WO2009040534A1 - Diagnostic method - Google Patents

Abstract

A method for typing a strain of transmissible spongiform encephalopathy in an animal, said method comprising (A) obtaining a sample of truncated prion protein that is resistant to digestion by protease and that is characteristic of transmissible spongiform encephalopathy; (B) determining the identity and amounts of protein types by establishing the extent and site(s) of truncation; (C) determining the relative proportions of said identified and quantified protein types in relation to the sum of their concentration within the sample; and (D) comparing said relative proportions of said protein types to those found in specific types of TSE and relating the result to the specific TSE type found.

Description

Diagnostic Method

The present invention relates to a method of typing strains or forms of transmissible spongiform encephalopathies or prion disease found in infected animals, as well as to diagnostic kits and reagents used in the method. In particular, the applicants have found that the method provides a technique for distinguishing between experimentally transmitted BSE in sheep and classical scrapie in sheep, as well as between unusual types of scrapie. Thus the technique is useful in detecting whether BSE does exist in the sheep population of the country, as well as clarifying the different types of scrapie, which appears now to be a heterogeneous disease.

The mammalian prion protein, PrP^c, is a membrane-anchored protein, predominantly expressed in neural tissue. Abnormal forms of the prion protein, PrP^Sc, are associated with the neurodegenerative disease transmissible spongiform encephalopathy (TSE) .

Many techniques have been described to detect the presence or levels of PrP^Sc as a means for diagnosing TSEs. Generally, in these methods, samples from a test animal are digested using a protease enzyme that digests normal protein but only partially digests PrP^Sc (at its N-terminus) leaving a protease resistant core protein (PrP^res) Detecting the presence of the PrP^res core proteins provides a basis for TSE disease diagnosis. Other methods, including for example those described in WO02/082919 subjects the samples to digestion conditions in which different "signature" peptides are released from PrP^c as compared to PrP^sc and these are detected using mass spectrometry and a range of standard peptides. However, extension of this method to typing of TSE strains has not been suggested.

TSEs occur as many variants or strains. These are characterised by differences in incubation time, disease progression, distribution of the PrP^Sc molecules or aggregates on whole- organism, tissue and cellular level, and by differences in protease resistance of PrP^Sc. Different TSEs are known for different mammals, but cross-species transmission is possible. A well-known example is the link between human new variant Creutzfeld-Jacob disease (nvCJD) and bovine spongiform encephalopathy (BSE) . Scrapie, a sheep TSE, has been present in the UK national flock and in many other countries for hundreds of years. No proof for transmission of scrapie to humans or cattle exists. Similarly there is no evidence to suggest that BSE has transmitted naturally to sheep, although experimental transmission is possible.

Moreover, possible occurrence of BSE in sheep may be masked by the occurrence of scrapie. Although no proof has been found that BSE has entered sheep flocks, differential testing for BSE is carried out when sheep samples have tested positive for TSE.

An additional complication is the existence of sub-types of scrapie, some of which, such as CH1641, may resemble BSE when analysed by biochemical techniques such as immunoblotting, but are in fact different and have their own, though largely unknown, risk of cross-species transfer. For these reasons, furthering identification and characterization of strains and subtypes of ovine TSE is very important.

Though ultimately strains can only be defined by bioassay, a range of tests have been used for TSE characterization including identification of characteristic lesions in the brain , patterns of PrP^Sc deposition by IHC, or on differences in banding pattern of PrP^Sc by immunoblotting. The latter screening tests make use of the partial resistance of PrP^Sc to proteases, notably proteinase K (PK) , and strain-related differences therein. For this type of screening test, brain tissue is typically homogenized and subsequently digested by incubation with PK under controlled conditions. This eliminates most proteinaceous material, including PrP^c, and partially digests PrP^Sc to leave a protease resistant core fragment, PrP^res. Prp^res can be detected by Western blotting following separation by gel electrophoresis to confirm TSE diagnosis. The PrP^res banding pattern can also show differences for different strains of TSE from the same type of organism. Stack et al . , Acta Neuropathologica 2002, 104:279- 286 demonstrated clear differences in PrP^res banding patterns between experimental BSE and scrapie in sheep. These strain- dependent differences are attributed to a different distribution of accessible PK cleavage sites, associated with the conformation of PrP^Sc. However, ID Western-blotting of PK-treated samples provide relatively low-resolution data. Even with the use of different antibodies to cover different epitopes, resolution is limited to within several amino acid residues.

Development of a higher resolution method with which to detect differences between the N-terminal PK cleavage sites of PrP^res and their relation to the various TSEs is therefore considerably important. Edman degradation has been used on a number of occasions to sequence HPLC-separated Lys-C digestion products of examples of PrP^res, allowing in each case, the most abundant N- terminal PK site to be determined. However, Edman degradation requires a significant amount of starting material and does not easily lend itself to high-throughput applications.

With recent rapid advances in analysis of protein-derived peptide sequences in the field of mass spectrometry, detection of the N-terminal cleavage sites of PrP^res by mass spectrometry has become a viable alternative. Separation by liquid chromatography followed by mass spectrometry to confirm the molecular weights of Lys-C digested hamster PrP^res peptides has been reported previously. Other workers have used MALDI-ToF analysis to identify N-terminal PK-cleavage sites of PrP^res obtained from brain homogenates of subjects affected by various human prion diseases and found marked differences in the main PK cleavage site specific for each disease. In a study utilizing electrospray ionisation of LC-separated tryptic digests of PrP^res obtained from mice inoculated with various TSE strains and detection by tandem (ion trap) MS, the applicants have identified several PK cleavage sites and found strain-dependent differences in their relative abundance. Although these mass spectrometry-based studies have provided important insight in correlation between qualitative differences in N-terminal amino acid profiles and TSE strains, accurate information on the relative frequency of PK cleavage at the various sites and correlations with TSE strains has thus far not been obtained.

The applicants have found however, that high resolution TSE strain typing can be effected by analysing the relative proportions of protease resistant protein types having similar terminal peptide sequences that are formed or exist in truncated prion protein fragments such as PrP^res that arise from TSEs, for example using a detailed analysis of peptides produced from fragments such as PrP^res based upon mass spectrometric measurements .

According to the present invention, there is provided a method for typing a strain of transmissible spongiform encephalopathy in an animal, said method comprising

(A) obtaining a sample of truncated prion protein that is resistant to digestion by protease and that is characteristic of transmissible spongiform encephalopathy;

(B) determining the identity and amounts of protein types by establishing the extent and site(s) of truncation,

(C) determining the relative proportions of said identified and quantified protein types in relation to the sum of their concentration within the sample; and

(D) comparing said relative proportions of said protein types to those found in specific types of TSE and relating the result to the specific TSE type found. Generally, in step (A) above, the truncated prion protein that is resistant to digestion by protease and that is characteristic of transmissible spongiform encephalopathy is obtained by subjecting a sample containing prion protein obtained from an animal having or suspected of having a TSE and digesting this with a protease, and in particular a non-specific protease, to remove normal protein, leaving PrP^res. However, analagous proteins (PrP^D) may be formed following endogenous proteolysis (Jeffrey, M., et al.(2006) J Comp Pathol, 134, 17-29). This naturally truncated PrP^Sc may be obtained by purifying PrP^Ξc using well known methods such as use of immunoprecipitation for example as described by Korth, C et al., (1997) Nature, 390, 74- 7, or differential detergent precipitation (PrP^D) as described for example by Hope et al 1988, Eur J Biochem, 172, 271-277 (1988).. In all these cases, the truncated prion protein will have a range of types, each with a predetermined N- or C- terminal end and may be useful in strain differentiation.

In particular the selected PrP^res proteins have predetermined N- terminal ends, as in many forms of the TSE, PrP^Sc is cleaved at specific cleavage points spaced from the N-terminal end, leaving a characteristic N-terminal sequence (s). However, it has been reported that in some atypical strains, fragments derived from the normally protease resistant core at the C-terminal end are formed and these fragments will have a recognisable and constant C-terminal motif. In these cases therefore, the C-terminal sequence may be used in the process.

The applicants have found that the relative proportions of the amount of each particular cleavage site within truncated proteins such as PrP^res and PrP^D is a key feature of the type of TSE and therefore can be used as a basis of strain identification. Thus, if one can detect and measure the amount of the prP^res types (by way of measuring the terminal peptide sequences) produced during for example, a non-specific protease digestion process, a clear indicator of the precise type of strain being investigated can be determined. Protein types having similar truncation sites will contain the same sequence of amino acids within the terminal peptide domains.

Suitably all the various truncated protein types within the sample are detected and quantified in order to provide accurate typing. However, in some cases, the relative amounts of a selection of the truncated protein types may be sufficient to characterise as strain type.

Since the cleavage sites, for example the N-terminal cleavage sites of the various non-specific proteases can be determined, for example using techniques set out below, it is possible to identify and quantify the fragments using various techniques. The cleavage sites may be analysed by determining the terminal amino acid(s) of the truncated prion protein (s).

For example immunological methods may be used in step (B) . Antibodies specific for the N-terminal end (or the C-terminal end in the case of fragments having a characteristic C-terminus) may be bound to the resultant peptide mix, and then quantified using conventional methods. For example, the antibodies may be labelled with a detectable label such as a visible label, or they may be detected and quantified using a subsequent labelling methodology.

Knowing exactly where the cleavages occur allows design of antibodies able to specifically detect the region of cleavage and quantify them by immunoassay. Immunoassays can also be used with reference standards for calibration/interpolation and thus quantification in a conventional manner.

In a particular embodiment, there is provided a method for typing a strain of transmissible spongiform encephalopathy in an animal, said method comprising (A) digesting a sample containing prion protein with a protease so as to remove normal protein and leave protease resistant fragments of abnormal prion protein (Pr^res) of predetermined N- terminal ends; (B) determining the identity and amounts of all fragments having said predetermined N-terminal end peptides within the sample; (C) determining the relative proportions of said identified and quantified peptides in relation to the sum of their concentration within the sample; and (D) comparing said relative proportions to the relative proportions found in specific types of TSE and relating the result to the specific type found.

However, in a preferred embodiment, the step (B) of the method of the invention is effected by carrying out a further digestion, under suitably denaturing conditions, and detecting and quantifying the resultant N- or C- terminal peptides in particular on the basis of the mass/charge ratio.

Thus in a particular embodiment, the invention provides a method for typing a strain of transmissible spongiform encephalopathy in an animal, said method comprising

(1) digesting a sample containing prion protein with a protease so as to remove normal protein and leave protease resistant fragments of abnormal prion protein (Pr^res) ;

(2) further digesting the products of step (1) for example under denaturing conditions to obtain a range of specific peptides with the predetermined N- or C-terminal ends;

(3) detecting the said specific peptides for example on the basis of the mass/charge ratio thereof, and determining the identity and amounts of peptides, preferably all peptides, within the sample;

(4) determining the relative proportions of said identified and quantified peptides in relation to the sum of their concentration within the sample; and (5) comparing said relative proportions to the relative proportions found in specific types of TSE and relating the result to the specific type found.

As before, the specific peptides will generally have a predetermined N-terminal end, produced as a result of cleavage of the products of step (1) by the further digestion, but in some cases, peptides generated from the C-terminal end may be characteristic of the PrP^res type.

The method of the invention has been found to provide particularly good results, in determining even very small differences such as found between strains of scrapie, as well as being able to determine different strains such as ovine BSE (experimentally infection) or unusual scrapie strains.

In particular in the method of the invention, step A or step (1) as applicable is suitably carried out on a sample known to contain prion protein. Such samples, which may be taken from animals, at φost-mortem or from in vitro studies, for example those carried out on cell lines or the like, are generally tissue samples such as brain or lymphoid tissues such as spleen, tonsil, pre-scapular lymph node, mesenteric lymph node, retropharyngeal lymph node, mediastinal lymph node, recto-anal mucosa associated lymphoid tissue (RAMALT) , spleen, and gut associated lymphoid tissue obtained mainly from the ileum and colon, as well as blood. However, generally brain material is the preferred source.

The non-specific proteinase used in step A or step 1 of the reaction may be any one of a variety of available proteases which cut proteins at multiple or random positions. These include for example dispase, atrolysin A, B,. C, E or F, envelysin, thimet oligopeptidase, matrilysin, vibriolysin, coccolysin, mycolysin, meprin A, astacin, leishmanolysin, peptidyl-asp metalloendopeptidase, autolysin, deuterolysin, bothrolysin, stromelysin 1 and 2, bacillolysin, thermolysin, aeromonolysin, leucolysin, mycolysin, pseudolysin, peptidyl-lys- metalloendopeptidase, aureolysin, neprilysin, β-lytic metalloendopeptidase, petidyl-asp metalloendopeptidase, ophiolysin, pitrilysin, insulysin, serralysin or proteinase K (PK) . The particular protease selected in any particular case will suitably be one which yields multiple fragments, with a ragged N-terminal in the particular species being investigated. A particular example of such as protease is Proteinase K, which is widely known and used in the field.

The specific protease used in step (2) of the particular method of the invention is suitably one which selectively cuts at specific residues only. Thus examples include trypsin which cuts at lysine-C or arginine-C, but others include endoproteinase-Arg-C , endoproteinase-Asp-N , chymotrypsin, endoprόteinase-Glu-C, pepsin, proline-endopeptidase. The selection of specific protease in any particular case will depend upon the particular sequence being studied. It is necessary to ensure that the protease used at this point results in a good range of characteristic or signature peptides. These will in general have common or similar C-terminal ends (or N- terminal ends in the case of C-terminal fragments) , which is/are dependent upon the cleavage site of the specific protease and the frequency with which that appears in the range of the peptides resulting from the non-specific protease cleavage. If there is more than one cleavage of the specific protease within this range, then the resulting peptides will have more than one C-terminal end, but these will be broadly similar. If there is only one such site (as there is for example, at the N-terminal end, when trypsin is used after PK cleavage of ovine prion protein from classical scrapie; other potential cleavage sites are essentially blocked by proline) , all the peptides will have a common C-terminal end. Depending upon the particular protease being used, this step would usually be carried out under denaturing conditions, following reduction and alkylation, for example by addition of high concentrations of urea, as is well known in the art, and illustrated hereinafter, to ensure that the PrP^res is unfolded sufficiently to expose the cleavage sites of the specific protease. However, whether this is necessary or not will depend upon the nature of the protein and the nature of the protease.

Suitable methodology used in step • (3) determines the peptides on the basis of the mass/charge ratio and this includes various methods including those in which the sample is introduced into the mass spectrometer using a liquid spray such as an electrospray or from a dried sample by laser desorption (e.g. MALDI) techniques. In a particular embodiment, the technique used is one that may be used in high-throughput application such as MALDI-ToF mass spectrometry. The use of the latter technique is further advantageous because a predominant single charged ion is generally produced during the mass spectrometry, rather than the many multiply charged species which may be obtained using other mass spectrometry methods, which result in highly complex signals .

However, the signals obtainable using these methods may be highly complex, and in particular embodiment, these may be simplified by utilising a preliminary chromatography step, and in particular a liquid chromatography step, to simplify the resultant signal.

Conveniently, the identification and quantification is effected in the same mass spectrometry technique, by use of appropriate reference standards, for calibration and interpolation purposes, and ideally by adding synthetic internal standard peptides at a known concentration, enabling account to be taken of sample related molecules which interfere with mass spectrometry analysis. The reference standard peptides have an amino acid sequence which is identical to that of the target sequence and the internal standards would be analogues modified to have a characteristic molecular weight, for example by covalent modification or by the use of specific isotopes within the peptide. Suitable modifications include acetylation, amidation, anilideation, phosphorylation, or the like. Isotopic labelling can involve the use of stable isotopes such as ¹³C, ¹⁵N, ²H, ¹⁷O or ³⁴S as would be understood in the art. These reference and internal standards will allow calibration of the results so that accurate quantitation of the target peptides is possible. The technique is described for example in WO02/082919, the content of which is incorporated herein by reference. In that reference, the technique is used to distinguish between normal and disease states, but it is not used in relation to typing of individual disease strains.

However, in view of the fact that the peptides generated will have different N- or C-terminal ends, which are of known sequence, it is possible that they may be detected and quantified in step (3) using immunological methods. Thus step (3) will comprise a method in which the peptides are simply detected using immunological methods as outlined above, with or without prior isolation and/or separation of peptides by methods well known to those skilled in the art (such as liquid chromatography) .

Quantification in this case, may be slightly easier than if the immunological methods were applied initially following the nonspecific digestion as a calibration peptide in the form of the peptide from the conserved region as discussed more fully below are also available.

It is important for the accuracy of the results however, that all peptides within the sample with the common C-terminal (or N- terminal in the case of C-terminal fragments) are identified and quantified. In a fully characterised type of TSE, it should be possible to identify all the peptides which would be produced in step (2) in order to obtain a correct concentration result for use in step (4) . Provided all are identified and quantified in step (3) , an accurate quantitation of the total sum of the peptides having an N-terminus determined by the cleavage site of the non-specific protease, within the sample is obtained. However, there is always a possibility that the strain under consideration is an unusual (uncharacterised) strain of some sort, in which the cleavage sites may be different. Therefore, in order to confirm that the total concentration of peptide may be quantified to allow accurate relative proportions of the specific peptides, the amount of a peptide from a region consistently resistant to digestion by the non-specific protease (a conserved region) , is also determined in step (B) or step (3) as appropriate, and this is used to confirm the total peptide content of the sample.

The nature of the conserved regions and thus the peptides generated therefrom will be variable depending upon factors such as the type of TSE being considered as well as the nature of the non-specific protease used. However, a particularly convenient region will be that located at the C-terminus of the TSE sequence which is generally highly conserved, except in the case of some atypical strains as mentioned above.

The source of the samples which are subject to analysis using this method may be any suitable biological organism which suffers from TSE. These include humans, ruminants (such as cattle and sheep as well as goats, cervids, such as deer, or felines. The particular peptides detected in each case will be different, but can be determined using methods as outlined for example herein.

In a particular embodiment, the method is used to determine the strain of sheep TSE. In a particular embodiment, the non-specific protease used in step (1) is Proteinase K (PK) , the specific protease used in step (2) is trypsin, and the fragments identified in step (3) comprise those of SEQ ID Nos 1-9 as shown in Table 1.

Table 1

*This shows the position of the sequence within the amino acid sequence of ovine prion protein, as published at (Goldmann, W. , Hunter, N., Foster, J. D., Salbaum, J.M. , Beyreuther, K. and Hope, J. Proc. Natl. Acad. Sci. U.S.A. 87 (7), 2476-2480 (1990)) for the homozygous ARQ (136, 154, 171) genotype) , the content of which is incorporated herein by reference.

This range of sequences have been found to be useful in the determination of most classical forms of scrapie, and the relative concentrations of these provide the means for readily determining the difference between a classical scrapie and ovine BSE as illustrated hereinafter.

In particular, SEQ ID NOS 6, 7, 8 and 9 appear to be dominant in BSE strains. Therefore, if the relative amounts of these are combined, they may be used to determine the relative BSE "character" of the strain. Where the relative proportions of these three combined exceeds about 70%, the strain may be classified as a BSE strain, although any result greater than about 50% total of these three peptides may be the subject of further investigation as outlined below.

The presence of significant quantities of peptide of SEQ ID NO 8 is an indicator of BSE character, although this peptide is also quite prevalent in some unusual strains of scrapie such as CH1641. Since preliminary studies have indicated that N-terminal sequence coverage for CH1641 is incomplete, further examination of CH1641 should reveal additional peptide sequence (s) which will distinguish it from BSE.

Conversely, the dominant peptides in classical scrapie strains are SEQ ID NOS 3, 4 and 5. Thus high percentages of these in a sample will indicate that the sheep from which the sample was taken was affected by classical scrapie only.

In this embodiment however, it is still preferable to include the identification and quantification of a peptide from a conserved region in step (3) to confirm that the relative concentration of the peptides is correct. Unusual strains of scrapie may exist, such as CH1641 where further peptides may be formed when the process of steps (1) and (2) are followed. If these are not detected, the relative proportions of the remaining peptides will be misinterpreted and so an inappropriate diagnosis may be made.

In a particular embodiment, the amount of a conserved sequence selected from SEQ ID Nos 10-12, and in particular SEQ IS NO 12, which is a peptide derived from the C-terminal region, as shown in Table 2

Table 2

is also identified and quantified in step (3) and used to confirm that the concentration of the conserved peptide is equivalent to the total of all the individual N-terminal peptides measured in step (3) .

Also in this embodiment, the applicants have found that some unusual strains of scrapie may resemble ovine BSE to a greater extent than would be expected. In particular, one unusual scrapie strain which has been designated CH1641 has been found to show a significant amount of BSE character in the method of the invention.

As a result, the procedure may be expanded to detect additional peptides which are obtainable from the unusual strains, in order to confirm the result.

In order to achieve this additional characterisation, investigation of further peptides, such as those which result from the non-specific protease digestion at cleavage sites elsewhere in the sequence of the prion protein, for example upstream or downstream of the cleavage site of the specific protease may be investigated. Thus particular peptides may comprise C-terminal peptides. One classification of unusual strains are described as atypical. These have different sensitivity to PK, and this results in generation of some lower mass peptides, as well as the ^Λusual' ragged ended PrP^res (Klingeborn, et al 2006 J. General Virology 87, 1751) .

The applicants believe that peptides derived from the C-terminal end of normal scrapie when the method of the invention is applied may differ from that of atypical scrapie. Thus for example the presence of a C-terminal peptide of SEQ ID NO. 13, derived from a region downstream of the trypsin binding site, may be prevalent in normal scrapie, whereas a 7kDa peptide of SEQ ID NO 14 may be more prevalent in atypical scrapie strains. Table 2A

Depending upon the conditions used in the reaction, these peptides may be found alone or they may form part of a longer peptide. In some instances, there may be more than one such peptide, as a result for example of "ragged" cutting by the nonspecific protease in this region also, and therefore, the precise nature and number of the cleavage sites will have to be determined in order to accurately quantify the further characteristic peptide, in a similar fashion to that discussed above in relation to the specific peptides used in the initial typing process.

The distinction between different scrapie strains previously not found by low-resolution methods such as Western blotting, provided by quantitative analysis of PrP^res N-terminal peptides proves that quantitative N-terminal amino acid profiling by LC- in particular, SRM has added value. In the experiments set out below, SRM was used rather than SIM for ion detection, because of the increased specificity. Initial experiments established that detection by SRM results in higher signal-to-noise compared to detection by SIM. Typically, response factors in SIM mode were of the order of 10e6, whereas SRM mode resulted in response factors of the order of 10e4. However, due to the higher s/n, limits of detection and quantification were higher because peak integration was less accurate. [Other types of mass spectrometry instrument offering high mass resolution may enable specific and sensitive detection in SIM mode]

SRM for peptide quantification is less commonly used compared to quantification of non-peptidic small molecules, although examples have appeared in recent literature. Peptides tend to fragment into a relatively large number of different channels with similar abundance rather than a limited number of dominant fragments. The latter would be more useful for developing SRM methods. In addition, peptides isolated from protein starting material are present in a mixture in which several compounds can be expected to have similar features. For example, a tryptic peptide mixture is expected to contain mostly peptides with Lys or Arg at the C-terminus, all resulting in abundant fragment ions of m/z 148 and m/z 175, respectively, at higher collision energies. As these fragments are not very specific, they do not make a good choice for SRM method development. The tryptic peptide series representing the 'ragged' PrP^res N-terminus in our experiments, all have the same C-terminal sequence, which gives rise to several similar y-ions and internal fragments (Figure 1) . PrP^res purified with the combined NaPTA and SCC methods gave less additional signal in LC-SRM chromatograms than with NaPTA alone, but this effect is difficult to distinguish from generally lowered signal abundance. Preparations by NaPTA-only generally allow unambiguous identification of the peptides of interest, as exemplified by the chromatogram in Figure 2B. Additional peaks in the chromatograms were occasionally observed, but were minor and at different retention times than the analytes of interest. The choice of the right combination of precursor and fragment ion is of great importance for peptide quantification from a mixture. In our initial method development work, several higher abundance, lower- (~< 400 amu) m/z fragments were included in addition to the combination of precursor and multiply charged y-ions listed in Table 3 below. Although these transitions provided higher ion counts, the increase in background resulting from the apparently increased number of compounds with signal at that particular transition resulted in elevated detection limits and thus compromised rather than aided quantification. It is not advisable to use fragment ions that are common between the different precursor ions, even if the precursor ions differ in m/z value, especially as the retention times are also sufficiently close for some peaks to overlap

(Figure 2) . Fortunately, larger y-ions specific for each of the peptides are also present in the MS/MS spectra, thus proving unique precursor-fragment pairs to specifically detect each of these peptides. The specificity of detection of these peptides is also aided by the LC separation, providing retention time as an additional parameter for unique identification of a peptide. This is demonstrated by the relatively straightforward identification of the N-terminal peptides prepared from a biological sample (Figure 2B) . Only one of the selected peptides, G96-K109, proved occasional difficulties due to interfering signal from an unknown compound eluting 30 s earlier. Sufficient purity of the sample was one of the considerations in the comparison between different PrP^res isolation methods described, but no proof was found that the method that provided the highest quantity of peptides of interest introduced more than reasonable amount of interfering peaks. The reduction in overall signal upon including the SCC purification step following the NaPTA method prevented us from making this distinction. Apparently, the specificity of detection by SRM permits us to use the purification method that provides the highest analyte quantity.

The gain in resolution provided by mass spectrometry-based detection is unparalleled. At least it is sufficient to discern the exact PK cleavage site to the highest resolution required, which for this purpose is down to the individual amino acid residue .

MS and MS/MS experiments of the N-terminal tryptic peptides of PK treated PrP^Sc have allowed selection of combinations of precursor and fragment ion pairs for SRM method development that are sufficiently specific, in spite of extensive sequence similarities between these peptides. The specificity of analysis by LC-SRM was sufficient to allow unequivocal identification and quantification of these peptides from tryptic digests of PK treated PrP^Sc isolated from 0.5 g sheep brain material, upon precipitation by NaPTA. Not only have differences in N-TAAP been identified between scrapie and experimental BSE in sheep, but also not previously resolved fine detail in PK cleavage sites between natural scrapie and an experimental scrapie subtype, SSBP/1. Absolute quantification using SRM has allowed us to identify strain- dependent differences that would be more difficult to pinpoint by lower resolution or qualitative N-terminal amino acid profiling methods. These methods could be expanded to the typing of many other types of TSE as discussed above.

The invention will now be particularly described by way of example with reference to the accompanying diagrammatic drawings in which:

Figure 1 shows the MS/MS spectra of synthetic analogues of p_rp²⁷-³o ^-terminal tryptic peptides (a) sextuply protonated GQPHGGGWGQPHGGGGWGQGGSHSQWNKPSKPK ( [G77-K109+6H]⁶⁺) , collision offset 25 V (b) quintuply protonated. GQPHGGGWGQPHGGGGWGQGGSHSQWNKPSKPK ( [G77-K109+5H]⁵⁺) , collision offset 40 V (c) quintuply protonated

GGGWGQPHGGGGWGQGGSHSQWNKPSKPK ( [G81-K109+5H]⁵⁺) , collision offset 30 V (d) quadruply protonated GGSHSQWNKPSKPK ( [G96-K109 + 4H]⁴⁺), collision offset 25 V;

Figure 2 shows LC chromatograms with peptide detection by SRM of (a) 0.1 pmol/μL equimolar mixture of synthetic analogs of PrP peptides (b) PrP peptides extracted from 0.5 g sheep brain (individual, VRQ/VRQ genotype) . 5μL of a total of 35 μL of extract derived from this sample was injected on column, thus quantities correspond to approx. 15% of starting material. Peak identification: 1, G96-K109; 2, G94-K109; 3, G85-K109; 4, G89- K109; 5, G81-K109; 6, G77-K109;

Figure 3 shows Abundance of PK digestion products of PrPSc, following: (i) purification by SCC (ii) without further purification (iii) purification by NaPTA followed by SCC (iv) purification by NaPTA only. Replicates were performed with 5 mL 10% brain homogenate starting material. Resulting pellets were reduced, alkylated and methanol precipitated. (A) Western blot detection using 6H4 antibody. Lanes: M, marker; +, pooled natural scrapie; -, ovine control brain; Pellets were dissolved in 40 μL Prionics Blue buffer, of which 4μL loaded per lane (50 μug tissue equivalent) . Ov+ve lane: sheep positive control prepared in accordance with the standardized VLA hybrid method⁶. (B) LC-SRM detection of N-terminal peptides formed following tryptic digestion upon dissolving the reduced and alkylated pellet in 6 M urea, pH 8.3. Each peptide concentration is given as the average (± SD) over four parallel processing replicates each following the same procedure, and was determined from a single LC-SRM measurement for each replicate. (C) Relative N- TAAP of NaPTA-prepared and quantified pooled homogenate;

Figure 4 is a series of graphs showing the normalised abundance (%85=100xC85/ (C85+C89+C94+C96) ) of four N-terminal tryptic peptides isolated from various TSEs in sheep. Natural scrapie: n=3, ARQ/VRQ genotype, SSBP/1: n=4, VRQ/VRQ genotype, CH1641: n=5, AHQ/AHQ genotype;

Figure 5 illustrates the relative positions of the cleavage sites giving rise to many of the peptides used in the evaluation of the method of the invention;

Figure 6 contains graphs showing a comparison of results obtained with natural scrapie in various genotypes of sheep;

Figure 7 contains graphs showing a comparison of results obtained using the method of the invention with natural scrapie, an unusual scrapie field case and an experimentally induced BSE; Figure 8 is a graph illustrating the differences which are found in the BSE characteristics of the N-terminal peptides of various TSEs;

Figure 9 is a graph showing the results using a mixed infection model, where scrapie and BSE brain homogenates were combined in mixed proportions; and

Figure 10 shows the results of some reanalysis carried out to add the peptide of SEQ ID NO 8 into the range of peptides identified and quantified.

Example 1 METHODS Chemicals and synthetic peptide standards.

All chemicals and reagents were of reagent grade quality and were obtained from Sigma-Aldrich (Poole, Dorset, UK) , unless otherwise stated. Trypsin (V5111, >5000 u mg^"1) was purchased from Promega (Madison, WI, USA) .

Solvents for LCMS including H₂O were HPLC Gradient grade Chromasolve (Riedel de Hahn - Sigma-Aldrich) and formic acid (puriss. pa for mass spectroscopy) and heptafluorobutyric acid (puriss. pa for ion chromatography) were obtained from Fluka (Sigma-Aldrich) .

Prionics-check WESTERN homogenization buffer was prepared as recommended by the manufacturer (Prionics AG, Schlieren, Switzerland) .

"Iodide solution" contained potassium iodide (0.9 M), sodium thiosulphate (9mM) , sodium phosphate (15 mM, pH 8) and sarkosyl (1 % w/v) , and 20 % sucrose/ iodide solution was prepared by dissolving sucrose (1Og) in iodide solution (33 ml) and diluting to 50 ml with water. Phosphate buffered saline (0.1 M, pH 7.0) was prepared by mixing appropriate quantities of Na₂HPO₄.2H₂O, NaH₂PO₄ and NaCl in H₂O.

Synthetic analogs of sheep PrP derived peptides (Table 3) were obtained from Peptide Protein Research Ltd. (Eastleigh, Hampshire, UK) purified to 98% and used without further purification for method development and quantification. Each peptide was obtained as its TFA salt and dissolved in H₂O to a known concentration. Thus obtained stock solutions were diluted further into equimolar mixtures of 50 pmol/μL of each peptide. These mixtures formed the bases of the dilution series used as external calibration standards.

Table 3: N-terminal PrP^Sc27-30 tryptic peptides and ions used in detection with mass spectrometry.

*Average m/z, as observed with triple quadrupole instrument Brain samples

Brain homogenates were derived from field scrapie cases (pooled scrapie brain) or from sheep inoculated via the intracerebral route with brain homogenate from well characterised serially passaged scrapie isolates (CH1641 and SSBP/1) or from cattle with BSE. The genotypes of these sheep, all of the Cheviot breed, with respect to their PrP alleles was AHQ/AHQ (CH1641) , VRQ/VRQ (SSBP/1) and ARQ/ARQ (BSE) . Homogenate was prepared from a VLA maintained scrapie-free NZ sheep flock for use as negative controls. Brain samples were prepared as 10% weight/volume homogenates in Prionics homogenization buffer by ultrasonication for 60 seconds.

Extraction of PrP^Sc -derived peptides

Homogenates were centrifuged at 2000 rpm for 2 minutes to remove cellular debris and subsequently divided into 3-5 itiL aliquots. These aliquots were processed by precipitation with sodium phosphotungstic acid (NaPTA) as described by Wadsworth {Wadsworth, Joiner, et al. 2001 The Lancet 2001; 358:171} and/or sucrose cushion centrifugation method (SCC) as described by Hope et al. {Hope J, Multhaup G, et al. 1988 Eur. J. Biochem. 172:271}, to extract the PK-treated PrP^Sc. The combined method has been described in detail by Howells et al . {Howells L J. Gen. Virol, submitted}. Briefly, lU/μL benzonase in 0.02 M MgCl₂ was added to each aliquot at 50 μL/mL homogenate and incubated for 30 minutes at 50⁰C. This was followed by a 60 minute incubation with Proteinase K (>30 units/mg, P2308) at a concentration of 100 μg/mL homogenate. Subsequently the protease was deactivated by heating in a boiling water bath for 15 min. Note that an enzyme blocker could not be used as samples are subject to a second enzymatic proteolysis later in the procedure.

For protein precipitation by NaPTA, a 4% phosphotungstic acid solution (80 μL/mL homogenate, in 170 mM magnesium chloride) was added, followed by incubation at 37⁰C for 30 min with continuous agitation. All samples were then centrifuged for 1 h at 22,000 g, after which supernatants were discarded and the pellets resuspended in 0.4 ml sarkosyl (0.1 % w/v in phosphate buffered saline, pH 7.4) and EDTA (0.1 ml, 250 mM) . The suspensions were re-centrifuged at 22,000 g for 30 min and the supernatants discarded. For additional p_rp^27"30 purification by SCC, resulting pellets were resuspended in water (1.25ml) and iodide solution (2.5ml), carefully transferred onto a cushion of 20% sucrose/ iodide solution (1.25ml) in a 5ml tube and centrifuged at 287,000xg for 1.5h at 10⁰C (Beckman L8-60 centrifuge with swing out head) . The supernatants were discarded and the pellets, essentially PrP^27"30, washed by resuspending in 0.5ml water and centrifuging at 13,000rpm for 0.5h. The pellets were stored at - 80⁰C until further processing.

Reduction and alkylation were performed as follows: the prP^res pellets were solubilised in guanidine hydrochloride (50 μl, 6M in 50 mM Tris, pH 8.0), reduced with 2 mM dithiothreitol (5 μl) at 95 ⁰C for 20 min. and alkylated with 4-vinylpyridine (6mM, 5 μl) at room temperature for 1.5 h. Insoluble material was discarded following centrifugation (RCF is 8000 g, for 2min, bench top microfuge) and the protein isolated from the supernatant by precipitation with cold methanol (0.25 ml at -20 ⁰C) , maintaining at -20 ⁰C overnight before centrifugation at 10,000 g for 10 min at -9 °C. The supernatant was discarded and the pellet re-suspended in cold methanol (-20 ⁰C) , centrifuged at 10,000 g for 2 min at -9 ⁰C, and after discarding the supernatant, the pellet was allowed to dry in air at room temperature for 2-4 h.

For tryptic digestion the dried PrP^res pellet was suspended in 10 μL freshly prepared urea (6M) , upon which 10 μL Tris/methylamine solution (150 mM Tris buffer pH 8.0 containing 60 mM methylamine and 15 mM calcium acetate) and the synthetic trypsin substrate boc-val-leu-lys-7-amido-4-methylcoumarin (bocVLK-AMC, 0.4 ng in 2 μl) , which functions as a quality control of tryptic digestion, were added. Trypsin was dissolved in H₂O to 20 ng/mL and a 2 μL aliquot added immediately. Samples were incubated at 30 ⁰C for 18 h and digestion was terminated addition of 12 μL 5% formic acid. Thus, for each 0.5 g brain starting material, 36 μL PK-treated and trypsin digested PrP^Sc sample was obtained. Digests were stored at -20 °C awaiting analysis by LCMS/MS.

Western blotting

Pellets obtained from 0.5 g tissue and processed up to and including reduction and alkylation were homogenized in 40 μL

Prionics blue buffer and incubated at 100⁰C for 10 minutes. From this mixture, 4μL was loaded in each lane of a precast NuPAGE 12% Bis-Tris 17-well gel (Invitrogen, Paisley, Renfrewshire, UK) . Positive controls were prepared by the VLA hybrid method, which included a PK digestion step. {Stack, Chaplin, et al. 2002 supra.} Immunodetection was performed using the 6H4 antibody to PrP as described elsewhere{Stack, Chaplin, et al. 2002 supra.}.

Liquid Chromatography and Mass Spectrometry To separate the diagnostic N-terminal tryptic peptides of PK- treated PrP^Sc, an Ultimate 3000 liquid chromatography system (Dionex, Camberly, Surrey, UK) was used. Chromatographic peaks were analysed with an API2000 triple quadrupole mass spectrometer (Applied Biosystems, Warrington, Cheshire, UK) .

The synthetic peptide analogs {Table 3} were used to optimize and develop the LC and MS/MS (SRM) methods and served as calibration standards for quantification. As these peptides are relatively basic without hydrophobic amino acid residues, they are not very effectively retained on a standard reversed-phase

C18 LC column. This is reflected by their MEEK index {Meek Proc. Natl. Acad. Sci USA 1980, 77, 1632}. A nanoLC strategy was developed by Howells et al. {Howells L supra.} to specifically retain and separate these hydrophilic compounds. This method was transferred to the capillary LC set-up used in the current work. The LC sequence adopted is as follows: the trapping cartridge (small molecule cap-trap, 004/25108/31, Michrom Bioresources Inc. CA, USA) is prewashed separately from the analytical column with 100% aqueous phase containing 0.1% heptafluorobutyric acid (HFB) prior to injection of the samples (5 μL, "microlitre pick- up" injection mode) onto this column. Urea and other residual polar reagents and buffers are washed from the pre-column by a 0.02 % HFB in 0.1 % formic acid running for 10 min at 20 μL/min. A 5 μL/min gradient running from 0% to 40% solvent B (0.1% FA in acetonitrile) over 18 minutes was used to elute and separate over a C18 analytical column. Capillary columns, either a Pepmap C18 (150x0.3 mm) (Dionex, Camberly, Surrey ) or an ACE-AQ (150x0.3 mm, 100 A pore size, 5 μm particle size) (Hichrom Ltd, Reading, Berkshire) were used. The column was connected to the 10-port after a Krudkatcher in-line filter (0.5 μm pore size) (Phenomenex, Macclesfield, Cheshire) . Both precolumn and analytical column were subsequently washed by 80% B for 5 min at 5 μL/min. The analytical column was re-equilibrated in 100% A (consisting of 98% H₂O, 2% acetonitrile and 0.1% FA when a Pepmap column was used, 100% H₂O with 0.1% FA in case of an ACE- AQ column) for 20 minutes prior to injection of the next sample. Both the trapping cartridge and the analytical column were inside a column oven maintained at 30⁰C.

Peptides were eluted from the analytical column into the API2000 triple quadrupole mass spectrometer (Applied Biosystems, fitted with a Turbolon source. Curtain gas pressure was set to 15 psi, GSlto 20 psi and GS2 off. Source temperature was set to ambient while the interface heater was on. The electrospray voltage was set to 5kV. Ql and Q3 operated with unit resolution. Data were acquired in SRM mode and single transition was selected for each analyte (Table 3) . A dwell time of 100 ms was used for each transition with a pause of 5 ms. Data analysis and quantification

Peptide MS/MS spectrum interpretation was aided by use of the freely available fragment ion calculation tool "MS product" {Baker P & Klauser K }.

LC-MS/MS data were acquired, and calibration lines and analyte concentrations calculated based on peak area, using Analyst software version 1.4. Analytes were quantified by external calibration based on standard runs spanning the concentration range of 2 fmol/μL to 1 pmol/μL (5 μL injections) . Linear- through-zero or quadratic regression was used to fit the calibration standard data, depending on which gave the better regression coefficient. Analyte concentrations were calculated from duplicate dilution series prepared from independently prepared 50 pmol/μL stock mixtures. All analyte and calibration standard peaks were manually verified and re-integrated as necessary. Results MS and MS/MS of PrP^res N-terminal peptides

Electrospray mass spectra of the synthetic analogues of the selected PrP^res N-terminal tryptic peptides (Table 3) displayed a multiply protonated ion series for each peptide (data not shown) . For example, charge states between 3+ and 7+ were observed for G77-K109. As each peptide contained the highly basic sequence KPSKPK (SEQ ID NO 15) at the C-terminus in addition to one or more histidines towards the N-terminus, this degree of protonation is easily explained. For each of the peptides, the P_maχ-1 protonation state (where P_max is the maximum number of protons that a peptide can accommodate in the gas phase, assumed here to equal the number of basic amino acid residues plus the N-terminal amine) gave the highest peak in the electrospray mass spectrum

Fragment spectra were acquired of the two most abundant charge states, P_max-1 and P_max-2, for each peptide at various collision offsets. As a rule, the P_max-1 charge states readily fragmented into one or more singly charged b-ions and their multiply charged y-ion counterpart. Peptides in the P_max charge state fragmented similarly. The P_max-2 charge states required somewhat higher collision energies for the same degree of fragmentation, and gave spectra containing a larger proportion of fragments of relatively low m/z, mainly C-terminal ions, immonium ions and internal fragments. This is illustrated by the fragment spectra of [ (G77-K109)+6H]⁶⁺ and [ (G77-K109) +5H] ⁵⁺ (Figure IA and IB). These results are in keeping with the ^Λmobile proton model' {Dongre, Jones, et al . 1996 J. of the American Chemical Society, 1996, 118, 8365}, which states that sufficient proton density on the peptide backbone is required to promote charge- directed fragmentation. It is inferred thus that, in the P_max-2 charge state, proton density is sequestered to such an extent by the basic C-terminus and the histidine residues, that the probability of charge-directed fragmentation of more N-terminal amide bonds is reduced.

For both the [ (G77-K109) +6H] ⁶⁺ (Figure IA) and [ (G85-K109) +5H] ⁵⁺ ions (data not shown) , the b₂ ⁺/y₃i⁵⁺ and b₂ ⁺/y₂₃ ⁴⁺ fragment ion pairs, respectively, dominated the spectra at collision energies corresponding to the onset of fragmentation. This effect can be attributed to the presence of a proline residue in position three. Amide bond dissociation is known to be enhanced directly N-terminal to Pro. For [ (G81-K109) +5H] ⁵⁺ however, b/y ion formation did not appear to be restricted to dissociation N- terminal to the proline residue in position seven (Figure 1C) . A (a₄ ⁺+ b₄ ⁺) /V₂₅ ⁴⁺ (m/z 330.1, m/z 358.2 and m/z 640.2) ion pair was observed in higher abundance than b₆ ⁺/y₂₃ ⁴⁺ (m/z 543.2 and 593.5) at any collision offset. In addition, various other b⁺/y⁴⁺ combinations were identified. The greater distance of the proline residue from the N-terminus in G81-K109 presumably results in reduced probability of the N-terminal proton to direct the formation of a b₆ ⁺/y₂₃ ⁴⁺ pair. Interestingly, the MS/MS spectrum of [G96-K109 +4H]⁴⁺ displayed a peak (assigned to the y₁₃ ³⁺ ion, m/z 494.6) in addition to and higher than the Yi₂ ³⁺ ion at m/z 475.3 (Figure ID). In keeping with b-ion formation mechanisms, a b/ ion was not observed however. [G96-K109 +4H]⁴⁺ formed an additional fragment ion at m/z 556.5 (Figure ID). This peak was of similar intensity as that assigned to Yi₃ ³⁺, but was not used as a transition for SRM, as ions of m/z 556.5 appeared in the MS/MS spectra of all of the peptides studied, apart from that corresponding to the m/z of the P_max-1 charge state of [ (G77-K109)+6H]⁶⁺ (Figure IA). Given the mass accuracy limitations of the triple quadrupole mass spectrometer, up to 10 different fragments could be assigned to m/z 556.5. Tentatively assignment of m/z 556.5 to y_g ²⁺ or yis³⁺ ions seems appropriate since the peak appears as part of multiply charged y-series.

The multiply charged y-ion fragments of the P_max -1 charge states were thus selected for SRM detection used in the following sections, since these were the most specific and abundant products .

SRM chromatograms of standards, samples and controls

The SRM ion chromatograms showed relatively symmetrical peaks for both an equimolar mixture of the synthetic analogues (Figure 2A) and the tryptic peptides isolated from pooled sheep brain tissue (Figure 2B) . Peaks were not fully resolved in the time domain, but the selectivity of SRM detection allowed integration of each component without cross talk.

Each of the tryptic peptides for which a synthetic analogue was produced and an SRM detection method developed, was observed to some extent upon LC-SRM analysis of digested PrP^res extracted from scrapie-infected sheep brain. Negative control samples simultaneously prepared from brain tissue derived from a scrapie-free sheep did not show any peaks with a signal-to-noise ratio higher than 3, that could be assigned to N-term of PrP^res. Comparison of the chromatograms in Figure 2 reveals that that the sites in PrP^Sc most frequently cleaved by PK are N-terminal to Gly85, Gly89 and Gly94, whereas sites N-terminal to Gly77, Gly81, and Gly96 are present with a relatively low abundance. Quantification of the PrP^res N-terminal tryptic peptides

The LC-SRM methods were used to detect and quantify each of the six N-terminal peptides by reference to external calibration standards. For each peptide somewhat different limits of detection (LODs) , defined as the lowest concentration of analyte that generated a minimum signal-to-noise ratio of 3, were found, which are listed in Table 4.

Table 4

Table 4 : Summary of quantification parameters as applied to differentiate between naturally transmitted scrapie, SSBP/1 scrapie, CH1641 scrapie and BSE.

For most peptides except G77-K109 and G81-K109, LODs were achieved of 10-20 fmol injected on-column. The LODs for G77-K109 and G81-K109 were somewhat higher and more variable between different batches of analyses. The LODs being higher for these peptides was attributed to the fact that their signal tended to be distributed over a larger number of charge states, resulting in a lower response factor for the selected transition. The batch-to-batch variability of the LODs of G77-K109 and G81-K109 was ascribed to the relatively wide and occasionally tailing chromatographic peaks of these two peptides. This could possibly be caused by effects due to their larger size combined with polarity and basicity, resulting in ionic interaction with column contaminants or carrier material under slightly sub- optimal chromatographic conditions . LODs and peak shapes significantly improved when ACE-AQ columns were used instead of Pepmap columns. As a precaution to errors arising from variable quantification parameters, LODs were estimated for each batch individually based on the performance of the calibration standards, which was then taken into account in the evaluation of the results.

Calibration curves for the peptides G85-K109, G89-K109, G94-K109 and G96-K109 all showed excellent linearity with R values better than 0.99. Occasionally, calibration curves for G77-109 and G81- 109 provided less linearity but R values better than 0.99 were obtained when quadratic regression was used instead.

Limits of quantification (LOQs) and accuracies were established using dilutions between 2 fmol/μL and 1.0 pmol/μL of synthetic peptide mixtures in the same buffer as the isolated PrP^Sc after tryptic digestion. LOQs were defined as the concentration above which the accuracy of the calculated value was within 20% of the theoretical value (n=3) . Different LOQs were obtained for the different peptides. The best values were obtained for G94-K109 and G96-K109, whereas LOQs for G77-K109 and G81-K109 were relatively poor. The data in Table 2 are given as obtained using the Pepmap column, which corresponds to the conditions used to analyse samples prepared from ovine brain material and presented below, but improved LOQs were obtained when the ACE-AQ column was used. Accuracies obtained between the LOD and the LOQ were typically within 25-50% of the theoretical value, allowing confirmation of the presence of a peptide and put a lower limit on its concentration, but with the appropriate caution with respect to the accuracy of the calculated value.

Quantitative N-teπninal Amino Acid Profiling of PrP^res

To assess the quantitative capabilities of prP^res profiling by LC-SRM based N-TAAP and to simultaneously evaluate whether it was possible to reduce sample preparation time and possibly increase the yield of the N-terminal peptides, different combinations of previously used PrP^res isolation methods were tested and compared. Preparation methods evaluated were: precipitation by sodium phosphotungstic acid (NaPTA) , centrifugation through a sucrose cushion (SCC) , these two methods combined{Howells L supra.} or neither method applied. The sample preparation methods have been compared for both pooled (Figure 3) and individual (data not shown) samples from sheep naturally infected with classical scrapie. With detection by Western blotting the gold standard TSE differential screening method, the preparations were analysed by both Western blotting and LC-SRM. Using pooled brain homogenate allowed us to compare the preparation methods in replicate. Six 4 mL aliquots were subjected simultaneously to each of the preparation methods up to precipitation in ice-cold methanol, thus including reduction and alkylation. Two of the six resulting pellets were subsequently solubilised in SDS-buffer and analysed by WB (Figure 3A) , whereas the four remaining pellets were solubilised in buffer containing 6 M urea, subjected to tryptic digestion and quantification by LC-SRM (Figure 3B) .

Figure 3 clearly shows that NaPTA on its own produced the most intense signals. In contrast, no significant signal was obtained from samples prepared by just the SCC method. Combination of the NaPTA and SCC methods produced less signal compared to NAPTA alone, though more signal than just the SCC method. Preparation of PrP^res by centrifugation only yielded better signal compared to additional application of the SCC method, but not as much as when a NaPTA step was included. Therefore the NaPTA method without any additional purification was used in all further experiments .

The data in Figure 3 also show that the efficiency of the extraction methods can be variable even when the aliquots have been processed in parallel. This is most dramatically demonstrated by the differing intensities in the NaPTA only

Western blot. We have verified that variations originating from the process of PrP^res isolation is the main contributor to the standard deviations shown Figure 3B. For comparison, the coefficient of variation of the peptide concentration between the processing replicates with the NaPTA only method ranges from 31% to 56%, whereas the variation in accuracy of quality control standards of the different peptides is between 7% and 17% (100 fmol/μL) . Hence a better way to represent the N-TAAP is when variation in absolute concentration of PrP^res isolation is corrected by presenting the abundance of each of the peptides as a fraction of the sum of their concentration, as illustrated by the relative profile given in Figure 3C.

The Western blot in Figure 3A shows that the isolation methods used in conjunction with LC-SRM analysis do not introduce changes that are detectable by WB. There is no change in migration pattern of the NaPTA/SCC lanes compared to the lane from PrP^res prepared by the VLA hybrid method (lane 14, Ov+ve) . For the NaPTA and SCC methods, stronger bands are observed at molecular weight above 30 kDa, known to occur upon precipitation of PrP^Sc in organic solvents. No N-TAAP differences have been detected either by LC-SRM within the limits of experimental variation, as shown in Figure 3B.

All preparations apart from the hybrid method included a reduction and alkylation step. Although reduction and alkylation of PrP^Sc is known to prohibit detection by the 6H4 antibody, the tissue equivalent for these lanes was considerably higher compared to the Ov+ve lane (100 mg compared to 0.5 mg, respectively) .

Differentiating ovine TSEs by N-Terminal Amino Acid Profiling The N-terminal amino acid profiles of PrP^Sc27-30 extracted from sheep brain tissue were compared. N-TAAPs were determined of naturally transmitted scrapie and of experimental transmissions of BSE, CH1641 and SSBP/1. SSBP/1 is an experimentally defined scrapie isolate. CH1641 is a scrapie subtype of which the prP^res displays a banding pattern similar to BSE upon detection by Western blotting, but can be differentiated from BSE by transmission to mice and by immunohistochemistry. Our data showed that the absolute abundance of a given peptide for a given TSE varies considerably for the individual animals. For example, the measured concentration of G94-K109 varied between 10 and 44 fmol/μL in samples prepared from BSE infected tissue (n=4),and between 3 and 18 fmol/μL for CH1641 (n=5) . This variation presumably corresponds to variations of the total PrP^Sc present in the individual tissue samples, and to variations in processing between samples as observed for pooled brain homogenate (Figure 3B and C) . Hence taking the variability of the total amount of prP^res isolated out of the equation is necessary when comparing N-TAAPs between strains, and we determined the average and standard deviation of the percentage abundance of each peptide for each of the strains as shown in Figure 4. The peptides G77-K109 and G81-K109 were found below the limit of quantification in all but the naturally transmitted scrapie samples, where their abundance was relatively low, similar to the data from pooled samples in Figure 3. Therefore, these peptides were excluded from the data analysis.

Figure 4 shows clear differences in N-TAAP corresponding to various TSEs. Comparing the data for natural scrapie and SSBP/1 on one hand, and CH1641 and BSE on the other, there are marked differences in the relative abundance of G89-K109 and of G85- K109, being lower for CH1641 and BSE, and G96-K109, being higher for CHl641 and BSE. These findings are in agreement with the molecular weight trends observed with Western blotting, where the bands from BSE and CH1641 are shifted to a lower molecular weight compared to classical scrapie and SSBP/1.

Comparing the data from classical scrapie with SSBP/1 scrapie in Figure 4, we see that the ratio of the abundance of G89-K109 to G94-K109 is significantly higher for SSBP/1 scrapie. Western blot analysis shows no differences in molecular weight of prP^res for classical scrapie compared to SSBP/1, although distinction based on glycoform ratios is possible. Although no clear distinction could be made between CHl641 and BSE based on the current set of peptides selected detected to determine the N- TAAPs.

Example 2

Using the methodology described in Example 1 above, further investigations using peptides of SEQ ID NO 1-9 and 12, as well as were carried out on a range of sample from sheep suffering from various forms of scrapie, including some unusual strains and experimentally induced BSE.

The results are illustrated graphically in Figures 5-10.

Figure 5 illustrates the relative positions of the main PK cleavage sites within ovine prion protein for natural scrapie. Mean profiles for ARQ/VRQ (14 animals), VRQ/VRQ (8 animals) and ARQ/ARH (4 animals) scrapie N-TAAPs are shown in Figure 6. Whilst the first two are broadly similar, the ARQ/ARH derived samples gave a consistent pattern in which the peptide representing PK cleavage at G89 was most abundant. However, it is clear that the three peptides, those of SEQ ID NOS 3, 4 and 5 are clearly the most prevalent in scrapie type. Therefore, the sum of these, as compared to the total amount of peptide in the sample gives an indicator of the "scrapie" character of a particular strain.

[By also measuring the amount of the C-terminal peptide of SEQ ID NO 12, it is possible to estimate the recovery/coverage of the N-terminal peptides, by adding up the concentration of all the N-terminal peptides and dividing this by the concentration of the C-terminal peptide. This should ideally be 1, assuming the N-term peptides detected represent all cleavage points and that that C-terminal peptide is indeed conserved. If it is not, that indicates that a peptide may have been missed in the analysis. This methodology allowed the identification of the peptide of SEQ ID NO 8, which is now used in analysis] Similar results for a BSE strain as compared to a natural scrapie strain and an unusual strain PG1697/00) are shown in

Figure 7. It is clear, that in the case of BSE, peptides of SEQ ID NOS 6. 7 and 9 are dominant, and, in total represent 71 % of N-terminal peptides, wheras for scrapie it is only 20 %. The unusual strain, showed an increased level of these peptides at 58%, but this is still significantly below the BSE strain levels .

Furthermore, the results of these tests (Figure 7) , and in particular considering the content for the peptide of SEQ ID NO 12, indicated that for BSE, despite giving 71 % BSE trait, only a modest proportion (26%) of the N-terminal peptides were recovered (later studies indicated that peptide W102-K109 may be responsible for the balance; Figure 10) , whereas for scrapie we have 20% BSE trait but recover 91% of N peptides.

Figure 8 provides a general indicator for the extent to which the proportion of BSE characteristic differs between TSEs. The y axis gives the % BSE trait for individual sheep and the x axis indicates the TSE type. On average the relative abundance of BSE characteristic peptides is a lot lower for classical scrapie compared to BSE, as expected and averaged 14 % compared with 71% for BSE 68 % for CH1641, 58% for the atypical scrapie and 37 % for SSBP/1. Some data points where the fraction seems higher actually have a rather low total yield, explaining the result. This is especially noticable for SSBP/1 data and some CH1641 data.

The scatter is associated with greater uncertainty associated with particular samples containing peptides at low abundance. On average, it is clear that the relative abundance of BSE characteristic peptides is a lot lower for classical scrapie compared to BSE.

In order to further investigate the diagnostic potential of the technique, scrapie and BSE brain homogenate was mixed in various proportions before PK digestion and processing using the method outlined in Example 1. The results of this work are shown in the graph on Figure 9. On the y axis are the % TSE marker peptides and on the x axis the of BSE in the mix up to 100% BSE or 100 % scrapie. It is clear that, from about 50 % BSE, distinctions can be made between the two TSE types. This suggests that the method of the invention could detect BSE, even in the case of a mixed infection.

Once the discovery of the omission of the peptide of SEQ ID NO 8 was discovered as mentioned above, experiments were carried out to determine the contribution it makes to the BSE N-TAAP. A few selected samples were reanalyzed, and the amended data is shown in Figure 10. Clearly the peptide of SEQ ID NO 8 represents an important cleavage site, and it represents a further indicator for both BSE and CH1641 strains.

Claims

Claims

1. A method for typing a strain of transmissible spongiform encephalopathy in an animal, said method comprising (A) obtaining a sample of truncated prion protein that is resistant to digestion by protease and that is characteristic of transmissible spongiform encephalopathy;

(B) determining the identity and amounts of protein types by establishing the extent and site(s) of truncation; (C) determining the relative proportions of said identified and quantified protein types in relation to the sum of their concentration within the sample; and

(D) comparing said relative proportions of said protein types to those found in specific types of TSE and relating the result to the specific TSE type found.

2. A method according to claim 1 wherein step (A) comprises subjecting a sample containing prion protein obtained from an animal having or suspected of having a TSE and digesting this with a protease to remove normal protein to leave PrP -,^rres

3. A method according to claim 2 wherein the protease is a non-specific protease such as Proteinase K.

4. A method according to any one of the preceding claims wherein in step (B) , the sample is subjected to a further digestion under denaturating conditions to obtain a range of specific peptides, derived from the N- or C-terminus of the truncated prion protein, and detecting the quantifying peptides of similar sequence.

5. A method according to claim 4 wherein all peptides within the sample are detected and quantified.

6. A method according to claim 4 or claim 5 wherein the peptides are detected and quantified on the basis of their mass/charge ratio.

7. A method according to claim 6 wherein the peptides are introduced into a mass spectrometer using a liquid spray such as an electrospray for detection and quantification, or detection and quantification is carried out on a dried sample using a laser ablation or desorption mass spectrometry technique.

8. A method according to claim 7 wherein the mass spectrometry uses MALDI technology.

9. A method according to any one of claims 6 to 8 wherein the peptides are subject to a preliminary chromatographic separation.

10. A method according to any one of claims 4 to 9 wherein the further digestion is carried out using a specific protease.

11. A method according to claim 10 wherein the specific protease is trypsin.

12. A method according to any one of the preceding claims wherein the sample is homogenised brain tissue.

13. A method according to any one of the preceding claims wherein the amount of a peptide from a region consistently resistant to digestion by the non-specific protease, is also determined and this is used to confirm the total peptide content of the sample and so the extent of the N- or C-terminal peptides detected.

14. A method according to claim 13 wherein the said region which is consistently resistant to digestion by the non-specific peptide is a C-terminal region.

15. A method according to any one of the preceding claims which is used to determine the strain of sheep TSE.

16. A method according to claim 1, said method comprising (1) digesting a sample containing prion protein with a nonspecific protease so as to remove normal protein and leave protease resistant fragments of abnormal prion protein (PrP^res) ;

(2) further digesting the PrP^res, for example under denaturing conditions with a specific protease to obtain a range of specific peptides, derived from the N- or C-terminus of PrP^res;

(3) detecting the said specific peptides for example on the basis of the mass/charge ratio thereof and determining the identity and amounts of all said specific peptides within the sample; (4) determining the relative proportions of said identified and quantified specific peptides in relation to the sum of their concentrations within the sample

(5) comparing said relative proportions to those proportions found in specific types of TSE and relating the result to the specific type found.

17. A method according to claim 16 which is used to determine the strain of sheep TSE, wherein the non-specific protease is PK, the specific protease is trypsin, and the peptides identified comprise those of SEQ ID Nos 1-9.

18. A method according to claim 17 wherein the amount of a conserved sequence selected from SEQ ID Nos 10-12, is also measured in step (3) and used to confirm that the total concentration of the selected conserved peptide is equivalent to the total of all the individual N-terminal peptides measured.

19. A method according to claim 17 or 18 wherein the relative amounts of SEQ ID NOS 6, 7, 8 and 9, either alone or in combination, is used to determine the relative BSE character of the strain.

20. A method according to any one of the preceding claims wherein a further peptide, which results from the non-specific protease digestion at a cleavage site upstream or downstream of the cleavage site of the specific protease is identified and quantified to further characterise the strain type.

21. A method according to claim 20 wherein the method is a method according to any one of claims 17 to 20, and wherein said further peptide comprises SEQ ID NO 13 or SEQ ID NO 14.