication-generated heat is responsible for the effect (16). Treatment of the prion agent with detergents (e.g., N-lauroyl sarcosinate) can extend the incubation period of disease without reducing titers (20, 36). The PMCA conversion buffer contains Triton X-100, a detergent that can alter the relationship between titer and incubation period (37). The PMCA process can generate new strains; therefore, the observed differences between titers and incubation periods may represent strain interference between HY TME and another strain(s) resulting from the PMCA process (2, 4, 15). The mechanism responsible does not result in a permanent modification of the PMCA-generated HY TME agent, since a second i.c. hamster passage performed using a 1% (wt/vol) brain homogenate from hamsters inoculated with a 10⁻² dilution of the PMCA-generated HY TME agent resulted in all (n = 5) of the animals developing clinical disease by 60 ± 3 days postinfection, which is similar to the result seen with the brain-derived HY TME agent (Table 1). Finally, the results presented here suggest that the titer of PMCA-generated prion agent derived from previous studies may be higher than the incubation period would indicate.

Table 2.

Relationship between titer and incubation period of brain-derived and PMCA-generated HY TME agent

PMCA-generated HY TME inoculum titer (i.c. LD50/g)	Brain-derived HY TME inoculum titer	Increase in no. of days (%) of incubation with brain-derived vs PMCA-generated HY TME agent	P value
106.5	106.3	12 (17)	<0.05
105.5	105.3	14 (15)	<0.05
104.5	104.3	10 (10)	<0.05
103.5	103.3	75 (40)	>0.05
102.5	102.3	52 (28)	<0.05
101.5	101.3	22 (10)	>0.05