Nathan R. Deleault, Ralf W. Lucassen, and  Surachai Supattapone

Department of Biochemistry, 7200 Vail Building, Dartmouth Medical School, Hanover, New Hampshire 03755, USA

Correspondence and requests for materials should be addressed to S.S. (supattapone@dartmouth.edu).

Much evidence supports the hypothesis that the infectious agents of prion diseases are devoid of nucleic acid, and instead are composed of a specific infectious protein [1]. This protein, PrP^Sc, seems to be generated by template-induced conformational change of a normally expressed glycoprotein, PrP^C (ref. 2). Although numerous studies have established the conversion of PrP^C to PrP^Sc as the central pathogenic event of prion disease, it is unknown whether cellular factors other than PrP^C might be required to stimulate efficient PrP^Sc production. We investigated the biochemical amplification of protease-resistant PrP^Sc-like protein (PrPres) using a modified version [3] of the protein-misfolding cyclic amplification method [4]. Here we report that stoichiometric transformation of PrP^C to PrPres in vitro requires specific RNA molecules. Notably, whereas mammalian RNA preparations stimulate in vitro amplification of PrPres, RNA preparations from invertebrate species do not. Our findings suggest that host-encoded stimulatory RNA molecules may have a role in the pathogenesis of prion disease. They also provide a practical approach to improve the sensitivity of diagnostic techniques based on PrPres amplification.

We previously showed that PrPres amplification in vitro shares many specific features with the
pathogenic process of prion propagation in vivo, including strain and species specificity [3]. In a
typical amplification reaction, diluted prion-infected brain homogenate (0.1% w/v) is mixed with
either 5% (w/v) normal brain homogenate or buffer control and incubated overnight at 37 °C.
Hamster Sc237 PrPres is amplified about sixfold under these conditions (Fig. 1a, compare M with
Sc). While characterizing the biochemical requirements of PrPres amplification reactions, we were
surprised to discover that treatment of such reactions with DNase-free, heterogeneous pancreatic
RNase abolished PrPres amplification in a dose-dependent manner (Fig. 1a, first panel). In vitro
PrPres amplification was also abolished by treatment with purified pancreatic RNase A, RNase T1,
micrococcal nuclease, or benzonase (Fig. 1a). In a control experiment, we found that addition of
RNase A for 1 h after overnight incubation did not reduce the recovery of PrPres already amplified
(Supplementary Fig. S1).

Figure 1 Effect of various enzymes on PrPres amplification. Immunoblots of PrPres amplification reactions. Samples include mixtures of normal and diluted scrapie brain homogenate (M), diluted scrapie brain homogenate control (Sc), and mixtures of normal and diluted scrapie brain homogenate incubated with various enzymes in a dilution series. The values of X (manufacturer's U µl^-1) for enzymes are: DNase-free RNase = 0.5; RNase A = 0.005; RNase T1 = 50; micrococcal nuclease = 0.005; benzonase = 2.5; RNase V1 = 0.01; RNase H = 1; RNase-free DNase = 1; EcoRI = 1; apyrase = 0.1; heparinase III = 0.025.

To ensure that the commercial nuclease preparations we used were not contaminated with
proteases, we measured the levels of PrP^C and PrPres after overnight incubation with various
nucleases. These measurements confirmed that levels of PrP^C (Fig. 2a) and input PrPres (Fig. 2b)
were both unperturbed by addition of enzymes that inhibited PrPres amplification.

Figure 2 Nucleases do not cause proteolytic, steric or end-product inhibition of PrPres amplification. a, Effect of DNase-free RNase on PrP^C levels. Standard amplification mixtures were incubated overnight with RNase. Serial dilutions of each sample are shown. b, Effect of nucleases on PrPres levels. Samples of diluted scrapie brain homogenate were treated with specified nucleases at concentrations designated 'X' in Fig. 1. c, Effect of inactivated nucleases on PrPres amplification. Samples include mixtures of normal and diluted scrapie brain homogenate (M) and diluted scrapie brain homogenate control (Sc). d, Effect of nucleotides on PrPres amplification.

To test whether inhibition of PrPres amplification might be mediated by end products of RNase
digestion, we measured directly the effect of cyclic 2',3'-guanidine monophosphate (GMP) and 3'
-cytidine monophosphate (CMP) on PrPres amplification. Neither of these nucleotides inhibited
PrPres amplification in vitro at concentrations up to 1 mM (Fig. 2d). Our control experiments rule
out the possibility that contaminating proteases, steric hindrance, or digestion end-products account
for the inhibition of PrPres amplification by specific nucleases. Taken together, these experiments
indicate that RNA is required for PrPres amplification in vitro.

We next sought to determine whether a preparation of isolated RNA molecules could reconstitute
the ability of nuclease-treated normal brain homogenate to amplify PrPres. Remarkably, total RNA
isolated from hamster brain successfully reconstituted the ability of benzonase-pre-treated brain
homogenate to amplify PrPres in a dose-dependent manner (Fig. 3a). By contrast, purified heparan
sulphate proteoglycan (HSPG) failed to reconstitute PrPres amplification (Fig. 3a). Other
polyanions, such as ssDNA (Fig. 3b), polyadenylic acid, heparan sulphate, pentosan sulphate and
polyglutamic acid (data not shown) also failed to stimulate PrPres amplification. In this and other
reconstitution experiments, benzonase-treated control lanes have a greater level of PrPres
amplification than the diluted scrapie brain homogenate control samples, indicating that the
benzonase pre-treatment reactions were incomplete. Empirically, we found that it was necessary to
perform benzonase pre-treatment reactions at 4 °C to avoid denaturing PrP^C before the addition of
polyanions.

Figure 3 Reconstitution of PrPres amplification with RNA. Immunoblots of PrPres amplification reactions. Samples include mixtures of normal and diluted scrapie brain homogenate (M) and diluted scrapie brain homogenate control (Sc). Indicated samples were pre-treated with benzonase before reconstitution assays as described in Methods. a, Reconstitution with total hamster brain RNA or HSPG. b, Reconstitution with 0.5 mg ml^-1 total hamster brain RNA or 0.5 mg ml^-1 random, synthetic 23-base DNA oligonucleotide. c, Reconstitution with 0.5 mg ml^-1 total (T), filtrate (F), retentate (R) and 50% formamide retentate (R*) samples from RNA fractionated by ultrafiltration.

To estimate the molecular size of the RNA species capable of reconstituting PrPres amplification,
we fractionated our preparation of total hamster brain RNA by ultrafiltration through a filter with a
relative molecular mass cutoff of approximately 100,000 (M_r 100K). Using agarose gel
electrophoresis, we detected all of the ribosomal RNA bands in the retentate and all of the transfer
RNA in the filtrate (data not shown). Using these samples, we discovered that the filter retentate
was capable of reconstituting PrPres amplification to a level slightly lower than unfractionated total
brain RNA. By contrast, the filtrate was not able to reconstitute PrPres amplification (Fig. 3c).
These data indicate that most of the reconstitution activity is conferred by RNA molecules >100K
in size (>300 nucleotides).

There is currently a need to develop more sensitive diagnostic tests for prion disease—this might be
achieved by increasing the efficiency of PrPres amplification techniques. We therefore investigated
whether addition of total hamster brain RNA could increase the efficiency of PrPres amplification in
vitro in brain samples not pre-treated with nuclease. We mixed a more dilute homogenate of
prion-infected brain (0.02% w/v) with 5% (w/v) normal brain homogenate overnight without
sonication, and measured PrPres amplification. Our results show that addition of total hamster brain
RNA to this mixture of intact brain homogenates significantly stimulates PrPres amplification over
baseline (Fig. 4a). As a control, we confirmed that addition of RNA did not alter the level of input
PrPres or PrP^C in these samples (Supplementary Fig. S3). Densitometric measurements indicate
that PrPres in 0.02% (w/v) prion-infected brain homogenate samples is amplified about sixfold after
overnight incubation, similar to the PrPres amplification level previously reported for 0.1% (w/v)
prion-infected brain homogenate3. By contrast, PrPres in samples amplified with RNA is amplified
about 24-fold, indicating that addition of RNA increases the efficiency of in vitro PrPres
amplification about fourfold. Addition of RNA also increased the efficiency of PrPres amplification
of sonicated protein-misfolding cyclic amplification (PMCA) reactions (Supplementary Fig. S4).

Figure 4 Stimulation of PrPres amplification with RNA. a, Immunoblot of PrPres amplification reactions. Samples include mixtures of 5% (w/v) normal and 0.02% (w/v) scrapie brain homogenate (M) and diluted scrapie brain homogenate control (Sc). Indicated samples contained 0.5 mg ml^-1 total hamster brain RNA (+ ). b, Agarose gel electrophoresis of total RNA prepared from various species. c, Immunoblot of species-specific stimulation of PrPres amplification with RNA. Total RNA (0.5 mg ml^-1) prepared from various species was added to PrPres amplification reactions.

To assess the specificity of RNA-mediated stimulation of PrPres amplification, we isolated total
RNA from several sources, including Escherichia coli, Saccharomyces cerevisiae,
Caenorhabditis elegans, Drosophila melanogaster, and mouse and hamster brain. Agarose gel
electrophoresis analysis of these preparations revealed the expected band patterns for each species
and confirmed that each preparation contained high-quality, non-degraded RNA (Fig. 4b).
Furthermore, each of these preparations was substantially free from contaminants as judged by
optical spectroscopy (A₂₆₀/A₂₈₀ > 1.9; where A indicates absorbance and subscript numbers
indicate wavelength). Notably, among the six preparations of RNA tested, only hamster and mouse
brain RNA could stimulate PrPres amplification in vitro (Fig. 4c). This apparent species specificity
cannot be attributed to tissue specificity because total hamster liver RNA also stimulated PrPres
amplification (data not shown). This argues that mice and hamsters express specific RNA
molecules required for PrPres amplification. Additional experiments show that the RNA stimulation
activity within the Trizol-extracted hamster brain RNA preparation was irreversibly destroyed by
glyoxylation, but not by deproteination, heating to 60 °C, or transient exposure to 50% formamide
for 1 h (Supplementary Fig. S5).

If PrPres amplification studies accurately model PrP^Sc formation in vivo, the results presented here
represent a significant advance in our understanding of the mechanism of prion conversion.
Previously, it has been shown that purified PrP^C can be converted into protease-resistant PrPres in
vitro in the absence of cellular cofactors [7]. However, the fact that a 50-fold molar excess of
purified PrPres is required to drive conversion of purified PrP^C suggests that efficient PrPres
formation may depend on the presence of cellular factors other than PrP^C (ref. 8). On the basis of
the results presented here, we propose the hypothesis that specific RNA molecules are cellular
cofactors for PrP^Sc formation. Consistent with our hypothesis that specific RNA-converting factors
stimulate PrP^Sc formation, nucleic acids bind avidly to and promote conformational change of
recombinant PrP (refs 9–14). However, it is important to note that full-length, refolded recombinant
PrP lacking post-translational modifications cannot undergo stoichiometric conversion to PrPres
(Supplementary Fig. S6), and therefore the results of biophysical studies using recombinant PrP
cannot be directly related to the results described here. It has been proposed that PrP^Sc molecules
might bind to specific host RNA molecules to generate strain diversity [15]. Whether the
RNA-converting factors we describe are also involved in generating strain diversity remains to be
determined. Finally, it is important to emphasize that the existence of RNA-converting factors is
fully consistent with the protein-only hypothesis proposed previously [1], because the nucleic acids we
describe are host-encoded and not contained within the infectious agent.

Methods:

Animal and reagent sources: Specific-pathogen-free female golden Syrian hamsters at 3 weeks
old were purchased from Charles River Laboratories. Apyrase, DEPC, cyclic 2',3'-GMP,
3'-CMP, heparinase III, heparan sulphate proteoglycan (Mr >200K), polyadenylic acid (M_r
200–2,000K) and polyglutamic acid (M_r 50–100K) were obtained from Sigma; RNase-free
DNase, micrococcal nuclease, RNase A and DNase-free RNase were obtained from Roche;
RNase T1 was obtained from Epicentre; recombinant benzonase nuclease was purchased from
Novagen; EcoRI was obtained from Gibco BRL; and RNase H and RNase V1 were obtained
from Ambion.

In vitro PrPres amplification: In vitro PrPres amplification [3] and PMCA [4] were performed as
previously described, except that normal brain homogenates were prepared with EDTA-free
complete protease inhibitors (Roche) to facilitate experiments involving metal-dependent enzymes.
Two millimolar MgCl₂ was added to reactions with benzonase and 2 mM CaCl₂ was added to
reactions with micrococcal nuclease and apyrase. All amplification and control reactions were
performed at 37 °C for 16 h. For PrPres detection, protease digestion was performed with
50 µg ml-1 proteinase K for 1 h at 37 °C and immunoblotting was performed with 3F4 monoclonal
antibody (Signet). For PrP^C detection, samples were not subjected to proteinase K digestion
before immunoblotting. All protein electrophoresis experiments shown were performed on 12%
SDS polyacrylamide gels and reference M_r for such experiments are shown.

Nuclease inactivation: Micrococcal nuclease and benzonase were inhibited with 5 mM EDTA.
RNase A (50 µg) was incubated with 1% DEPC in 100 µl at 25 °C for 2 h. After incubation, the
reaction was dialysed twice against 1  l  10 mM Tris pH 7.2 at 4 °C using a Pierce 3500 MW
Slide-A-Lyzer minidialyis unit to remove free DEPC. Control samples containing active RNase A
were dialysed in parallel. Protein recovery >90% was confirmed by BCA assay (Pierce). Active
and inactivated nucleases were added to amplification reactions at concentrations designated 'X' in
Fig. 1. 'No enzyme' control samples were processed in parallel.

Reconstitution assays: Nuclease digestion before reconstitution was performed by incubating a
batch of normal brain homogenate (10% w/v) with benzonase (final concentration of 2.5 U µl^-1)
and 2 mM MgCl₂ for 16 h at 4 °C in the absence of detergents. Benzonase was then inactivated by
the addition of 5 mM EDTA before reconstitution with RNA or other polyanions.

Preparation and measurement of RNA: RNA was isolated from animals <5 min after death
using rotor–stator homogenization, extraction with Trizol reagent (Invitrogen) for 5 min at 25 °C,
and isopropanol precipitation according to manufacturer's instructions, using RNase-free reagents,
containers and equipment. For yeast, cell walls were disrupted during extraction as previously
described, using Trizol in place of pheno [16]. All RNA solutions were alcohol precipitated, washed
and resuspended in RNase-free water before use. The concentration and purity of each solution
was determined by spectroscopic measurement of absorbance at l₁/ l₂ = 260/280 nm and
confirmed by electrophoresis on 1% agarose gels stained with ethidium bromide.

RNA size fractionation: Total hamster brain RNA (0.4 mg) was diluted into 0.8 ml RNase-free
water, loaded in 0.2-ml batches onto four separate Schleicher and Schuell Centrex UF-05 (100K
cutoff) ultrafiltration devices, and centrifuged for 15 min at 3,000g. The devices were then washed
with an equal volume of water. The filtrates were pooled and retentate fractions collected by briefly
centrifuging the ultrafiltration devices upside down into new microcentrifuge tubes. Parallel samples
of denatured retentate were prepared in 50% formamide to disrupt all intra- and intermolecular
interactions.

Reverse transcriptase polymerase chain reaction: RT–PCR was performed using the One
Step RNA PCR kit (AMV) from Takara/Fisher following the manufacturer's instructions, using the
PrP-specific primers 5'-CGAACCTTGGCTACTGGCTGCTG-3' and
5'-GCTTGATGGTGATATTGACGCAGTC-3', and the following parameters: reverse
transcription at 50 °C for 15 min, heat inactivation of reverse transcriptase at 94 °C for 2 min,  X 25
PCR cycles (94 °C for 30 s, 55 °C for 30 s, 72 °C for 90 s). Products were run on a 1% agarose
gel and stained with ethidium bromide.

Supplementary information accompanies this paper:
http://www.nature.com/cgi-taf/DynaPage.taf?file=/nature/journal/v425/n6959/abs/nature01979_fs.html&dynoptions=doi1066663288
 

References:

1.
    Prusiner, S. B. Novel proteinaceous infectious particles cause scrapie. Science 216, 136-144 (1982)

2.
    Prusiner, S. B. (ed.) Prion Biology and Diseases (Cold Spring Harbor Laboratory Press, Cold
    Spring Harbor, New York, 1999)

3.
    Lucassen, R., Nishina, K. & Supattapone, S. In vitro amplification of protease-resistant prion
    protein requires free sulfhydryl groups. Biochemistry 42, 4127-4135 (2003)

4.
    Saborio, G. P., Permanne, B. & Soto, C. Sensitive detection of pathological prion protein by
    cyclic amplification of protein misfolding. Nature 411, 810-813 (2001)

5.
    Lockard, R. E. & Kumar, A. Mapping tRNA structure in solution using double-strand-specific
    ribonuclease V1 from cobra venom. Nucleic Acids Res. 9, 5125-5140 (1981)

6.
    Banks, G. R. A ribonuclease H from Ustilago maydis. Properties, mode of action and substrate
    specificity of the enzyme. Eur. J. Biochem. 47, 499-507 (1974)

7.
    Kocisko, D. A. et al. Cell-free formation of protease-resistant prion protein. Nature 370, 471-474 (1994)

8.
    Caughey, B., Horiuchi, M., Demaimay, R. & Raymond, G. J. Assays of protease-resistant prion
    protein and its formation. Methods Enzymol. 309, 122-133 (1999)

9.
    Derrington, E. et al. PrPC has nucleic acid chaperoning properties similar to the nucleocapsid
    protein of HIV-1. C. R. Acad. Sci. III 325, 17-23 (2002)

10.
    Moscardini, M. et al. Functional interactions of nucleocapsid protein of feline immunodeficiency
    virus and cellular prion protein with the viral RNA. J. Mol. Biol. 318, 149-159 (2002)

11.
    Gabus, C. et al. The prion protein has RNA binding and chaperoning properties characteristic of
    nucleocapsid protein NCP7 of HIV-1. J. Biol. Chem. 276, 19301-19309 (2001)

12.
    Gabus, C. et al. The prion protein has DNA strand transfer properties similar to retroviral
    nucleocapsid protein. J. Mol. Biol. 307, 1011-1021 (2001)

13.
    Nandi, P. K., Leclerc, E., Nicole, J. C. & Takahashi, M. DNA-induced partial unfolding of prion
    protein leads to its polymerisation to amyloid. J. Mol. Biol. 322, 153-161 (2002)

14.
    Cordeiro, Y. et al. DNA converts cellular prion protein into the -sheet conformation and inhibits
    prion peptide aggregation. J. Biol. Chem. 276, 49400-49409 (2001)

15.
    Weissmann, C. A 'unified theory' of prion propagation. Nature 352, 679-683 (1991)

16.
    Chapon, C., Cech, T. R. & Zaug, A. J. Polyadenylation of telomerase RNA in budding yeast. RNA
    3, 1337-1351 (1997)

Additional References:

Acknowledgements. The authors thank G. Saborio, C. Soto, V. Ambros, C. Cole and W.
Wickner for helpful advice. This work was supported by the Burroughs Wellcome Fund Career
Development Award, the Hitchcock Foundation, and an NIH Clinical Investigator Development
Award.

Competing interests statement. The authors declare that they have no competing financial
interests.

1. Adler V, Zeiler B, Kryukov V, Kascsak R, Rubenstein R, and Grossman A, "Small, Highly Structured RNAs Participate in the Conversion of Human Recombinant PrP^Sen to PrP^Resin vitro".

2. Frenster JH, "Ultrastructural Probes of Active DNA Sites, and the RNA Activators of DNA".

Top of Page - Euchromatin Network - Current Research - Forums - Other Sites - Future Events -

For Further Information or Feedback:
e-mail:   frenster@euchromatin.net
Phone:   +1 650 367 6483
Fax:   +1 650 364 1773