Patent Application
20120009595
The present invention relates to the use of protease resistant binding agents, typically polyionic materials such as polyanionic materials including pentosan polysulphate, dextran sulphate or other polyanionic polyglycosides or polycationic materials including polybrene, polyamidoamine dendrimer or poly(diallyldimethylammonium chloride), under selective conditions to capture aggregates of protein molecules in the presence of non-aggregated molecules of said protein. Optionally, the extent of aggregation of the protein is determined.
Prion diseases, also referred to as transmissible spongiform encephalopathies or TSEs, have been recognised for some time. Bovine spongiform encephalopathy (BSE) was first reported in 1985. The first cases of variant Creutzfeldt Jakob disease (vCJD) were reported in 1996. vCJD is a fatal neurodegenerative disease in humans believed to be caused by the consumption of BSE contaminated meat. The incubation time between infection to clinical symptoms in the human may be many years.
The only identified component of the prion, the agent causing prion diseases, is PrPSc, an abnormal isoform of PrPC (PrPSc is also referred to as PrPres and PrPC also referred to as PrPSen). PrPSc has previously been regarded as being distinguished from PrPC in that it is comparatively protease resistant. Recently however, it has been published that there is a protease sensitive form of PrPSc, i.e. that there is an infective form of PrP that is protease sensitive.
It may be that the infective but protease sensitive PrPSc is able to aggregate (i.e. is aggregating in nature) but not yet aggregated or at least only partially aggregated.
Both protease insensitive and protease sensitive forms of PrPSc and core portions of PrPSc left after partial protease digestion (often referred to in the art as PrP27-30) are referred to herein as PrPSc except where the context indicates that a specific one of these is meant. Also, the term ‘aggregating proteins’ is used to include both aggregated protease resistant PrPSc and similar forms of other proteins as well as infective non-aggregated or partially aggregated forms of PrPSc or other proteins, which may include the newly observed protease sensitive infective PrPSc.
PrPC is a GPI anchored glycoprotein of unknown function. Although some other markers for prion diseases have been suggested PrPSc remains not only an obligatory prion component, but also the only reliable and universally accepted marker for this family of diseases.
The once favoured methodology for assaying for the presence of PrPSc is to subject a sample to proteolysis with Proteinase K for a period sufficient to destroy PrPC and then to determine the presence of surviving PrPSc by an immunoassay using an antibody which is not selective for PrPSc in the presence of PrPC (Serban et al., Neurology, Vol. 40, No. 1, January 1990). The use of the protease naturally excludes the presence of an antibody as a capture agent or as a detection agent during the proteolysis step. The protease must be removed or deactivated before the antibody can be introduced. It would be desirable to avoid this limitation on the procedure.
The assay depends on the complete removal of PrPC to avoid false positives and upon the conditions not being such as also to degrade PrPSc to avoid false negatives. Such conditions of selective proteolysis need to be developed for each type of sample to be assayed. The sensitivity of the resulting assay is limited. For instance, in assays of bovine brain tissue, the sensitivity may only be such that a reliable positive result is obtainable at about the time that the animal would have been likely to show clinically observable symptoms of BSE. Thus, the assay has a sensitivity limit in the region of 1 μg/ml, corresponding to 104-105 prion infectious units.
There was a need for more sensitive and specific diagnostic tests for prion diseases. In particular, an ante-mortem test using blood or other sample types was required to assess the disease status of a particular animal. In the absence of such a method extensive slaughtering of cattle is required once an affected animal is identified within a herd. It was again vital that a diagnostic test be developed to screen the human population and to protect individuals from potential infection from donated blood, surgical procedures and organ and tissue transplants.
U.S. Pat. No. 5,977,324 and U.S. Pat. No. 6,221,614 both describe methods of binding PrPSc using phosphotungstic acid (PTA). PTA is a non-specific protein precipitant that will also bind to, and precipitate, a wide range of proteins other than PrPSc. The concentration of the proteins in the sample will also greatly affect the recovery using PTA.
Plasminogen has been reported to bind PrPSc selectively with respect to PrPC and was proposed for use in diagnostic assays (Fischer et al. Nature, 2000, Nov. 23, 408 (6811): 479-83). However, this method has not proved sufficiently useful in practice. Plasminogen is also identified in a related disclosure, US 2002/0004586, as being a factor which selectively binds PrPSc.
According to U.S. Pat. No. 6,419,916 and related disclosures, the polyamine compound Superfect™ (a branched polyamine mixture produced by heat induced degradation of a PAMAM dendrimer) and other similar branched polyamines are capable of clearing PrPSc from cells in vitro. The mechanism is unclear. It is speculated that such compounds may bind directly to PrPSc arranged as an amyloid with exposed negatively charged moieties and induce a conformational change under acidic conditions. It is said that the effect cannot simply involve binding of PrPC and inhibiting synthesis of PrPSc because existing PrPSc is cleared. The polyamine is found to make PrPSc protease sensitive provided the pH is below 4. It is deduced that the polyamines act in an acidic cell compartment in the in vitro PrPSc clearance experiments.
It would appear from this work that it would be speculative to conclude that such polyamines bind PrPSc. A number of other possibilities are advanced. No selectivity for the binding of PrPSc over PrPC is shown or suggested. Furthermore, it cannot be deduced that any binding that occurs is more than transitory, just serving to alter the conformation of PrPSc so as to allow protease attack. Also, the action of the polyamines appears to require a low pH. Our own investigations in fact show that such dendrimer polyamines do not bind PrPSc at such low pH.
Pentosan polysulphate (poly-b-xylose-2,3-disulphonate, PPS) is one of a range of large polysulphonated polyglyco-sides (PGs) (MW 8,000-12,000). Made from beechwood, it is an inexpensive compound that has been used for many years as an anticoagulant similar to heparin, also a PG. PGs including PPS and other polyanions are known to bind both PrPC and recombinant PrP (recPrP), see for instance Brimacombe D B et al, Biochem J, 1999 Sep. 15; 342 pt 3, 605-13. PPS has accordingly been proposed as a potential therapeutic agent for preventing or treating TSE diseases. It has not however been shown to remove existing PrPSc in vivo or in vitro.
In manufacture, sawdust from beechwood is extracted to produce the soluble sugar polymer of xylose (a five member ring sugar) called pentosan. This polymer is then subjected to a sulphation reaction using a mixture of chlorosulphonic acid and pyridine, which results in 3 out of 4 of all the sugar ring hydroxyls having a sulphate ester added to them. The total sulphate content is then about 50-55% by weight which is more than in heparin, in which it is about 30-35%. The only other similar molecule that approaches this high degree of sulphation is dextran sulphate (40-45%). Pentosan has quite a low MW of 3.5-7.0 K.
No selectivity for binding by polyanions or polycations of PrPSc with respect to binding of PrPC had been reported. As described in U.S. Pat. No. 7,659,076 and in U.S. Ser. No. 12/644,283, of which this application is a continuation in part, we established conditions under which polyionic materials bind aggregated altered proteins like PrPSc and further established conditions under which such polyanions bind these abnormal forms but do not bind their non-aggregated normal forms like PrPC, the binding being sufficiently strong and under preferred conditions sufficiently selective to be useful in assays for the presence of the aggregated altered protein (e.g. PrPSc).
The market for therapeutic proteins is becoming increasingly large. It is vital for such products that they can be produced to a particular quality, and that that quality can be guaranteed to the end user. The quality of the protein products is affected by aggregation of the proteins, which may occur at any phase of the production process, or after production. It would therefore be useful to be able to monitor the state of aggregation of therapeutic proteins during the protein production process, in order for example to identify steps in the process that lead to aggregation of the product, and to be able to quantify the level of aggregation in the final product so that the quality of the final product may be indicated, along with appropriate storage conditions and shelf life.
The products of aggregation can range from small aggregates consisting of small numbers of protein molecules (oligomers) to much larger aggregates consisting of hundreds or more molecules. Large aggregates may be monitored by several techniques, such as dynamic light scattering (DLS), multi-angle light laser scattering (MALLS), UV spectroscopy, light obscuration, micro-flow imaging (MFI) or nanoscale tracking analysis (NTA). However, the small aggregate oligomers and the larger sub-visible particulates in the 0.1-10 μm range are much more difficult to measure (see Satish Singh et al. “An Industry Perspective on the Monitoring of Subvisible Particles as a Quality Attribute for Protein Therapeutics” J. Pharm. Sciences 2010, 99, 8, 3302-3321). It is advantageous to be able to detect and/or to quantify these smaller aggregates as their presence may be early indicators of more extensive future aggregation in a therapeutic protein product, and may exacerbate aggregation by “seeding” the formation of larger aggregates.
Accordingly, there is now provided in a first aspect of the invention, a process for the selective binding of aggregated protein molecules, preferably molecules of a therapeutic protein, in the presence of non-aggregated molecules of the protein, comprising contacting under selective binding conditions a solution containing both said aggregated and non-aggregated protein molecules with a polyionic material having a binding avidity for said aggregated protein molecules as present in the sample. The binding conditions may include the presence of a competition agent in solution, which competition agent has a lesser binding avidity for the aggregated protein molecules than does the polyionic material.
The polyionic material may be polyanionic or polycationic in character and many different organic polyionic materials are described hereafter for use in the invention. In this context, proteins are not themselves considered to be polyionic materials, so the polyionic materials are non-proteinaceous.
The polyionic material may be in solution or may provide a surface presenting ionic surface groups. In the latter case, the surface may be that of a polymer having said ionic groups covalently bonded within the structure of the polymer or produced by modification of surface groups of the polymer. An example of a suitable polyanionic polymer is Nafion, a perfluoronated sulphonated hydrocarbon polymer available as beads or as sheets. Polycationic polymers may also be used.
Alternatively, the surface is that of a substrate having coated thereon or bonded thereto a substance presenting said ionic groups. An example of a suitable polymer having such surface groups is a non-charged plastics surface activated with maleic anhydride and derivatised with TRIS to produce surface carboxyl groups or with a polycationic material. Polycations or polyanions may instead be passively coated on polymers such as polystyrene.
In the case of a polyanionic material, whether used in solution or coated on a solid surface, the polyanionic material may preferably be a polyanionic polyglycoside.
Generally, the competition agent has a lesser density of ionic groups than the polyionic material. Without being bound by theory, it is likely that the findings described in detail herein are due to aggregated molecules of the protein having more binding sites for interaction with ionic groups than the non-aggregated molecules of the same protein. A competition agent having one or a few ionic groups is able to interact with a certain avidity with either the aggregated or non-aggregated forms of the protein but a polyionic material is able to bind the aggregated form of the protein simultaneously through many ionic groups, leading to it having a higher avidity for the aggregated than for the non-aggregated form.
Our experimental results with infected bovine brains indicate that both immobilised polyanions (such as dextran sulphate) and polycations (such as polyethyleneimine) are able to capture the abnormal form of the prion protein PrPSC in brain homogenates. The signal obtained using an anti-prion protein antibody/enzyme conjugate is approximately 3 to 5 times higher for the best polycationic capture surface than for the best polyanionic capture surface. Further, our experimental results also show that immobilised organic polyanions (such as dextran sulphate) or organic polycations (such as p-DADMAC) are able to capture aggregated forms of human recombinant IgG and of recombinant human serum albumin in the presence of the non-aggregated form. The signal obtained using an appropriate antibody conjugate shows that these immobilised polyanions are capable of detecting small aggregates or oligomers consisting of a limited number of molecules. The heat-induced aggregation of monoclonal IgG can also be monitored using this method.
In both cases the detergent Sarkosyl (N-lauroyl-sarcosine) can act as a competition agent helpful for improving the specificity of capture of the aggregated protein and avoiding a signal from the non-aggregated protein, when using a non-specific protein antibody (i.e. one which is reactive with both the aggregated and the non-aggregated protein).
Also, when analysing for PrPSc, partially digesting the sample with trypsin substantially increases the signal from PrPSc when using either polycationic or polyanionic compounds, but has no effect on specificity (indicating that, under the conditions employed, trypsin is removing an inhibitor of polyion binding of PrPSc, rather than preferentially digesting PrPC, as has been observed for proteinase K).
In addition to the ionic interactions proposed, there may be additional hydrophobic binding between other regions of the aggregate and the polymers employed. These will strengthen further the binding interactions.
“Avidity” here is used in the usual meaning of the overall binding strength of a molecule with many binding sites with a multivalent binding agent and in contrast to “affinity”, being the binding strength between each individual binding site and of the molecule and the binding agent.
The competition agent if used is preferably an amino acid amide of a fatty acid, such as n-lauroylsarcosine. Such materials have detergency properties, but in this context may well simply be acting as monovalent binding agents via their terminal COO− group or as partially polyvalent agents through the formation of micelles.
In a further aspect, the present invention provides a process for the selective binding of an aggregated form of a, preferably therapeutic, protein in the presence of the non-aggregated form of the protein, comprising contacting a solution containing both said aggregated and non-aggregated forms with a polyanionic polyglycoside under conditions such as to provide selective binding of said aggregated form.
In preferred embodiments of each aspect of the invention said aggregated form of a protein is an immunoglobulin or an albumin, such as IgG or a serum albumin e.g. human serum albumin. However, the invention in all its forms is broadly applicable to the selective binding of aggregated forms of proteins.
Organic polycationic selective binding agents that can be used include polyethyleneimines, polyamines, including poly-lysines, polyamidoamines, e.g. PAMAM dendrimers, poly-quaternary amines such as poly(diallyldimethylammonium chloride) and 1,5-dimethyl-1,5-diazaundecamethylene poly-methobromide (also known as hexadimethrine bromide or Polybrene).
The preferred polyanionic polyglycoside is a polysulphonated polyglycoside. However, other anionic sites such as carboxylic acid groups or phosphate groups may be used as well or instead.
Preferably, the polysulphonated polyglycoside is pentosan polysulphate (PPS) or dextran sulphate.
Other polyanionic pentosan or dextran derivatives may be used as the polyanionic polyglycoside.
A high level of sulphonation (or other anionic group) is preferred.
The levels of sulphonation of the carrageenins, dextrans and pentosan are high. If a low proportion of the potential sulphonation sites is actually taken up by sulphate groups then it may be found that the compounds do not interact with the binding sites in the PrPSc selectively.
Suitable anionic organic selective binding agents may include:
Pentosan polysulphate (MW 3500-5000), Dextran sulphate 500 (MW 500,000), Iota-carrageenan, Lambda-carrageenin and carrageenans, e.g. Kappa-carrageenan, Heparins and heparans, Dextran sulphate 8 (MW 8,000), sulphonated polyglycosides such as fucoidan, keratin sulphate, hyaluronic acid polysulphate, colominic acid (bacterial polysialic acid), carrageenan types iii and iv, dermatan sulphate, heparan sulphate, furcellaran, sulphated commercially available polysaccharides e.g. polysorbate, sizofuran, xanthan gum, starch, cellulose compounds, pectin, gastric mucin, ceratonia, agars, acacia gum, Sulphated Glycoside 1, Sulphated Glycoside 2, N-acetyl-D-glucosamines, or Dermatan sulphate L-iduronic acid.
The polyionic material may be one selected to have the ability under non-selective conditions to bind both aggregated forms of a protein and also the non-aggregated form of the protein as well as the ability to bind the aggregated form selectively under appropriate conditions.
The desired selectivity is obtainable by suitable adjustment of the reaction conditions, particularly the presence and concentration of the competition agent, the pH and the detergency. Preferably therefore the pH is so selected as to provide said selective binding.
The pH is preferably from 5.6 to 9, e.g. from 7 to 9, more preferably 8 to 9, e.g. from 8.2 to 8.6, especially 8.4, particularly when the detergents described below are used. Suitable buffers include phosphate buffers and Tris buffers.
The salt concentration is preferably not higher than 250 mM and is preferably significantly less, e.g. not above 100 mM.
Preferably, a detergent is present which promotes said selective binding whether by virtue of detergency or by acting as a competition agent.
Particularly preferred for this purpose are detergents which are an amino acid amide of a fatty acid, e.g. n-lauroylsarcosine or other fatty acid sarcosines. The presence of such a detergent/competition agent is especially, preferred when the selective binding agent is polyanionic.
Preferably the concentration of this detergent is at least 0.05% by weight, more preferably at least 0.1%, preferably at least 0.2%, e.g. 0.2 to 2%, more preferably 0.5 to 1.5%, but greater amounts may be used.
Other detergents having a similar effect may be used, including CHAPS, Brij, Octyl-β-glycoside, Tween 20, Triton X-100 and Nonidet P-40. The use of high concentrations of sodium dodecylsulphate (SDS) is however undesirable. Combinations of n-lauroyl sarcosine (or similar) with other detergents are suitable, preferably containing 0.5 to 2%, e.g. about 1% sarcosine detergent, e.g. with 0.5 to 2%, e.g. about 1%, of one of the detergents listed above, particularly Triton X-100 or Nonidet P-40.
In some circumstances we have found that the presence of trypsin, chymotrypsin, proteinase K, or another such protease can be helpful to prevent inhibition by unknown materials of the binding of aggregating protein to either polyanionic or polycationic selective binding agents. This is especially the case where the sample contains a relatively high level of other proteins, such as is the case if a PrPSc positive brain sample is diluted with a PrPSc negative brain material. Additional matrix inhibition prevention can be obtained by including other enzymes of a degradative nature including DNase and collagenase.
The selective binding agent after binding to said aggregated form of the protein may be captured with an immobilised capture agent and the presence or amount of a complex formed between said selective binding agent and said capture agent may be determined.
Said capture agent may be a lectin (where the binding agent is suitable, e.g. is a polyglycoside) or an antibody reactive with said selective binding agent. Said selective binding agent may be provided with a selectively bindable tag moiety and said capture agent may then bind to said tag moiety.
Optionally and alternatively, the selective binding agent is immobilised to a solid medium prior to exposure to said sample. The selective binding agent may be provided with a selectively bindable tag moiety and may be immobilised to said solid medium via said tag.
Where a bindable tag moiety is present it may for instance be biotin, fluorescein, dinitrophenol, digoxygenin, or (His)6.
The selective binding agent may be immobilised directly to a solid rather than through a bindable tag. For instance PG's may be directly coupled by covalent coupling through remaining hydroxyl groups of the PG using solid phases derivatised with for instance epoxy or vinylsulphone groups.
In each aspect of the invention, whether the binding of the aggregated protein takes place before or after the immobilisation or capture of the selective binding agent, the immobilised selective binding agent/aggregated protein complexes are preferably subjected to a washing step to remove non-aggregated protein to improve selectivity. The washing step is preferably conducted using a solution containing a said competition agent, which may be a detergent solution, which preferably again comprises a detergent that whether by virtue of its detergency or otherwise promotes selective binding. This is preferably an amino amide of a fatty acid, e.g. n-lauroyl sarcosine or another fatty acid sarcosine. Preferably, the concentration of the sarcosine detergent in the washing step is at least 0.05%, preferably at least 0.1%, more preferably at least 0.2%, e.g. 0.2 to 2%, preferably 0.5 to 1.5%. Other detergents may also be present and the wash is preferably buffered to a pH in the range of 5.6 to 8.4.
Said binding of aggregated protein may be qualitatively or quantitatively determined by conducting an immunoassay for the aggregated protein after separation of bound aggregated protein from non-aggregated protein.
Also, once the aggregated form of the protein has been selectively bound and optionally after the non-aggregated form has been removed, further polyanionic material, e.g. anionic polyglycoside (suitably labelled with a tag or detectable label) may be bound to the already bound aggregated protein to form a sandwich (e.g. polyglycoside-aggregated protein-polyglycoside label) which may then be quantitated or detected. Selective binding conditions may not be necessary when carrying out the second part of sandwich formation.
As mentioned above, the selective binding agent may be immobilised to a solid material either before or after being contacted with the aggregated protein. Separation of the sample from the solid material may then be used to remove the non-aggregated form of the protein from the assay leaving only the aggregated form for further determination.
In this context, solid support materials include not only macroscopic or handlable materials such as microtitre plates, dipsticks and laminar flow devices, but also microbeads such as superparamagnetic microbeads, which may be separated off by filtration or by magnetic capture. Biotin or other tags may be conjugated to dextran sulphate or PPS and like materials by standard chemical methods. About one in ten of the sugar backbone residues in PPS is a uronic acid methyl ester and this provides one route for coupling via their carboxyl residues. Other known routes for coupling are hydroxyl (one in four is still free after the sulphation reaction), or end group reducing sugar. Biotin is a convenient bindable tag moiety to employ for binding of the polyanionic material or other selective binding agent to a solid material derivatised with avidin or a material with avidin binding properties such as steptavidin, Neutravidin or Captavidin™.
Other molecules suitable for use as bindable tag moieties will include all those which are readily conjugated to the polyionic material and which lend themselves to capture by a suitable capture agent. For instance, a molecule such as fluorescein may be conjugated to PPS or like molecules by reacting an amino fluorescein derivative with the uronic and side chains of pentosan polysulphate in the presence of carbodiimide EDC(1-ethyl-3-(3-dimethyl-aminopropyl)-carbodiimide) and may be captured by a suitable readily available antibody, which may itself be immobilised to the solid material. Other tags suitable for antibody capture in this way include dinitrophenol DNP, digoxygenin, nucleic acid or nucleic acid analog sequences, and (His)6. Binding agents other than antibodies may also be used, e.g. complementary nucleic acid or nucleic acid analog sequences.
Alternatively however, a capture agent may be used which selectively binds the polyionic material itself rather than through a tag moiety. For instance, polyglycosides may be bound by a suitable lectin or by a suitable antibody. Antibodies for binding PPS are for instance disclosed in Kongtawelert et al; J. Immunol. Methods 1990, Jan. 24; 126(1); 39-49. Standard techniques for immobilising such antibodies are well known in the art.
Any known or in future devised method for determining the presence or amount of aggregated proteins such as PrPSc, aggregated IgG or aggregated human serum albumin (without needing selectivity to exclude the non-aggregated form such as PrPC, non-aggregated IgG or non-aggregated human serum albumin) can be used to determine the presence or amount of the aggregated form once it has been selectively bound by the selective binding agent and unbound non-aggregated form protein has been separated therefrom, suitably by immobilisation of the bound and washing away of residual unbound. Such methods include the known ELISA, RIA, IRMA and other forms of immunoassay, for instance the method embodied in the Bio-Rad Platelia™ BSE Detection Kit and described in Serban et al.
Depending on the form of the assay used, it may be desired or required to elute the captured aggregated form protein from the selective binding agent prior to the assay. In conducting such an elution step, the presence of a chaotrope such as guanidine thiocyanate may be desirable at a concentration of at least 1M, preferably 2 to 6 M, e.g. 4 to 6M. Alternative chaotropes may be used including urea.
Additionally or alternatively, a competition agent having a still higher avidity may be used to displace the protein from the selective binding agent. Sodium dodecyl sulphate (SDS) is suitable for this and is preferably used at a concentration of 0.5 to 1% by weight, preferably above 0.75%.
Other proteins that may be selectively bound and determined according to the invention include the β-amyloid protein and tau protein which form plaques in Alzheimer's disease. Still others include a peptide hormone, a growth hormone, erythropoietins, or interferons.
Without wishing to be bound by the following theory, it is thought that PPS and similar molecules function in the invention by binding pairs of negative sulphate groups to pairs of positive amino acids (Lys and Arg) in the relevant proteins or via the proteins' polyhistidine metal binding sites. Binding to the aggregated forms may be stronger due to the increased number of binding sites presented by the aggregating protein. Suitable anionic detergents may compete more effectively for binding with the non-aggregating form to enhance selectivity. Suitably, the selectivity obtained is such that the avidity for binding to the aggregating protein is at least three times that for the non-aggregated form, preferably at least 10:1.
In a further aspect, the invention includes a process for separating aggregated IgG from non-aggregated IgG comprising selectively binding aggregated IgG to a binding agent in the presence of an amino acid amide of a fatty acid. Preferred conditions for such binding are as set out in detail above and the bound protein may be assayed as described.
The invention will be further described and illustrated by the following examples making reference to the accompanying drawing in which:
Biotin was conjugated to pentosan polysulphate using standard chemical methods. The biotinylated pentosan polysulphate was allowed to bind to the rogue prion protein in brain homogenates and after binding the pentosan polysulphate/prion complexes were captured using streptavidin-derivatised superparamagnetic beads. The captured rogue prion was subsequently eluted from the beads and detected using the immuno-based Bio-Rad Platelia™ BSE Detection Kit; the latter kit is unable to differentiate the normal and rogue prion protein and will give a signal with both proteins. A bank of two BSE-infected and two uninfected bovine brains were investigated and used to demonstrate that the pentosan polysulphate, under the specific conditions described, could be used to specifically capture rogue prion protein from the brain homogenates.
After performing the immunodetection in the microtiter plate-based Platelia™ assay the signal in each well was measured at a wavelength of 450 nm using an ELISA reader.
The signal from the two BSE-infected brain homogenates containing rogue prion is significantly higher than in the uninfected normal brain homogenates.
The Bio-Rad Platelia™ BSE Detection kit cannot differentiate between normal or rogue prion protein. Normally, the specificity for rogue prion protein is achieved by prior digestion of the sample with proteinase K which removes the protease susceptible normal prion protein. Any rogue prion protein in the sample is more resistant to protease digestion and remains and is subsequently detected by the Platelia™ assay. In this experiment we have demonstrated an alternative approach to protease digestion of the sample. We have used defined conditions under which biotinylated pentosan polysulphate in solution can specifically bind to the rogue prion protein in the sample. The rogue prion/pentosan polysulphate complex can then be captured using streptavidin superparamagnetic beads. After washing, the rogue prion protein can subsequently be eluted and detected in the immunoassay. Normal prion protein is not captured by this protocol and is washed away and is therefore not detected in the immunoassay. We have demonstrated that by using this technique we could correctly detect rogue prion protein in two BSE-infected bovine brains and no signal was observed in two normal bovine brains.
Biotin was conjugated to pentosan polysulphate using standard chemical methods. The biotinylated pentosan polysulphate was used to coat streptavidin-derivatised super-paramagnetic beads. The coated beads were then used to specifically capture the rogue prion protein from brain homogenates. The captured rogue prion protein was subsequently eluted from the beads and detected using the immuno-based Bio-Rad Platelia™ BSE Detection Kit; the latter kit is unable to differentiate the normal and rogue prion protein and will give a signal with both proteins. A bank of three BSE-infected and three uninfected bovine brains were investigated and used to demonstrate that the pentosan polysulphate, under the specific conditions described, could specifically capture rogue prion protein from the brain homogenates.
After performing the immunodetection in the microtiter plate-based Platelia™ assay the signal in each well was measured at a wavelength of 450 nm using an ELISA reader.
The signals from the three BSE-infected brain homogenates containing rogue prion protein is significantly higher than in the uninfected normal brain homogenates.
The Bio-Rad Platelia™ BSE Detection kit cannot differentiate between normal or rogue prion protein. Normally, the specificity for rogue prion protein is achieved by prior digestion of the sample with proteinase K which removes the protease susceptible normal prion protein. Any rogue prion protein in the sample is more resistant to protease digestion and remains and is subsequently detected by the Platelia™ assay. In this experiment we have demonstrated an alternative approach to protease digestion of the sample. We have used defined conditions under which pentosan polysulphate can specifically capture the rogue prion protein from the sample. This captured rogue prion protein is eluted and detected in the immunoassay. Normal prion protein is not captured by the pentosan polysulphate and is washed away and is therefore not detected in the immunoassay. We have demonstrated that by using this technique we could correctly detect rogue prion protein in three BSE-infected bovine brains and no signal was observed in three normal bovine brains.
Approximately one in ten of the sugar residues in the poly-xylose backbone of pentosan sulphate is substituted with a uronic acid residue, this in turn is substituted with a methyl ester on some of the carboxyl groups, thus a number of free carboxyl groups exist in the molecule and can be derivatised with carbodiimide to form active esters. These in turn may be substituted with amino species to generate an amide bond. In this particular case, EDC and NHS are chosen to form the active ester and biotin hydrazide is chosen as the amino species. Two reactions were performed, a one step reaction in which biotin hydrazide is present initially and no NHS is added, and a second reaction in which NHS/EDC is allowed to react simultaneously with PS and biotin hydrazide.
Pentosan sulphate (Norton Healthcare) was a gift from Stephen Dealler
Biotin hydrazide 100 mg, Pierce#21339 mw 258.33 batch AH41461
EDC [1-ethyl-3-(3-dimethylaminopropyl)carbodiimide methiodide] Sigma #16, 534-4, 1 g,
NHS [N-Hydroxysuccinimide] Sigma #H7377 5 g mw 115.1
Dialysis tubing mwco 3.5 k Pierce # 68035
DMSO Sigma
The two reactions were conducted using the following protocols in two versions, with and without NHS:
Dissolve 100 mg of biotin hydrazide in 6 ml of DMSO in a glass vial, this may require warming and/or ultrasonication. The final concentration is thus 16.7 mg/ml or 65 mM. Take 1,000 mg of pentosan sulphate and dissolve in 10 ml of a 50/50 mixture of DMSO and water; this can be done in a plastic universal container. Dissolve 100 mg of EDC in 1 ml of DMSO in a glass vial, it may need warming. Dissolve NHS (approx 40-50 mg) in 1.0 ml of water.
The reaction is performed in conical bottom polystyrene universal containers, with a small circular magnetic stirrer bar (approx 10 mm dia) on a magnetic stirrer base and fitted with a combination pH electrode of 12 mm dia (or less).
Place 5.0 ml of pentosan sulphate solution in the reaction vessel, add 1.0 ml of biotin hydrazide solution, stir well and record the pH. A value of 7-8 can be expected. Add 0.2 ml of EDC solution and whilst continuously stirring, record the pH and add 10 μL aliquots of 1 N HCl from a glass micro-syringe and needle, recording the pH after every addition. Continue additions of acid until the pH is in the range 5-6. This is necessary as the reaction generates OH ions. The reaction should remain clear and colourless throughout. If any white precipitate of biotin hydrazide is formed, then the concentration of DMSO should be increased, the target value is >/=50%. Leave the reaction for 2-3 hours at room temperature (or overnight if this is more convenient).
Record the final pH of the reaction mixture. Add an equal volume of 1M NaCl to dilute the DMSO down to 25% and displace ionically bound hydrazide and transfer the entire contents to a 35 cm length of 2.2 cm dia dialysis tubing. Note the DMSO concentration is reduced to 25% to avoid damage to the dialysis tubing, the tubing should also be tested with water prior to use to detect any pinholes and should be only 1/3 full to allow for swelling on dialysis. Dialyse overnight against 2 L of water and repeat this several times, the more dialysis the better as pentosan sulphate tends to strongly retain basic ions by non-covalent ionic inter-action by virtue of its strong negative charge. Freeze dry the dialysed solution and record the dry weight. The final product should be a firm white cake. Yields can vary a lot, but 50-60% is typical, most of the loss occurs on dialysis, due to MW heterogeneity of the pentosan sulphate and loss of species with a MW of less than 3,500.
This reaction is carried out essentially as above except that 1.0 ml (44 mg) of NHS is added to the reaction vial prior to the addition of the EDC reagent which starts the reaction. The initial pH may be in the range of 6-7 and should be adjusted down with 1 N HCl to approx pH 5-6.
After calculating the recovery from the dry weight, make up a solution of 10 mg/ml in water and scan the spectrum from 200 to 400 nm. Peaks should be seen at 260 and 280 nm, though one or both may be unresolved shoulders. This adsorption is due to pyridine residues incorporated into the molecule during the sulphation step. They can be used to monitor the concentration of pentosan sulphate, eg during chromatography. Pentosan can be monitored by UV absorption at 260 nm, or at lower concentrations by the Toluidine Blue metachromasia assay.
After performing the immunodetection in the microwell plate-based Platelia™ assay the signal in each well was measured at a wavelength of 450 nm using an ELISA reader. The rogue prion protein could be readily detected in the spiked serum sample that had not been treated with pentosan polysulphate. In contrast the pentosan polysulphate-treated sample gave no signal in the test demonstrating that there was no detectable rouge prion protein remaining in this sample.
This experiment demonstrates that pentosan polysulphate can be used to effectively remove rogue prion protein from samples of interest.
Biotin was conjugated to pentosan polysulphate using standard chemical methods. The biotinylated pentosan polysulphate was used to coat streptavidin-derivatised superparamagnetic beads. The coated beads were then used to establish conditions of detergent under which the pentosan polysulphate could bind the rogue prion protein but not the normal cellular prion protein.
Preparation of Pentosan Polysulphate Coated Magnetic Beads.
After performing the immunodetection in the microtiter plate-based Platelia™ assay the signal in each well was measured at a wavelength of 450 nm using an ELISA reader.
At all concentrations of N-lauroylsarcosine there was a discrimination between BSE-infected and normal brain. 0.2% N-lauroylsarcosine was the best concentration of detergent and allowed the pentosan polysulphate to bind to and capture the rogue prion protein without binding or capture of the normal prion protein. In the absence of N-lauroylsarcosine, even though SDS detergent was present, there was no discrimination of pentosan polysulphate binding to rogue prion and normal prion protein. Under these conditions the pentosan polysulphate bound both the normal and rogue prion protein.
Biotin was conjugated to pentosan polysulphate using standard chemical methods. The biotinylated pentosan polysulphate was used to coat streptavidin-derivatised super-paramagnetic beads. The coated beads were then used to establish conditions of pH under which the pentosan polysulphate could bind the rogue prion protein but not the normal cellular prion protein.
(Sigma-Aldrich Company Ltd., S-2415) were washed by magnetic capture in three consecutive 1 ml volumes of TBS (50 mM Tris, 150 mM NaCl, pH 7.5).
After performing the immunodetection in the microtiter plate-based Platelia™ assay the signal in each well was measured at a wavelength of 450 nm using an ELISA reader.
At a pH of 7.5 and lower the pentosan polysulphate-coated beads could bind both normal and rogue prion protein. At pHs of 9.6 and higher the pentosan polysulphate-coated beads could not bind both forms of the prion protein. At pH 8.4 the pentosan polysulphate-coated beads captured the rogue prion protein but did not capture the normal prion protein. At this pH the pentosan polysulphate shows specificity for binding to the rogue prion protein.
The specificity of binding under the test conditions of pentosan polysulphate to rogue prion protein is dependent upon the pH. At pH 8.4 pentosan polysulphate binds rogue prion protein but cannot bind the normal prion protein. At pHs of 7.5 and lower both normal and rogue prion are bound whereas at pHs of 9.6 and higher there is no binding of rogue or normal prion protein. Therefore, for specific binding of pentosan polysulphate to rogue prion protein under these conditions a pH close to 8.4 should be used.
PrP can be bound to immobilised polyanions. In the absence of competing polyanions in the capture buffer both PrPres and PrPc are captured. Specificity for capture of PrPres can be achieved by including in the capture buffer a polyanion of lower charge density than that of the capture polyanion. In this example dextran sulphate is used as the high charge density capture polyanion and N-lauroyl sarcosine (which forms multi-molecular detergent micelles) and pentosan polysulphate or fucoidan are used as the weaker charge density competing polyanions.
In the absence of competing polyanion in the capture buffer the overall signal is lower and there is no difference in signal from infected or normal brain i.e. there is no specific capture of PrPres. The signal from infected brain, however, is increased by including a competing polyanion in the capture buffer and the signal from the corresponding normal or uninfected brain is suppressed. In this example, the best differentiation between infected and normal brain is achieved by the use of 1% (w/v) N-lauroyl sarcosine in the capture buffer. In addition, a differentiation between infected and normal brain can be achieved with fucoidan or pentosan polysulphate. With pentosan polysulphate the differentiation can be increased by increasing the concentration of the competing polyanion, pentosan polysulphate, in the capture buffer from 0.1 to 1 mg/ml. As a control, if dextran sulphate is included in the capture buffer the signal, as expected, is reduced to background as it competes for and inhibits the binding of the PrP to the immobilised dextran sulphate.
In this experiment it was demonstrated that PrPres could be specifically captured to a polyanionic surface. In this instance, the surface was provided by derivatised maleic anhydride polystyrene. Uncharged polysorp and maxisorp wells were used as controls. In other experiments it has been demonstrated that these uncharged surfaces can be derivatised with polyanionic dextran sulphate and can then bind PrPres.
The anionic polystyrene surface, under the conditions used in this experiment, specifically captured PrPres. Uncharged plastic did not have this effect unless it had been coated with a polyanionic ligand.
Positive Sample: A 25% suspension of brain homogenate known to be positive for PrPSc
Negative Sample: A 25% suspension of brain homogenate known to be negative for PrPSc
Maxisorb plates were coated according to the following coating protocol. 1 mg of Polybrene was coated onto the plates in carbonate buffer at pH 7.4 and left overnight, washed 3 times with PBS. The plates were then coated with 1 mg of dextran sulphate in PBS. After 6 hours, the plates were washed 3 times with PBS, then blocked with 5% BSA by adding 400 μl of 5% BSA solution and leaving for 30 minutes. Plates were then washed 3 times with PBS and allowed to dry.
Preparation of Sample Dilution in Negative Brain
Preparation of Sample Dilution in Water
40 μl of sample was mixed with 60 μl of H2O and 25 μl of capture buffer, 250 mM Tris pH 8.4, 5% BSA, 5% Triton X-100, 5% sarkosyl, 1.25 mg/ml trypsin.
Assays were Performed According to the Following Assay Protocol:
The results obtained were as follows:
Dilution of 10 mg of brain
homogenate in −ve brain
Dilution of 10 mg of brain
homogenate in H2O
These results are presented graphically in
We have shown that, under defined conditions, various selective capture agents are specific for the capture of aggregated pathogenic prion protein such that normal unaggregated prion is not captured. The aggregated prion protein has an extensive beta-pleated sheet structure whereas normal prion is mostly alpha helix in structure. This example demonstrates that other aggregated beta-pleated sheet proteins such as tau aggregates that are found in Alzheimer's Disease can similarly be selectively captured.
Results with the Anti-Tau Antibody
It is known that the brains from most aged individuals contain aggregated tau but in Alzheimer's Disease there are more of these aggregates than in age matched controls. Here, the selective capture agent is capturing these aggregates. In this example, trypsin digestion decreases the binding of the protein and reduces the signal but, under these conditions, does not reduce it to back-ground. The ratio of signal after treatment with trypsin to the signal without treatment was much higher in the Alzheimer's brains than in the controls. This suggests that there is more protease resistant aggregates of tau protein in Alzheimer's brain compared to the age matched controls.
1 mg of Hexadimethrine bromide (Polybrene) (100 μl of 10 mg/ml in carbonate buffer pH 7.4) was coated onto Maxisorb plates and left over night at RT°.
Each plate was then washed 3 times with PBS and coated with 1 mg of Dextran Sulphate (MW 500000) (10 mg/ml stock in Tris buffer pH 8.6) and left at RT° for 4 hrs.
The plates were then washed 3 times with PBS and then blocked with 300 μl of 5% BSA solution in TBS and left at RT° for 30 minutes.
The plates were then washed 3 times with PBS.
250 mM Tris buffer at pH 8.4 containing 5% BSA, 5% Sarkosyl, 5% Triton
Weakly and strongly positive brains BI63 and SV10 (25% homogenate) were treated as follows to provide samples for assay.
25 μl of brain homogenate +25 μl of capture buffer, 250 mM Tris pH 8.4, 5% BSA, 5% Triton X-100, 5% sarkosyl, +65 μl of H2O.
To this sample 10 μl of various concentrations of Trypsin was added.
50 mM Tris pH8.4+1% sarkosyl
5 mg of Positive
brain Bi63
5 mg of Positive
brain SV10
The presence of Trypsin appears to have increased the signal. It also appears that a broad concentration range of Trypsin can be used without a detrimental effect on assay.
This example demonstrates the use of various poly cations for specific capture of PrPSc. The ligands were either passively coated onto polystyrene microplates or actively coated (i.e. bound), where appropriate, to maleic anhydride plates.
All the selective binding agents were immobilised overnight at 16-25° C. in 50 mM carbonate buffer pH 9.6 at a concentration of 10 μg/ml. After immobilization, the wells were washed ×3 with PBS and then blocked with 5% (w/v) BSA in PBS for 30 mins. After blocking, wells were washed ×2 with PBS before use. The PAMA dendrimer starbust, poly L-lysine and polyethyleneimine were coated onto both Maxisorp and maleic anhydride microplates whereas the polybreen and pDADMAC were coated onto the Maxisorp plates only.
The pDADMAC and PAMA dendrimer starburst poly cations work well as PrPSc-specific ligands when passively coated to polystyrene microplates. The pDADMAC works best in this series of binding agents. Polyethyleneimine works to some degree when immobilised on maleic anhydride microplates through its amino groups.
This experiment demonstrates that a variety of poly cations can be used to specifically capture PrPSc from infected brain under the given Capture Buffer conditions used. These agents can be passively or actively immobilised to polystyrene surfaces. Other experiments have demonstrated that maximum signal from 20 mg of positive brain can be achieved in the presence of 1% (w/v) N-lauroyl sarcosine in the Capture Buffer; without N-lauroyl sarcosine the signal is reduced. This illustrates that the capture agents perform best under defined buffer conditions.
pDADMAC, under defined conditions, has been shown to be specific for the capture of aggregated pathogenic prion protein; normal unaggregated prion is not captured. The aggregated prion protein has an extensive beta-pleated sheet structure whereas normal prion is mostly alpha helix in structure. It is postulated that the binding agent may recognize other aggregated beta-pleated sheet proteins such as beta-amyloid and tau aggregates that are found in Alzheimer's disease. The experiments below were performed in order to investigate this hypothesis.
The polycationic binding agent enables capture of the tau aggregates. When the captured tau is detected with the anti-tau antibody, the Alzheimer's disease brains all gave a high positive signal whereas the negative control brains gave a low negative signal. In conclusion, capture with a polycation under the specified conditions can enable differentiation of Alzheimer's disease brains from those brains without the disease.
The effect of different proteases on the effectiveness of capture of PrPSc to polycation-coated plates (formed by coating the wells with poly (diallyldimethyl ammonium chloride) (pDADMAC), (Aldrich Chemical Company Inc., catalogue number 40, 903-0) were investigated.
It has been demonstrated that the polycationic ligand under certain conditions is specific for binding to PrPSc. However, the signal can be reduced by matrix effects derived from constituents of the brain homogenate that can interfere with binding and reduce the signal. This matrix effect can be reduced and the signal from infected brain increased by the use of proteases. This study shows that trypsin, chymotrypsin and proteinase K are effective at removing the matrix inhibition; pronase (at the concentrations investigated) is less effective. Trypsin at a concentration of 6.25-1.15 mg/ml is equally effective whereas chymotrypsin is more effective as the concentration is increased. DNase has a demonstrable but smaller effect on removal of matrix inhibition.
The effects of pH and salt concentration on the effectiveness of capture of PrPSc to polycation-coated plates (formed by coating the wells with poly (diallyldimethyl ammonium chloride) (pDADMAC), (Aldrich Chemical Company Inc., catalogue number 40, 903-0) were investigated
As the pH of the Capture buffer is lowered the signal from the uninfected brain increases but the signal from the infected brain decreases. At pHs of greater than 8.0 the optimum positive to negative signal ratio is achieved.
As the salt concentration in the Capture buffer is increased the signal from the infected brain progressively decreases. This indicates that a low salt concentration or no salt is the optimum condition for the PrPSc capture.
The effect of different N-lauroyl sarcosine concentrations in the presence or absence of trypsin were investigated on the effectiveness of capture of PrPSc to polycation-coated plates (formed by coating the wells with poly (diallyldimethyl ammonium chloride) (pDADMAC), (Aldrich Chemical Company Inc., catalogue number 40, 903-0) were investigated.
In the absence of N-lauroyl sarcosine there is no signal from the infected brain with low concentrations of trypsin. At higher concentrations of trypsin, however, some signal is restored in the absence of N-lauroyl sarcosine.
Human recombinant monoclonal IgG1 (supplied by CalbioReagents, San Mateo, USA, purified by Protein A) at 10 mg/ml was aggregated by incubation at 50° C. and at 60° C. At various time points aliquots of the protein were diluted in an excess of capture buffer: 50 mM Tris pH 8.4, 1% (v/v) Triton X-100, 1% (w/v) N-lauroyl sarcosine to stop the aggregation process. Upon completion of the time course, 2 μg of the IgG from each time point was tested in triplicate in an immunoassay (hereafter called ‘the Seprion assay’) by capture in ligand-coated microplate wells for 60 min. The wells were coated by addition of 100 μl of a solution of 5% pDADMAC to each microplate well, incubated overnight at 22° C., washed 3× with Tris buffer pH 7.5, then 100 μl of 5% dextran sulphate in Tris buffer pH 7.5 was added, incubated 3 hours at 22° C., and washed 3× with Tris buffer pH 7.5. After washing, the captured aggregates were detected using an anti-IgG (H+L) HRP conjugate and TMB (3,3□,5,5□-Tetramethylbenzidine) substrate
The results of the aggregation at 50° C. are shown in
In order to demonstrate that the Seprion assay is detecting aggregates, human monoclonal IgG was aggregated at 60° C. for 60 min and then filtered through a 300 kDa nominal molecular weight exclusion filter (polyethersulfone Ultracel centrifugal membrane filter, Millipore Corporation). The following IgG fractions were tested in the Seprion assay: 1) non-aggregated (ie prior to aggregation) and not filtered; 2) non-aggregated but filtered; 3) aggregated and unfiltered; and 4) aggregated and filtered. The Seprion assay was performed in triplicate with the same amount of protein from each fraction, making the assumption that no protein had been lost in the filtration process.
The results from the assay are shown in
The background signal from the non-aggregated IgG was similar whether filtered or not, which is interpreted as indicating that there were no aggregates in this fraction and that there was no observable loss of protein due to the filtration process. There was a high signal from the aggregated unfiltered IgG, which was lost on filtration, indicating that the aggregates had been removed by the filtration process.
Thus, this Example demonstrates that the assay was detecting aggregates of IgG that were removed by filtration through the 300 kDa nominal molecular weight filter.
In order to further investigate the size of aggregates that are detected by the assay, aggregated recombinant human serum albumin (HSA) was tested three times under the following conditions: 1) prior to filtration; 2) after filtration through a 300 kDa nominal molecular weight filter; and 3) after additional filtration of the filtrate from 2) through a 100 kDa nominal molecular weight filter. The results are shown in
The results show a progressive loss of signal with each filtration step as aggregates of different sizes are removed from the HSA solution. Significantly, there was a higher signal from the HAS that had been filtered through the 300 kDa filter compared to the signal generated by subsequent filtration through the 100 kDa filter. This signal is likely to be generated by aggregate within the nominal molecular weight range of 100-300 kDa. Given that the molecular weight of HSA is approximately 68 kDa, the signal is likely to have been generated by aggregates comprising only a small number of rHSA molecules.
Whilst the invention has been described with particular reference to preferred embodiments thereof it will be appreciated that many modifications and variations thereof are possible within the general scope of the invention. Any variation of the invention as explicitly claimed which would operate in the same way to produce the same result is to be within the protection conferred by the application.
In this specification, unless expressly otherwise indicated, the word ‘or’ is used in the sense of an operator that returns a true value when either or both of the stated conditions is met, as opposed to the operator ‘exclusive or’ which requires that only one of the conditions is met. The word ‘comprising’ is used in the sense of ‘including’ rather than in to mean ‘consisting of’.
| Number | Date | Country | Kind |
|---|---|---|---|
| 0204797.5 | Feb 2002 | GB | national |
| 0216808.6 | Jul 2002 | GB | national |
| 0229614.3 | Dec 2002 | GB | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10506131 | Aug 2004 | US |
| Child | 12644283 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 12644283 | Dec 2009 | US |
| Child | 13183063 | US |
