Background Art C-reactive protein (CRP), named because of its reactivity with the C polysaccharides of pneumococcal cell wall was first discovered by Tillet, W. S. and Francis, T. in 1930 ( (1930) Journal of Experimental Medicine 52, 561-571). CRP is a positive acute phase protein, the plasma levels of which rise dramatically in response to tissue injury, infection or inflammation from less than one to several hundred micrograms per millilitre and fall just as rapidly after recovery or treatment. The half-life of CRP in the circulation is approximately 19 hours. CRP is synthesised almost exclusively by the hepatocytes however there is also evidence of extra-hepatic synthesis. The plasma concentrations of CRP increase as a result of the accelerated rate of transcription of the respective genes in the liver. Interleukin-1, interleukin-6 (IL-6) and transforming growth factor- ? 6 stimulate the transcriptional activation of CRP and other acute phase proteins. IL-6 which is produced by activated leucocytes, fibroblasts and endothelial cells is the major stimulant for CRP production.

CRP is a member of the pentraxin family of proteins and is well conserved biologically throughout evolution. It shows approximately 70% sequence homology with another member of this family Serum Amyloid P (SAP) which is also present in the plasma. No CRP deficiency diseases have ever been identified which suggests that it has a crucial in vivo role. CRP has been detected in all mammals, birds, amphibians and in marine teleosts and is composed of five identical subunits bound non-covalently to each other in a cyclic pentameric structure. Each subunit has 206 amino acids and an approximate molecular weight of 21,000. CRP requires calcium for ligand binding and five calcium dependent ligand binding sites are present on one face of the CRP molecule through which it recognises phosphorylcholine (PC, a hydrophilic portion of phosphatidylcholine on pneumococcal C polysaccharide) and other bacterial products. Also present on the CRP molecule is a hydrophobic pocket which accommodates the methyl groups of phosphorylcholine. PC is also found in phospholipids in cell membranes and plasma lipoproteins and in the complex polysaccharides of plants, fungi and bacteria.

CRP has two major biological functions: (i) A role in the innate immune response against infection: Here, CRP activates complement by the classical Clq pathway and reacts with Fey receptor and possibly other receptors on phagocytic cells. CRP activation of the classical Clq pathway differs from IgG activation of the pathway in that it is localised to the collagen-like regions instead of the globular head groups of Clq.

CRP is most efficient at early complement pathway activation. (ii) Removal of membrane and nuclear material from necrotic cells: CRP binds to damaged tissue, to nuclear antigens and to certain pathogenic organisms in a calcium dependent manner. The calcium sites react with components including phosphatidylcholine and sphingomyelin on damaged cell membranes when exposed during injury. CRP also binds to chromatin which has been exposed during necrosis or apoptosis and to small ribonuclear proteins on damaged cell membranes.

CRP detection has been used for many years in the diagnosis of tissue injury or inflammation as high concentrations can be found in patients suffering from infection, inflammation, trauma, malignancies, stress, arthritis, surgery and acute myocardial infarction. CRP detection is also found to be useful in monitoring efficacy of therapy to the aforementioned diseases.

Until recently, the prediction of cardiovascular disease relied on factors such as hypercholesterolemia and hypertension, however these tests failed to predict the risk of coronary thrombosis in one third of cases tested and so a non-conventional method such as CRP detection was considered. CRP is produced due to underlying vascular inflammation and it was found that baseline levels of CRP can predict future risk of stroke, myocardial infarction and peripherial arterial disease in both healthy and high-risk individuals such as smokers, the elderly and angina patients. The multiple Risk Factor Intervention Trial was the first to demonstrate the role of CRP and inflammation have as cardiovascular risk factors (Kuller, L. H. et al., (1996) American Journal of Epidemiology 144 (6), 537-547). Conventionally Troponin T and I

were used as markers for patients with acute coronary syndrome.

Troponin T is still found to be a more accurate predictor than CRP in the first 72 hours of illness however CRP can be used independently to predict both cardiac risk and coronary revascularisation during a six month follow up period. It has been shown in studies which had in- hospital death and acute myocardial infarction as end-points, it was found that CRP was not predictive (Heeschen, C. et al., (2000) Journal of the American College of Cardiology 35 (5) 1535-1542). However when the observation period was extended to after discharge from hospital, CRP detection proved to be predictive of future coronary events. Thus, Troponin and CRP detection facilitate risk evaluation at different stages of acute coronary syndrome and therefore the two tests together give more accurate predictions.

Conventional test systems for CRP such as nephelometry, immunoturbidimetry and enzyme-linked immunosorbent assay (ELISA) all depend on the availability of antibodies (immunoglobulins), either polyclonal or monoclonal, for the detection of CRP in both normal and disease-state plasma specimens obtained from human individuals.

Indeed, most ELISA tests use antibodies for both CRP capture and detection. Other factors of concern are the reproducibility and cost of antibody supply. Polyclonal antibodies are produced upon immunisation of an animal with the target antigen. It is generally accepted that that process is unsatisfactory for the large-scale production of immunodiagnostic reagents (antibodies) due to the variability in immune response between different immunised animals which results in the production of antibody preparations of differing specificities and

sensitivities of detection. Monoclonal antibody production does offer an alternative route of supply, however the costs involved for initial hybridoma generation and subsequent large-scale in vitro production of antibody can be prohibitive.

EP-A 1 076 240 describes and claims a method for determining levels of CRP using labelled PC, for example Eu3+ labelled PC. Two types of sandwich assay methods using PC are described, namely antibody-PC (Method B) and PC-antibody (Method C). The methods described in EP-A 1 076 240 have the drawback involved in the use of immunological methods set out above.

Furthermore, the detection of CRP in veterinary specimens, as a marker of animal infection or tissue inflammation, is not a straightforward matter due to the species specificity of antibodies directed against human CRP. In other words, test systems which can detect human CRP are species-specific and do not readily detect animal CRP molecules. This factor is of major concern from the viewpoint off animal welfare, whereby a generic CRP detection system for veterinary specimens would greatly facilitate improvement in both the diagnosis of animal disease and animal husbandry methods.

Disclosure of Invention The invention provides a method of detecting or determining C- reactive protein (CRP) in diverse animal species, which method comprises contacting a sample of a biological fluid from a human or non-human animal with a complex of a phosphorylated compound

containing a nitrogen moiety and a label which is directly detectable in a non-immunological assay, the phosphoryl moiety and the nitrogen moiety being positioned relative to each other so as to permit binding to calcium-dependent ligand binding sites present on CRP.

The method according to the invention enables one to carry out in a simple and rapid manner assays, some involving no signal amplification steps, which facilitate the non-immunological detection of CRP in a range of species including human, porcine and canine species.

These assays enable the determination of the normal versus disease-state where elevated CRP levels are present. The method is based solely on the interaction of CRP with the phosphorylated compound.

Preferably, the complex has the formula wherein: XisHoragroup - Z- (W) M-y

in which Z is a linker molecule; m is 0 or 1 W when present is a peptide or polypeptide; and Y is a label wherein when X is not H X'is a group S where S is a pyridoxal group, a tyrosinyl group or a group wherein R1 is H, (C1-C4) alkyl or a carboxy group; R2 is H or (Cl-C4) alkyl; R3 is H or (Cl-C4) alkyl; n is 1 or 2,

or wherein when X is H, X'is a substituted or unsubstituted adenosyl or guanosyl group linked directly through the sugar moiety to a label or indirectly to said label by means of a peptide or polypeptide molecule, the label being directly detectable in a non-immunological assay.

Certain of the nitrogen-containing phosphorylated compounds embraced by Formula I were investigated for their ability to bind with CRP by Tanaka, T and Robey, Frank, A ( (1983) The Journal of Immunological Methods 65, 333-341).

Further, preferably, the complex has the formula wherein: Ri is H, (C-C4) alkyl or a carboxy group; R2 is H or (Cl-C4) alkyl; R3 is H or (Cl-C4) alkyl; m is 0 or 1 ; n is 1 or 2 ;

Z is a linker molecule; W when present is a peptide or polypeptide; and Y is a label which is directly detectable in a non- immunological assay.

A preferred compound is one wherein Rl and R2 each represent a methyl group and n is 2 and Z is a cytidinyl phosphate moiety.

The inclusion of"W"is optional as indicated by the values for m, such that the label can be linked directly to the N-containing phosphorylated moiety.

When W is present, it is preferably a protein selected from bovine serum albumin (BSA), casein, ovalbumin, thyroglobulin and keyhole limpet haemocyanin (KLH).

A preferred protein is BSA.

Preferably, the label is selected from a microparticle, a gold label, an enzyme, a radio label and a lanthanide.

According to one embodiment, the label is a latex sphere or a gold label which is detectable in a nephelometric or turbidimetric assay as hereinafter described.

According to an alternative embodiment the label is the enzyme horseradish peroxidase which is detectable in an enzyme-linked sorbent assay as hereinafter described.

Preferably, the biological fluid is from a canine, porcine or human animal.

By biological fluid herein is meant a biological fluid as such including blood or a blood fraction such as plasma or a medium such as a buffer which is used to suspend or dilute the biological fluid.

In vivo and in vitro, C-reactive protein (CRP) interacts strongly, in a calcium (Ca2+) dependent manner, with nitrogen-containing compounds such as PC. Thus, PC binding offers a method to capture or immobilise CRP and facilitate subsequent detection.

Brief Description of Drawings Fig. 1A is a schematic representation of a turbidimetric assay system for CRP detection in accordance with the invention; Fig 1B is a schematic representation of a CRP-ELSA in accordance with the invention; Fig. 2 is a plot of A Absorbancessonm versus CRP concentration (pg/ml) for the turbidimetric assay system as described in Example 3; Fig. 3 is a plot of Absorbance450/620nm versus CRP concentration (p, g/ml) for the ELSA assay system as described in Example 8;

Fig. 4 is a linear regression analysis of CRP detection by the ELSA system versus a conventional nephelometric assay (R= 0.92) as described in Example 9; Fig. 5 illustrates the detection of CRP in porcine plasma specimens (g/ml) exhibiting a range of CRP concentrations as described in Example 9; and Fig. 6 illustrates the variation in CRP levels in canine serum specimens obtained prior to and post-cruciate ligament surgery, as measured in the ELSA and an immunoassay utilising IgG [anti-canine CRP] as described in Example 9.

Modes for Carrying Out the Invention The invention is described hereinafter with reference to a derivative of PC (cytidine 5'-diphosphorylcholine: CDPC) which contains a phosphorylcholine moiety and this was reacted with a carrier protein"W" (e. g. , bovine serum albumin: BSA), using schiff-base chemistry, to produce a stable phosphorylcholine-BSA (PC-BSA) reagent. CDPC has the following formula:

For example, the PC-BSA moiety can be either: (i) covalently coated directly onto microparticles and subsequently interacted with CRP from a number of species in a quantitative manner to form spectrophotometrically detectable complexes; or (ii) PC-BSA can be further modified by covalent attachment of horseradish peroxidase (HRP) to form PC-BSA-HRP which, in turn, can be used in a non- immunological enzyme-linked sorbent system (ELSA) to quantitatively detect CRP in biological fluids from various animal species.

In order to produce PC-BSA conjugates CDPC can be oxidised with sodium periodate to yield a reactive group capable of reacting with free amino (-NH2) groups present on the carrier protein BSA. This bond can then be stabilised by the addition of sodium borohydride which catalyses the formation of a stable schiff base. The PC-BSA conjugate still contains free-NH2 groups which can be used for example to facilitate molecular coupling to either carboxyl-activated latex microparticles via carbodiimide chemistry or to horseradish peroxidase via SMCC (4- (maleimidomethyl) cyclohexanecarboxylic acid N-

hydroxysuccinimide ester) activation of PC-BSA and reaction with SATA (S-acetyl thioglycolic acid N-hydroxysuccinimide)-activated horseradish peroxidase.

A microparticle based assay is based on the principle that immobilised PC-BSA present on latex microspheres, typically of 200nm diameter, will form a stable non-covalent complex with free CRP present in a buffer or biological fluid. This complex is spectrophotometically detectable by measuring the change in absorbance (AA) at a particular wavelength, preferably 550nm, with time (typically 5 minutes) which in turn is dependent on the CRP amount or concentration ( [CRP]) present in the reaction cuvette. Significantly, no signal amplification methodologies or antibody presence is involved in this method of detection. A schematic representation of the turbidimetric assay system is depicted in Fig. 1A.

In Fig. 1A immobilised PC (black circles) on latex particles binds free CRP in solution (indicated by dashed lines) to form insoluble complex. The extent of complex formation, measured spectrophoto- metrically, is proportional to CRP concentration.

The enzyme-linked sorbent assay (ELSA) is a competitive assay format for the detection of free CRP in biological fluids and diluents, the function of which is also antibody-independent. In essence, standards or specimens (with or without CRP) and phosphorylcholine-BSA-HRP (PC-BSA-HRP) conjugate are jointly added to microwells pre-coated with CRP. If there is no free CRP in the specimen (or zero standard) then the PC-BSA-HRP conjugate will preferentially bind to the CRP present

on the microplate. Following a wash step, to remove unbound material and substrate (tetramethylbenzidine) addition, an intensely blue coloured product (which turns yellow upon addition of acid with absorbance at 450nm) will develop due to HRP action. Conversely, if CRP is present in the specimen, then it will compete with immobilised CRP for binding to PC-BSA-HRP conjugate. On this occasion, the wash step will remove solution phase CRP: PC-BSA-HRP complexes resulting in less immobilised PC-BSA-HRP present in the microwell. Thus, reduced product formation is observed upon substrate addition in this case. In summary, ELSA colour development is inversely proportional to specimen CRP concentration. A schematic representation of the CRP- ELSA assay is depicted in Fig. 1B.

In Fig. 1B free CRP (pentagon) competes with immobilised CRP for binding to PC-BSA-HRP conjugate. Product formation, measured spectrophotometrically, is inversely proportional to CRP concentration.

Example 1 Conjugation of phosphorylcholine to bovine serum albumin (BSA) All manipulations were carried out at room temperature except where otherwise stated and were performed essentially as described by Eckersall P. D. et al ( (1989) Biochemical Society Transactions, 414.).

75mg (30mg/ml) CDPC was oxidised for 20 minutes by the addition of 2.5 ml of 0. 1 M sodium periodate, after which 0. 15ml of 1M ethylene glycol was added. Then 140mg BSA (28mg/ml) -previously resuspended in 5ml of 0.01 M bicarbonate buffer and dialysed at 4°C, with stirring-

was added to the CDPC. The pH of the reaction was maintained at 9.6 for 1 hour, using 5% sodium carbonate, after which reduction was performed overnight by the addition of 0.5M sodium borohydride (5ml).

The pH was then lowered to pH 3.5 for 1 hour by adding 1M formic acid and then neutralised with 1M sodium hydroxide. Following this, the reaction mixture was dialysed exhaustively at 4°C in 50mM potassium phosphate, 1mM EDTA, 150mM NaCl pH 7.8, with stirring, and then stored at-20°C until used for conjugation to horseradish peroxidase or coupling to microparticles.

Example 2 Covalent coupling of PC-BSA to carboxy-modified microparticles A solution containing the following components was prepared: 1001 0. 5M MES buffer pH 6.1, 546µl of deionised water, 1 00Ll 10% (w/v) latex particles, 230ut sulfo-NHS (50mg/ml) and 24µl 1-ethyl-3-(3- dimethylaminopropyl) carbodiimide (EDAC) (lOmg/ml). The components were mixed together on a shaker at room temperature for 30 minutes followed by centrifugation at 10, 000g for 20 minutes. The supernatant was discarded. The particles were resuspended by sonication in lml 50mM MES buffer pH 6.1 for washing and again centrifuged at l0, 000g for 20 minutes. The supernatant was discarded and the particles were resuspended by sonication in lml of PC-BSA protein stock (lmg/ml) in 0. 1M sodium borate buffer pH 8. 5 and mixed at room temperature for 1 hour with shaking. Then, zu 10. 25M ethanolamine hydrochloride solution in O. 1M sodium borate buffer was added and the

solution mixed for a further 30 minutes at room temperature. Following centrifugation at 10, 000g for 20 minutes and removal of the supernatant, the activated particles were resuspended in lml of a 1 Omg/ml BSA in 0. 1M sodium borate buffer (blocking buffer) by sonication and mixed for 30 minutes at room temperature. Finally, the particles (1% (w/v) suspension) were washed twice with 1ml aliquots of blocking buffer and finally resuspended in lml of 50mM Tris-HCl, 25mM NaCl, lOmM CaCl2 pH 7.4, with sonication, and stored at 4°C until required for use.

Example 3 Turbidimetric assay format Two diluents were employed for performance of a CRP turbidimetric assay, namely: (i) Specimen diluent comprised O. 1M glycine, 90mM NaCl, 50mM calcium chloride, 0.04% (w/v) sodium azide pH 7.6 ; and (ii) Particle diluent comprised 0.4% (w/v) PEG 6000 in sample diluent. Prior to specimen dilution, PC-BSA coated microparticles were diluted to 0.03% (w/v) in particle diluent to give 'working reagent'. Working reagent, standards, specimen diluent and specimens were then loaded onto a Cobos Mira (Cobos Mira is a trade mark) autoanalyser pre-programmed to dispense either standards or specimens (5p1 each) along with specimen diluent (45, u1) and working reagent (450ut) into reaction cuvettes, and brought to 37°C. The reaction was initiated by addition of the working reagent and allowed to proceed at 37°C for 5 minutes. The absorbance (550nm) of all standards and specimens was determined at T = 0 and at T = 5 minutes. Upon completion of the analysis, the SA value for each standard and specimen

was computed and specimen CRP concentrations were calculated from a plot of AAssonm (standards) versus [CRP] pg/ml as depicted in Fig. 2.

Example 4 SMCC (4-(maleimidomethyl ! cyclohexanecarboxylic acid N- hydroxysuccinimide ester, activation of BSA PC-BSA (lmg/ml) was modified with SMCC by adding 25mM SMCC to PC-BSA at a ratio of7. 5ul 25mM SMCC per mg PC-BSA.

The solution was mixed while making this addition and then incubated at room temperature for 30 minutes. The SMCC-activated PC-BSA (PC- BSA-SMCC) was then dialysed exhaustively against 50mM potassium phosphate, 1mM EDTA, 1 50mM NaCl pH 6.8, with stirring, at 4°C.

Example 5 SATA (S-acetyl thioglycolic acid N-hydroxysuccinimide) activation of horseradish peroxidase (HRP) Horseradish peroxidase (HRP) was prepared at a concentration of 4mg/ml in 50mM potassium phosphate, 1mM EDTA, 150mM NaCI pH 7.8. Sulphydral groups were introduced into horseradish peroxidase by adding 0. 05mg SATA per mg HRP (9 fold molar excess) and incubating at room temperature for lhour. Following this reaction, excess SATA was removed by exhaustive dialysis against 50 volumes of 50mM potassium phosphate, 1mM EDTA, 150mM NaCI pH 6.8, with stirring,

at 4°C. SATA-activated hoseradish peroxidase (SATA-HRP) was then stored at-20°C until required for conjugation to PC-BSA-SMCC.

Example 6 Conjugation of HRP and BSA-PC SATA-HRP was deblocked (i. e. , sulphydral groups exposed) by the addition of 500mM hydroxylamine (1/10 volume of SATA-HRP solution) and incubated, in the dark, at room temperature for 1 hour. PC- BSA-SMCC was then added to deblocked HRP in the ratio of lmg PC- BSA-SMCC per 4. 6mg of SATA-HRP, equivalent to a PC-BSA-SMCC: SATA-HRP molar ratio of 1: 7. This mixture was further incubated, in the dark, for 5 hours at room temperature. Six microlitres of 12mM 2- mercaptoethanol were then added for each mg of PC-BSA-SMCC present to react with, and block, any residual maleimide groups. Then, 2jLLl lOOmM N-ethylmaleimide (NEM) were addedper mg PC-BSA- SMCC present to react with both excess 2-mercaptoethanol and residual sulphydral groups on HRP. The mixture was then dialysed exhaustively against 50mM potassium phosphate, 1mM EDTA, 1 50mM NaCl pH 6.8 with stirring. PC-BSA-HRP conjugate formation was confirmed by SDS- PAGE analysis and functionality testing.

Example 7 Microtitre plate coating with human CRP Human CRP was diluted to lu-g/ml in 10mM sodium bicarbonate buffer pH 9.4. This solution was then added to NUNC Maxisorp (NUNC Maxisorp is a trade mark) flat bottomed plates (100µl per well). The coated plates were incubated at 4°C for approximately 16 hours to allow antigen attachment and then washed 4 times with PBST and tapped dry.

Microplates were blocked using a solution of 5% (w/v) Sucrose, 0. 5% (w/v) BSA in PBS for 1 hour at 37° (200 gl per well). The blocking solution was then removed, the microplates tapped dry to remove any remaining blocking solution and incubated overnight at 37°C to dry. The coated microplates were then stored at 4°C until required for use.

Example 8 ELSA assay format Human CRP standards (range from O-lO ! lg/ml) were prepared by serial dilution of a stock solution of CRP (24mg/ml) in 50mM MES, 25mM NaCl, 10mM CaCl2 and 0. 1% (v/v) Tween-20@ pH 6.0 as follows:

Human [CRP] Dilution (Rg/ml) 10.0 1/240 1.0 1/10 0.75 3/4 0.5 2/3 0.125 1/4 0.064 1/2 0.008 1/8 0.01 1/8 PC-BSA-HRP conjugate prepared in Example 6 and specimens for analysis were also diluted in 50mM MES, 25mM NaCl, l OmM CaCl2 and 0. 1 % (v/v) Tween-20 pH 6.0. Next, 50gel of standards or specimens diluted 1/40, 1/50 and 1/80 for porcine, human and canine CRP detection, respectively and 50p1 ofthe conjugate (diluted 1/8000) were added into microwells previously coated with 1 µg/ml human CRP (lOOng/well human CRP) as described in Example 7. After incubation at room temperature for 30 minutes, the microwells were washed 4 times with 50mM Tris-HCl, 25mM NaCl, IOmM CaCl2, 0. 1 % Tween-209 pH 7.4, with a 30 seconds soak stepper wash. The microwells were tapped dry and 100p1 TMB were added per well followed by a 15 minute incubation. The reaction was stopped by adding 100µl 1N H2SO4 and the

absorbance read at 450/630nm followed by calculation of specimen [CRP] from the standard curve depicted in Fig. 3.

Example 9 Assay Evaluation The CRP-ELSA described herein was used to evaluate the levels of CRP in 23 human serum specimens in parallel with analysis of identical specimens by conventional nephelometric analysis (Behring Nephelometric Systems, Dade Behring Marburg GmbH). The results are shown in Fig. 4.

Figure 4 illustrates that all specimens tested gave near identical results using both systems with an observed correlation coefficient of 0.92. To further evaluate the CRP-ELSA for the detection of porcine and canine CRP, seven porcine plasma and six canine serum specimens were analysed.

The results obtained following the analysis of the seven porcine specimens are depicted in Fig. 5. Specimens 1-3 exhibited high levels of CRP (> 20 pg/ml) while specimens 4-7 all contained lower levels of CRP (<10, ug/ml). This assay confirmed the utility of the CRP-ELSA described herein to detect variable levels of porcine CRP in clinical specimens.

Finally, Fig. 6 illustrates the results following the evaluation of canine serum CRP levels pre-and post-cruciate ligament surgery;

wherein the open bars represent CRP concentration (pg/ml) measured by ELSA in accordance with the invention and closed bars represent CRP concentration (jLLg/ml) measured by other method.

In Fig. 6, X-axis labels: numbers 1 and 2 refer to day 1 before surgery and immediately post-surgery, respectively. Number 3 corresponds to 12 hours post-surgery while numbers 4,5 and 6 relate to days 1, 15 and 48 post-surgery. It can be seen that alterations in canine CRP levels were apparent at day 1 which correspond to those expected following invasive surgical procedures. Furthermore, the subsequent recovery of the animal, post-surgery, is evident as the levels of CRP reduce to normal levels by day 15-levels which are maintained at day 48. These alterations in canine CRP levels were also detectable using an immunoassay method employing anti-canine CRP antibodies referred to herein as'other method'.