Previous Article | Next Article ![]()
Journal of Virology, April 2004, p. 3654-3662, Vol. 78, No. 7
0022-538X/04/$08.00+0 DOI: 10.1128/JVI.78.7.3654-3662.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Agence Française de Sécurité Sanitaire des Aliments, Unité Virologie-ATNC, 69364 Lyon Cedex 07,1 CEA, Service de Pharmacologie et d'Immunologie, CEA/Saclay, 91191 Gif sur Yvette Cedex,3 CEA, GIDTIP/DRM/DSV, CEA/Fontenay aux Roses, Fontenay-aux-Roses, France,4 Veterinary Laboratory Agency, VLA Lasswade, Penicuik, Midlothian EH26 OPZ, United Kingdom2
Received 27 August 2003/ Accepted 10 December 2003
|
|
|---|
|
|
|---|
The origin of the cattle BSE agent remains unknown, but it most probably originated from adaptation and recycling of the sheep scrapie agent in cattle (46, 47). Irrespective of the origins of BSE in cattle, the susceptibility of sheep to BSE has been demonstrated by using several different routes of infection (including the oral route) (2, 14-16, 29). The possibility of spread of this agent to sheep flocks is therefore of considerable concern, as exposure to BSE-contaminated meat and bone meal in some flocks is a real possibility (8). Thus, the precise identification of the BSE agent in sheep is needed, particularly as it may represent a potential new source of contamination for human beings. This identification of BSE in sheep cannot be made by analysis of clinical signs, as BSE-infected sheep show signs similar to those observed in naturally occurring scrapie (2, 15, 25).
The biochemical analysis of abnormal prion protein (PrP), the only specific marker of these diseases (6), can potentially contribute to strain identification. Abnormal PrP is derived from a normal host protein named PrPc. The abnormal PrP is partially resistant to protease digestion (PrPres) and gives a 27- to 30-kDa fragment after proteinase K (PK) treatment (6); in rodent scrapie a sequence of 62 N-terminal amino acids is digested, leaving a core of 141 amino acids (33). By using the Western blotting method and specific antibodies directed against the core globular domain of the PrP, three fragments can be separated by their relative molecular weights; these correspond to the diglycosylated, monoglycosylated, and nonglycosylated forms of the protein. While high levels of diglycosylated PrPres were found in sheep experimentally infected with BSE, natural sheep scrapie and BSE of cattle could not be readily distinguished according to ratios of the different PrPres glycoforms (5, 40). Some studies have identified a lower molecular mass of the unglycosylated PrPres in variant CJD compared to most other forms of CJD in humans and also in experimentally BSE-infected sheep compared to natural scrapie cases (2, 9, 21, 40). However, in sheep, a low molecular mass has also been found in the experimental scrapie isolate CH1641 (2, 22, 23, 40), although some molecular differences from the BSE agent in sheep have been reported (40).
Recently, an immunohistochemical approach for identification of BSE in sheep has also been achieved by characterizing the epitopes of abnormal PrP present in neurons and also in phagocytic cells of the lymphoreticular system and brain. This method does not use protease treatment, and the term PrPd is used to describe the abnormal disease-specific accumulations of PrP seen by immunohistochemistry. However, by using different peptide-specific antibodies directed against the PK cleavage region of the PrP, a more extensive digestion of the abnormal PrP was revealed in tingible body macrophages (TBMs) in lymphoid tissues and in glial cells and neurons in the brains of BSE-infected sheep compared to natural scrapie cases (27, 28).
Taken together, all of the data strongly suggest that PrPres from animals or humans infected by the BSE agent is more susceptible to PK treatment than PrPres found in scrapie-infected animals. As a consequence, a larger segment of the N-terminal sequence of PrPres is removed by proteolytic treatment, leading to a loss of the corresponding epitopes. This has already been illustrated in previous studies showing that the epitope recognized by the P4 monoclonal antibody (20) was specifically unexposed in BSE-infected cattle and sheep while kept in scrapie-infected sheep, apart from the CH1641 experimental scrapie isolate (40). On the basis of such observations, a new enzyme-linked immunosorbent assay (ELISA) was developed, which checks for the presence of the octa-repeat sequence following PK treatment as a way to identify the BSE agent in sheep (39).
In order to identify potential BSE cases in flocks, a series of 214 scrapie samples from affected French flocks was analyzed by the ELISA technique. In this screening step, most of the brain samples behaved like conventional scrapie samples, but 18 of them showed an increased protease cleavage without reproducing the behavior observed in a BSE-infected sheep control. Interestingly, when tested by Western blotting, 2 of these 18 samples, both originating from the same flock, showed an unglycosylated band with a lower molecular mass than in other scrapie samples and similar to that found in BSE-infected sheep. A precise characterization of these two cases was made in comparison to experimental BSE in sheep and natural scrapie cases, by Western blotting and by using an immunohistochemical analysis of PrPd processing in phagocytic cells of the lymphoreticular system and in brain.
|
|
|---|
Sheep TSE isolates. The sheep used in this study are described in Table 1.
|
View this table: [in a new window] |
TABLE 1. Sheep used in the study
|
(ii) Group 2. Group 2 consisted of one sheep with natural scrapie. R680 was a naturally diseased ARQ/ARQ British Suffolk sheep which died 892 days after birth. The animal showed typical vacuolar pathology and PrPd accumulation as previously characterized in this flock (18).
(iii) Group 3. Group 3 consisted of four Manech Tête Rousse sheep from the same affected flock for which the first official declaration of scrapie occurred in 1996. These sheep were euthanatized between December 1996 and February 1997 when they were between 2 and 5 years of age and presenting with clinical signs of scrapie. The histological examination of fixed brain stems confirmed the diagnosis of TSE. Only frozen frontal brain regions were available for biochemical studies. Of these four samples, two (O100 and O104) showed moderate or severe autolysis as determined by routine histology methods. Preliminary studies of PrPres patterns in sheep by using Western blotting had shown an usual pattern in the O100 and O104 isolates, with a low molecular mass of the unglycosylated PrPres, compared to previous data obtained in the laboratory (2).
(iv) Group 4. Group 4 included four sheep with histologically confirmed TSE detected from three other outbreaks within the French surveillance network. Biochemical studies were also performed on frontal brain regions, and histological analysis was performed on hindbrains.
ELISA procedure. The ELISA used in this study was designed to distinguish BSE from scrapie. It is similar in design to a rapid test used on a large scale in Europe (the Bio-Rad test formally developed by CEA). In this test PrPres is first purified and concentrated before being denatured and analyzed by using a sandwich immunoassay (19). The capture antibody, immobilized into the solid phase, recognizes an epitope in the N terminus of the PrP, whereas the tracer antibody binds to the C-terminal moiety. In the Bio-Rad tests used for the postmortem diagnosis of BSE in cattle, the PK digestion, which is a key step in this assay, is done in a controlled medium (a mixture of detergents and chaotropic agents) so that the N-terminal epitope recognized by the capture antibody is preserved during the proteolysis. By varying the conditions of PK treatment (notably by altering the nature and concentration of the detergents and chaotropic agents and the PK concentration), we have found that the PrPres associated with the BSE agent was more sensitive to PK digestion than other prion strains. Conditions were thus defined in which the epitope recognized by the capture antibody (here the octa-repeat sequence) was eliminated by the PK treatment in the BSE agent but remained preserved in scrapie strains. In practice, we used two sets of conditions of PK treatment. Compared to the Bio-Rad test validated for detection of PrPres in cattle BSE, this typing involves several modifications, including a change in the antibodies used for the sandwich immunoassay and the extensive use of PK inhibitors in order to have a better control of PK digestion. A more detailed description of this test will be given elsewhere (S. Simon et al., unpublished data). Under the first set of digestion conditions (set A), the critical epitope is conserved whatever the strain is. In the second set of conditions (set A'), the treatment preferentially eliminates the N-terminal epitope described above in BSE but not in scrapie. Ratios between the two measurements (A/A') were determined for differential detection of BSE and natural scrapie. This ratio is equal (or close) to unity with most natural scrapie isolates and is significantly above unity with the BSE agent.
Western blot procedure. Brain samples were homogenized at 10% in a 5% glucose solution by forcing the brain suspension through a 0.4-mm-diameter needle. A 330-µl volume was made up to 1.2 ml in 5% glucose before incubation with PK (10 µg/100 mg of brain tissue) (Roche) for 1 h at 37°C. N-Lauroyl sarcosyl at 30% (600 µl; Sigma) was added. After incubation at room temperature for 15 min, samples were centrifuged at 200,000 x g for 2 h on a 10% sucrose cushion in a Beckman TL100 ultracentrifuge. The pellets were resuspended and heated for 5 min at 100°C in 50 µl of denaturing buffer (sodium dodecyl sulfate, 4%; ß-mercaptoethanol, 2%; glycine, 192 mM; Tris, 25 mM; sucrose, 5%).
Samples were subjected to sodium dodecyl sulfate-15% polyacrylamide gel electrophoresis and electroblotted to nitrocellulose membranes in transfer buffer (Tris, 25 mM; glycine, 192 mM; isopropanol, 10%) at a constant 400 mA for 1 h. The membranes were blocked for 1 h with 5% nonfat dry milk in phosphate-buffered saline-Tween 20 (0.1%) (PBST). After two washes in PBST, the membranes were incubated (1 h at room temperature) with Bar233 monoclonal antibody (1/5,000 in PBST), which was raised against an ovine histidine-tagged recombinant protein (3) and recognizes the ovine PrP sequence from position 144 to 155 (FGNDYEDRYYRE) (5), or with P4 monoclonal antibody (1/5,000), which was prepared against a synthetic ovine PrP sequence from position 89 to 04 (GGGGWGQGGSHSQWNK) (from R-biopharm, Darmstadt, Germany) (20). After three washes in PBST, the membranes were incubated (30 min at room temperature) with peroxidase-labeled conjugates against rabbit or mouse immunoglobulins (1/2,500 in PBST) (Clinisciences). After three washes in PBST, bound antibodies were then detected with enhanced chemiluminescence (Amersham) or Supersignal (Pierce) chemiluminescent substrates. Migrations of the unglycosylated PrPres were compared for repeated runs of the samples, either on films after exposure of the membranes on Biomax MR Kodak films (Sigma) or by using pictures obtained with the Fluor-S Multimager (Bio-Rad). Molecular masses of the three PrPres glycoforms were determined in some selected samples by comparison of at least eight different runs of the samples of the center positions of the PrPres bands with a biotinylated marker (SDS-6B; Sigma), using Quantity One software (Bio-Rad). The intensities of PrPres signals were assessed by using the Fluor-S-Multimager software, allowing comparisons of reactivities with either Bar233 or P4 monoclonal antibody. Scoring of the reactivities used according to the decrease of P4 labeling: -, no labeling; ±, light labeling; +, moderate labeling; and ++, intense labeling. For quantitative studies of glycoform ratios, chemiluminescent signals corresponding to the three glycoforms of the protein were quantified by using a Fluor S-Multimager analysis system. Glycoform ratios were expressed as mean percentages (± standard errors) of the total signal for the three glycoforms (high, low, and unglycosylated), from at least eight different runs of the samples.
Immunohistochemical procedure. Tissues were fixed in 10% formalin and then routinely embedded in paraffin wax. To ensure adhesion, tissues sections (5 µm thick) were collected onto pretreated glass slides (StarFrost; Fisher Scientific, Pittsburgh, Pa.) and dried. Once they were dewaxed, rehydrated sections were used for disease-specific PrPd immunohistochemical analysis. Briefly, sections were immersed in 98% formic acid for 5 min, washed in running tap water, and then immersed in 0.2% citrate buffer to be autoclaved for 30 min at 121°C. Endogenous peroxidase was blocked by using 1% H2O2 in ethanol for 20 min. After rinsing in tap water and in PBST (0.2% Tween 20) buffer, nonspecific tissue antigens were blocked in normal horse serum for 60 min. Incubation with primary antibody was then performed overnight at 4°C. Different PrP antibodies (Table 2) were used, with the following specific concentrations: R521, 1/10,000; P4, 1/2,000; 505, 1/10,000); and R486, 1/10,000. The first three are directed against the upstream segment of the flexible tail (37) of the abnormal isoform of PrP, and R486 recognizes a more resistant region of the abnormal PrP (28). Bound antibodies were labeled by means of a commercial immunoperoxidase technique (Vector-Elite ABC; Vector Laboratories, Peterborough, United Kingdom). Once labeled, sections were immersed in 0.5% copper sulfate to enhance the intensity of the reaction product. Slides were counterstained with Mayer's hematoxylin.
|
View this table: [in a new window] |
TABLE 2. Antibodies used and their specificity for PrP epitopes
|
|
|
|---|
Western blotting study of sheep TSE isolates. All but one (O180, group 4) of the eight field scrapie samples studied by Western blotting gave a characteristic protein banding pattern corresponding to the three glycoforms of PrPres, with the diglycosylated (top), the monoglycosylated (middle), and the unglycosylated (bottom) bands (Fig. 1). It should be noted that samples O100 and O104 (group 3), strongly autolysed, showed only low levels of PrPres, and smears were regularly observed on Western blotting.
![]() View larger version (31K): [in a new window] |
FIG. 1. Western blotting analysis of PrPres with Bar233 (A and B) or P4 (C) monoclonal antibody. (A) Lanes: 1, cattle BSE; 2, BSE-infected sheep (intraperitoneal inoculation); 3, sheep scrapie case O100; 4, BSE-infected sheep (intrasplenic inoculation); 5, natural scrapie case. (B and C) Lanes: 1, sheep scrapie case O100; 2, BSE-infected sheep (intraperitoneal inoculation); 3, sheep scrapie case O104; 4, sheep scrapie case O58; 5, cattle BSE.
|
The molecular masses and glycoform ratios obtained from repeated runs of these two samples (O100 and O104) are shown in Fig. 2B, in comparison with another sample from a sheep (O58) with the same homozygous V136 R154 Q171 genotype from the same flock and with BSE infection in cattle or in sheep.
![]() View larger version (18K): [in a new window] |
FIG. 2. Glycoform ratios (A) and molecular masses (B) (means ± standard deviations) of PrPres detected by Western blotting with the Bar233 monoclonal antibody in sheep infected with natural scrapie or experimental BSE and in cattle with BSE.
|
Immunohistochemical study. (i) Lymphoid organs: TBMs. Lymphoid organ samples were immunolabeled with R521, P4, 505, and R486 antibodies. The results show that most antibodies labeled follicular dendritic cells (FDCs) and TBMs with different intensities (Fig. 3).
![]() View larger version (139K): [in a new window] |
FIG. 3. Comparison of immunolabeling of PrPd in lymph nodes with the P4, R521, 505, and R486 antibodies on serial sections. TBMs were labeled with all of the antibodies used in the natural cases of sheep TSE (brown deposits; arrow) (row 1) and in O100 (row 2) but not in experimentally infected sheep with BSE (row 3), where only R486 gave labeling. In row 2, the O100 sheep follicle shows numerous tissue spaces within the light zone and also shows separation of the follicular capsule from adjacent paracortical lymphocytes. These features are not present in the other two samples shown and are caused by postmortem changes consistent with autolysis. A x10 objective was used.
|
(ii) Central nervous system. At the level of the obex sections, the patterns of PrPd immunolabeling produced by different antibodies were compared for the different sheep analyzed (natural sheep TSE isolates and experimental BSE in sheep). For each brain stem sample, the dorsal nucleus vagus (X) and the hypoglossal (XII) and olivary nuclei were studied in order to characterize the intraglial and intraneuronal PrPd labeling patterns.
(a) Intraglial labeling. PrPd was detected with R521, P4, 505, and R486 antibodies in all naturally occurring TSE cases from the United Kingdom and France analyzed (groups 2, 3, and 4) (Fig. 4). The precise identification of the labeling was always indicated by the presence of a nucleus (in blue) closely associated with a coarse particulate type of PrPd deposit (brown deposits) (Fig. 4). This same pattern, with a similar intensity, was found for TSE cases O100 and O104 (Fig. 4). In contrast, in all BSE-affected sheep, whereas the 505 and R486 antibodies produced a similar intracellular PrPd labeling of glial cells, P4 and R521 did not produce this type of labeling (Fig. 4).
![]() View larger version (136K): [in a new window] |
FIG. 4. Comparison of intraglial PrPd immunolabeling (brown deposits; arrow) with the P4, R521, 505, and R486 antibodies on serial sections of brain stem. Intraglial PrPd was labeled with all of the antibodies used in the cases of natural sheep TSE (row 1) and in O100 (row 2) but not in sheep experimentally infected with BSE (row 3) (arrowheads), where only R486 and 505 gave labeling of glia. A x100 objective was used.
|
![]() View larger version (112K): [in a new window] |
FIG. 5. Comparison of intraneuronal PrPd immunolabeling (brown deposits; arrow) with the P4, R521, 505, and R486 antibodies on serial sections of brain stem. Intraneuronal PrPd was labeled with all of the antibodies used in the natural cases of sheep TSE (row 1) and in O100 (row 2) but not in sheep experimentally infected with BSE (row 3), where no intraneuronal immunolabeling was obtained with P4 (arrowheads). A x40 objective was used.
|
|
View this table: [in a new window] |
TABLE 3. Comparative intensity of intraneuronal immunolabeling in dorsal nucleus vagus (X), hypoglossal nucleus (XII), and olivary nucleus with the P4 and R486 antibodies
|
|
|
|---|
The identification of the precise cell type accumulating PrPd in TSE-affected sheep can be based on the shape and the localization of the implicated cells or on specific markers. In lymphoid organs, FDCs and TBMs were identified as accumulating abnormal PrP (31, 41), as well as different cell populations in the central nervous system (18, 38). Extracellular forms of PrPd released around neurons, astrocytes, and FDCs of sheep infected with scrapie are present as the full-length protein (26). In TBMs, glial cells, and neurons, intralysosomal PrPd appeared to be truncated by protein-degrading enzymes, and antibodies such as BG4 and FH11 did not label PrPd in these cells, as they are directed against the downstream segment of the N-terminal end of PrP (28, 37). It is therefore suggested that the N terminus of PrPd acquired from the extracellular space is digested within phagocytic cell lysosomes. A more detailed analysis using P4, R521, and 505 antibodies directed against the upstream segment of the flexible tail of the PrP molecule permits differentiation of BSE in sheep from different sheep scrapie sources (27, 28). This differentiation is possible because the degradation of the abnormal PrP in lysosomes seems to be more extensive in BSE infection than in natural scrapie. In sheep BSE, the P4, R521, and 505 antibodies do not label PrPd in TBMs, in contrast to what was observed for scrapie. The specific degradation of PrPd is, however, variable in different cells, being more extensive in TBMs, less so in glial cells, and less again in neurons. The precise site of truncation delimiting the most resistant domain of the abnormal PrP is therefore difficult to determine because of this cellular variation of PrPd labeling and because results can vary according to pretreatments used for PrPd antigen retrievals. Nevertheless, all of the results suggest that a more extensive truncation of the PrPd occurred in phagocytic cells and in neurons in BSE-infected sheep than with scrapie infection, and these immunohistochemical results are consistent with those obtained by using the Western blotting method after PK treatment. The more extensive protease digestion of the PrPres extracted from BSE-infected animals has been characterized by epitope mapping following Western blotting detection or by N-terminal sequencing of the PrPd protein following transmission in sheep (40), mice (4), and humans (34).
Following an initial screening by using an ELISA technique aimed at identifying the BSE agent in sheep based on the reactivity against an epitope close to the PK cleavage site, Western blotting studies of PrPres extracted from the brains of French sheep field TSE cases led to the finding of two particular cases (O100 and O104), of the same genotype (VRQ/VRQ) and from a same flock, that showed some similarities with BSE in sheep regarding the molecular mass of the unglycosylated PrPres. Other cases analyzed in this study (O58 and O54 [group 3], including a sheep of this same genotype from this flock, and four sheep from other TSE outbreaks [group 4]) showed the PrPres pattern previously described for natural scrapie cases, with a clearly distinct and higher molecular mass compared to that for BSE in sheep (2, 22, 40). So far, a single sheep TSE isolate, CH1641, with close similarities in the migration of unglycosylated PrPres to that found for BSE in sheep has been described. CH1641 has been experimentally maintained by serial passages in sheep from a scrapie case that occurred in the United Kingdom in the 1970s (2, 13, 23). The BSE agent and CH1641 in sheep also shared a strongly decreased recognition by the P4 monoclonal antibody compared to that in natural scrapie cases (40). P4 labeling was also decreased in O100 and O104 compared to a scrapie control of the same genotype and from the same flock, but to a lesser extent than for BSE in sheep. These results are in good agreement with those obtained by ELISA characterization of protease digestion, showing an increased protease digestion without reaching the level found for BSE in sheep. In contrast with immunohistochemical methods, biochemical analysis does not allow characterization of the cell types in which PrPres accumulates. The natural TSE sheep samples described above were studied by immunohistochemical methods, using antibodies close to the protease cleavage site, in brain and lymphoid tissues. All of them, including those from the two affected sheep with unusual Western blotting patterns, O100 and O104, had a typical scrapie-like signature by immunohistochemical analysis. We also observed differences of glycoform ratios between natural sheep TSE cases, including O100 and O104, and experimental BSE in sheep. These differences, mainly linked to the very high levels of diglycosylated PrPres in BSE-infected sheep, are quite consistent with those previously described for British cases of natural scrapie or experimental BSE, using a modified version of the Western blotting Prionics test and 6H4 antibody (40). Together these data suggest that O100 and O104 do not share all the features found in BSE in sheep. Nevertheless, criteria which allow the distinction of ovine BSE from scrapie need to be more clearly defined, especially when quantitative assessment of the glycoform ratios or differential reactivities of antibodies directed toward the protease cleavage sites are considered, as these methods are difficult to standardize.
The significance of the biochemical properties of sheep isolates O100 and O104 is unclear, and several hypothesis can be considered to explain these observations (1). Since the nonfixed brain material from these two sheep was autolysed, an endogenous protease action, might have predigested the tissue and abnormal PrP during the autolysis process, thereby facilitating or enhancing the PK action. However, among the frozen samples received in the laboratory for diagnosis, a number of field samples were also autolysed without showing any unusual protease cleavage. Similarly, use of variable concentrations of PK did not lead to changes in the molecular weight, and only the signal intensity of PrPres detection was modified (24, 30).
Several procedural steps which contribute to variations in electrophoretic mobility or/and glycoform ratios have previously been identified. These included metal ion chelations, prior to PK treatment (43), PrPres extraction under different pH conditions (49, 50), and the use of sodium phosphotungstic acid precipitation before PK treatment (44). Whereas in these situations changes in the PrPres patterns were the result of modifications in the laboratory methods, it should be emphasized that they could reveal a genuine biological diversity of these diseases.
Our results could be explained by the presence in these sheep of a particular strain of TSE agent with some properties close to those found in the BSE agent and in the CH1641 scrapie isolate. While the unusual electrophoretic features were similarly found in two sheep of the same genotype in the same flock, but not in some other sheep of this flock, characterization of the infectious agent involved in these scrapie cases is required, using mouse transmission studies in a panel of mice with different prn-p genotypes, as used for strain typing (7), and in other transgenic mouse models (10, 11, 42).
In conclusion, our results have reinforced the idea that uniform features could be found in sheep experimentally infected with BSE, including those infected with BSE from different geographical origins. Importantly, through the study of a panel of field sheep TSE cases, two with some biochemical features similar to those of ovine BSE were found. These two samples were initially identified from a group of 18 following application of a rapid ELISA test, which may be useful for screening of large series of samples. The accurate identification of the infectious agent strain in sheep, however, requires a careful analysis by the different available methods, including those which aid characterization of different molecular properties that may characterize TSE strains and mouse transmission studies.
Stéphane Lezmi was financially supported by a grant from Agence Française de Sécurité Sanitaire des Aliments (AFSSA).
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»