NOVEL ARTIFICIAL PHOSPHOLIPID-PROTEIN BIOCONJUGATES FOR BIOMOLECULAR RECOGNITION

Abstract

There is provided phospholipid-peptide or phospholipid-protein conjugates useful for identification of autoantibodies, and in particular autoantibodies directed towards phospholipids. There is further provided intermediates and methods for their preparation as well as synthesis of the phospholipid-peptide or phospholipid-protein conjugates.

Claims

1. A phospholipid protein conjugate of the formula (I): wherein L is a linker, such as 1,2,3-triazole, an amide with a general formula RC(O)NHR wherein R is a phospholipid and R is protein/peptide; a polyethyleneglycol (PEG) or their combination, and wherein the phospholipid and protein or a peptide are biologically complementary.

2. The phospholipid-peptide or phospholipid-protein conjugate according to claim 1, wherein the protein or peptide is selected from beta2-glycoprotein I, prothrombine, and synthetic peptides having at least 10 amino acid residues up to 100.

3. The phospholipid-peptide or phospholipid-protein conjugate according to claim 1, wherein the phospholipid is selected from cardiolipin, a fatty acid derivative of phosphoethanolamin, such as 1,2-distearoyl-phosphoethanolamine.

4. The phospholipid-peptide or phospholipid-protein conjugate according to claim 1, wherein the linker L is 1,2,3-triazole conjugated with an amide.

5. A phospholipid-peptide or phospholipid-protein conjugate according to claim 1, wherein the conjugate is selected from compounds of formula BC-1 and BC-2: wherein b2GPI denotes 2-glycoprotein I, and PT denotes prothrombin.

6. Phospholipid-protein conjugate of claim 1 having the following formula: ##STR00020## wherein: ##STR00021##

7. Phospholipid-protein conjugate of claim 1 having the following formula: ##STR00022## wherein: ##STR00023##

8. Phospholipid-protein conjugate of claim 1 having the following formula: ##STR00024## wherein: ##STR00025##

9. Phospholipid-protein conjugate of claim 1 having the following formula: ##STR00026##

10. Phospholipid-protein conjugate of claim 1, selected from the group consisting of a compound having the following formula (BC-7) and (BC-8): ##STR00027## wherein n is an integer in the range of 1-20 and Y.sup.+ is any appropriate cation.

11. Phospholipid-protein conjugate of claim 10, wherein Y.sup.+ is Na.sup.+, K.sup.+, or Ca.sup.2+.

12. Phospholipid-protein conjugate according to claim 10, wherein the protein/peptide is selected from the group consisting of -2-glycoprotein I, prothrombine, and synthetic peptides having at least 10 amino acid residues up to 100.

13. Phospholipid-protein conjugate according to claim 10, wherein the protein/peptide is selected from the group consisting of beta2-glycoprotein I, prothrombine.

14. A cardiolipin analogue selected from one of the following formulae (CL-1) to (CL-6) for conjugation with a protein or peptide: ##STR00028## ##STR00029## wherein Y.sup.+ may be any appropriate cation.

15. A cardiolipin analogue according to claim 14, wherein Y.sup.+ is Na.sup.+, K.sup.+, or Ca.sup.2+.

16. A protein or peptide derivative selected from the group consisting of compounds of the formula: ##STR00030##

17. The protein or peptide derivative of claim 16, wherein the protein is selected from the group consisting of -2-glycoprotein I, prothrombine, and synthetic peptides having at least 10 amino acid residues up to 100.

18. The protein or peptide derivative of claim 16, wherein the protein is selected from the group consisting of beta2-glycoprotein I, prothrombine.

19. Enzyme-linked immunosorbent assay (ELISA) for detection of autoimmune antibodies against the conjugates according to claim 1.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0044] FIG. 1 shows a synthethic strategy for preparation of novel bioconjugates by SPAAC click chemistry. (b) Detecting specific antibodies by the novel phospholipid-peptide and phospholipid-protein bioconjugates in homogeneous and 2D array formats.

[0045] FIG. 2 shows a general strategy for synthesis of novel cardiolipin analogues with improved chemical stability.

[0046] FIG. 3 shows a scheme of enzyme-linked immunosorbent assay (ELISA) for detection of autoimmune antibodies against antigens of the present invention.

[0047] FIG. 4 shows comparative performance of phospholipid antigens of the present invention in ELISA detection of autoimmune antibodies.

[0048] FIG. 5 shows synthetic strategies for cardiolipin-beta-2GPI and cardiolipin-PT complexes, particularly compounds 7 and 8 therein. Reagents and conditions: (i) 0.1 M bicarbonate buffer-DMSO 9:1, 4 C., 12 h; (ii) corresponding azide, CuSO4:TBTA 1:1.1, ascorbic acid, 1PBS-DMSO-t-BuOH 3:2:0.1, v/v/v; (iii) KMnO4, NaIO4, t-BuOHH2O 9:1, v/v, rt, 12 h; (iv) succinimide ester, N,N-diisopropylcarbodiimide, DMSO, rt, 12 h; (v) proteins 2, 3, 8, 0.1 M bicarbonate buffer-DMSO 9:1, v/v, rt, 12 h; (vi) KMnO4, NaIO4, t-BuOHH2O 9:1, v/v, 45 C., 1 h, microwave reactor; (vii) succinimide ester, N,N-diisopropycarbodiimide, DMSO, rt, 5 h; 3-azidopropan-1-amine, rt, 12 h.

[0049] FIGS. 6A and 6B shows results of IgG ELISA assay [absorbance for grouped patients (6B) and healthy controls (6A) across antigens], obtained using compounds 7-8 of FIG. 5 and controls. Negative and positive controls (HNP and HCL, Immunovision) are shown as a x and a star, respectively. Conjugates 5 and 6 in FIG. 6A corresponds to compounds 6 and 7 respectively of Scheme VI herein.

[0050] FIG. 7 shows data plot for observed correlations between aPLPs and clinical parameters for disease stated patients (SUH cohort). Students's t-test indicated a correlation between Smith positivity and elevated aPLP titers using exclusively antigen 8 of FIG. 5.

[0051] FIG. 8 shows linear regression plot for observed correlations between aPLPs and clinical parameters for disease stated patients (SUH cohort). Ordinary least squares (OLS) analysis was performed using Stata. The difference in aPLP titers for Smith negative and Smith positive groups was found to be statistically significant with p value of 0.002 for antigen 8 of FIG. 5.

DETAILED DESCRIPTION OF THE INVENTION

[0052] Herein, the bioconjugates in focus will be cardiolipin and phosphatidylserine attached to beta2-glycoprotein I (beta2-GPI), prothrombin and their short immunogenic fragments (peptide with a general sequence An, where A is an amino acid and n is between 10 and 50). Synthetic cardiolipin analogues with increased chemical stability are prepared in accordance with FIG. 2, and then applied in bioconjugation with beta2-glycoprotein I or its short fragment.

[0053] Next, chemical modification of the novel azide derivatives of cardiolipin and phosphotidylserine will be linked to the terminal alkyne or difluorinated cyclooctyne moiety of the corresponding proteins under CuAAC and SPAAC reaction conditions (FIG. 1a).

[0054] In preparation of phospholipid-peptide and phospholipid-protein conjugates for diagnostics and studies of autoimmune disorders, convenient bioconjugation strategies are on high demand. Among other methods available to date, two variants of azide-alkyne cycloaddition, namely a copper-catalyzed and strain-promoted click reactions (CuAAC and SPAAC, respectively), were recently demonstrated as promising synthetic approaches giving stable 1,2,3-triazole products both in vitro and in vivo (H. C. Hang, J. P. Wilson, G. Charron. Acc. Chem. Res. 2011, 44, 699, and J. A. Precher, C. R. Bertozzy. Nat. Chem. Biol. 2005, 1, 13). Thus, azide- and alkyne-functionalized fatty acid and lipid reporters were successfully attached to cells by CuAAC reaction providing powerful tools for monitoring lipid trafficking and metabolism (H. C. Hang, J. P. Wilson, G. Charron. Acc. Chem. Res. 2011, 44, 699). Furthermore, bioconjugation of biotin with several lipid reporters was achieved in high yields using CuAAC protocol (J. A. Precher, C. R. Bertozzy. Nat. Chem. Biol. 2005, 1, 13). The disadvantage of CuAAC reaction is a high toxicity of copper required for the reaction. This limits CuAAC bioconjugation strategy to exclusively in vitro conditions. In turn, SPAAC approach was extensively utilized for bioconjugation in vivo giving an opportunity to monitor cellular processes and metabolic pathways of whole organisms, although with a main focus on proteins and glycans (J. A. Precher, C. R. Bertozzy. Nat. Chem. Biol. 2005, 1, 13).

[0055] In order to perform bioconjugation by SPAAC approach, azide and cyclooctyne functionalities need to be attached to biomolecules (FIG. 1a). This can be done using commercially available reagents and following protocols described in literature. (M. G. Paulick, A. R. Wise, M. B. Forstner, J. T. Groves, C. R. Bertozzi. J. Am. Chem. Soc. 2007, 129, 11543) Providing high yields of the desired bioconjugates in mild nontoxic conditions, SPAAC click chemistry seems to be an ideal method of choice in preparation of novel phospholipid-protein conjugates which can be applied for bioanalysis both in vitro and in vivo.

[0056] Current bioanalysis can be performed in several formats requiring preparation of different reagents and application of conditions adapted for exactly this method. One promising method for detection of specific antibodies is a homogeneous (all-in-solution) analysis which is a simple and rapid technique (M. Chassignol, Y. Aubert, V. Roig, U. Asseline. Nucleos. Nucleot. Nucl. Acids 2007, 26, 1669). The homogeneous analysis can be performed using, for example, fluorescent labeling of target biomolecules and following their interaction by monitoring fluorescence signal. Immobilization on 2D array is another promising method of bioanalysis, which is especially powerful for serodiagnostics of human disorders due to high sensitivity of this method and possibility of target antibody's type validation. Proteins, peptides and DNA can be immobilized on 2D arrays following previously described methods (A. M. Rouquette, C. Desgruelles. Lupus 2006, 15, 403, and references cited therein. and A. Castro. H. Wang. WO 2007/061793 A2, PCT/US2006/044572). However, in case of relatively small lipid molecules a protein cargo is required for attachment to the array which protects lipid from interactions with surface and, therefore, prevents loss of its immunogenic activity. Previously, nitrocellulose strips containing natural cardiolipin attached to BSA and KLH proteins were prepared and further applied in serodiagnostic studies (A. Castro. H. Wang. WO 2007/061793 A2, PCT/US2006/044572). Providing high immunogenic activity right after preparation, the resulting conjugates were rapidly degraded under array storage by unknown mechanism (This observation has been recently made at the Central Research Institute of Epidemiology, Moscow, Russia, while repeating synthetic procedures described in WO 2007/061793 A2). Therefore, a perspective direction in this area is a preparation of synthetic cardiolipin analogues with increased chemical stability compared to natural prototypes followed by attachment to proteins/peptides and immobilization (J. S. Yadav et al. Tetrahedron Lett. 1996, 37, 6603 and K. Kasireddy et al. Bioorg. Chem. 2005, 33, 345).

EXAMPLES

Example 1Synthetic Approach to Protein Phospholipid Conjugates of Formula 10-11 of FIG. 2

[0057] The reagents and conditions for the synthesis according to FIG. 2 are:

[0058] a) 9-decenoic acid, DCC, DMAP, DCM, 0 C.; b) NaH, 10-bromo-1-decene, DMF, 0 C. to rt; c) DDQ, DCM, H.sub.2O, 0 C.; d) MeOPCl.sub.2, DIPEA, THF, 78 C.; HOCH.sub.2CH(OMPM)CH.sub.2OH, 78 C. to rt; 30% aq. H2O2, DCM, rt; e) DDQ, DCM, H2O, 0 C.; f) NaI, 2-butanone, 80 C.; g) NaIO.sub.4, KMnO.sub.4, tBuOH, H2O, rt; h) DIC, NHS, rt; i) Protein, buffer 12t, rt.

[0059] The anchoring of the analogues (8-9) of FIG. 2 to protein or peptide may be achieved by attachment of fatty acid residues with terminal double bonds oxidized with NaIO4-KMnO4, and afterwards NHS-assisted coupling to the protein. Alternatively, CuAAC and SPAAC approaches using azide-derivatives of the phospholipids 8-9 and the alkyne-labeled proteins can be performed. For these conjugations the above mentioned step h) is as follows:

[0060] SPACC Conjugation

[0061] h) DIC, NHS, 12t, rt; i) 3-azidopropan-1-amine, tBuOH, Et.sub.3N, 3t, rt; j) Protein/peptide cyclooctyne derivative, 3-24 t,rt.

[0062] CuAAC Conjugation

[0063] h) DIC, NHS, 12t, rt; i) 3-azidopropan-1-amine, tBuOH, Et.sub.3N, 3t, rt; j) Protein/peptide alkyne derivative, Cu-TBTA, ascorbic acid, 3-24 t,rt.

Example 2Fluorescent Labeling and ELISA Assays

[0064] Documented immunogenity of synthetic lipids accompanied by verified chemical structure and purity of the phospholipid-peptide and phospholipid-protein conjugates has provided novel scientifically and clinically important information on the phospholipid-antibody binding process.

[0065] These key studies can be performed in two tests. First, we will test the phospholipid-peptide and phospholipid-protein constructs of the present invention in a homogeneous assay using fluorescent labeling. We will also use enzyme-linked immunosorbent assay (ELISA) for targeting specific autoantibodies by the prepared bioconjugates. Herein, a secondary HPR-labeled antibody will signal interaction between the synthetic phospholipid-peptide and phospholipid-protein antigens and the antibodies in serum samples of patients diagnosed with APS and other autoimmune diseases (see e.g. FIG. 1b and FIG. 3).

Example 3Synthesis and Stability Studies of Protein-Cardiolipin Bioconjugates

[0066] Reagents and Solvents.

[0067] Cardiolipin sodium salt from bovine heart (Sigma-Aldrich); synthetic cardiolipin (Avanti Polar Lipids); phosphatidylcholine from egg (PC L--isomer; Avanti Polar Lipids); phosphatidylcholine synthetic (18:1 9-cis-isomer; Avanti Polar Lipids); KMnO.sub.4, NaIO.sub.4, Na.sub.2SO.sub.3, albumin from bovine serum (BSA), lysozime, myoglobin, hemocyanin from Megathura crenula (KLH), N-hydroxysuccinimide (NHS; all from Sigma-Aldrich), 2-glycoprotein I human, non-recombinant (Diarect antigens), N,N-diisopropylcarbodiimide (DIC; Fluka), DMSO (Sigma-Aldrich), t-butanol (Sigma-Aldrich), sodium dodecyl sulfate (SDS; Sigma-Aldrich).

[0068] Methods.

[0069] Reactions were carried out in 1.5 mL plastic eppendorf tubes; mixing was performed using Eppendorf thermomixer 5436 shaker, evaporation was performed using rotary centrifuge (HETOVAC VR-1).

[0070] Analysis. Lipid molecules were analyzed using MALDI HRMS (Bruker microflex). Protein-lipid conjugates were analyzed by LC-MS (Bruker).

##STR00015##

[0071] Synthesis of Oxidized Cardiolipin Cl.sub.ox (2)Step i

[0072] To solution of cardiolipin 1 (20 mg; natural (n) or synthetic (s)) in 2 mL t-BuOH 20 mg NaHCO.sub.3 in 100 mkl MQ water was added and mixed for 5 min. Afterwards 12 mg NaIO.sub.4 in 120 mkl water, 15 mg KMnO.sub.4 in 400 mkl water and 500 mkl t-BuOH were subsequently added. The mixture was kept at room temperature for 2 hours, then 150 mg Na.sub.2SO.sub.3 was added and mixed for additional 10 min. The resulting mixture was centrifuged at 12000 rpm for 5 min, and upper phase was taken with a pipette. A t-BuOH phase was washed with 5% aq. HCl in saturated aq. NaCl solution, evaporated to dryness on a speedvac and redissolved in DMSO.

[0073] In parallel, oxidation of cardiolipin was performed as described above in presence of 16 mg of phosphatidylcholine (Scheme V; PC natural (n) or PC synthetic (s)), which is known to prevent oxidation of cardiolipin upon exposure to air. Thus, the oxidation experiments gave four lipid reagents for further attachment to protein cargos: CLox(n), CLox(s), CLox(n)-PCox(n) and CLox(s)-PCox(s). The oxidation was confirmed by HRMS MALDI spectra.

##STR00016##

[0074] Synthesis of Activated Ester of Cardiollpin CL-NHS (3)Step ii

[0075] To a solution of oxidized lipid reagent 2 (10 mg of cardiolipin) in 1 mL DMSO 5 mg of NHS and 10 mkl of N,N-diisopropylcarbodiimide were added. Reaction mixture was stirred till the lipids were completely dissolved (approx. 2 hours). The reaction was left at room temperature overnight. Resulting solution was used in the next step without purification. Storage of CL-NHS: at 70 C.

[0076] Synthesis of Protein-Cardiolipin Conjugates (CL-Protein)

[0077] To a solution of 1 mg of protein in 100 mkl of 0.2 M carbonate buffer (pH 8.5) 875 mkl DMSO and 15 mkl of a solution of CL-NHS 3 (10 mg/mL) were added. Reaction was mixed for 20 min and then left overnight. Resulting reagent is useful for diagnostics according to literature without purification. We evaluated stability of the conjugates in the mixtures described below after 4 months of storage at +4 C. and after 1 year storage at 20 C. by LC-MS (Table 1). Natural lipid components were previously applied by other groups. As one can see, these reagents are rapidly degraded during storage, whereas using synthetic lipids and developed storage conditions (especially 3 and 4) the conjugates were stable up to 1 year at 20 C. without need in de-aeration. The best results of stability to degradation upon storage were shown by the novel conjugates 8, 10, 16 and 18 prepared herein. Stabilities are given in table 1 below.

TABLE-US-00001 TABLE 1 Composition and stability of the lipid-protein conjugate prepared in this study* Stability, % Re- Origin of 4 months 1 year agent Lipid Protein lipid Storage mixture no: No. Package component Component sample 1 2 3 4 1 2 3 4 1 Plastic vial CL(n) Bovine serum Natural 40 40 50 53 32 30 42 43 2 Plastic vial CL(s) albumin (BSA) Synthetic 66 68 70 66 64 58 63 66 3 Plastic vial CL(n)-PC(n) Natural 34 35 54 57 31 33 42 47 4 Plastic vial CL(s)-PC(s) Synthetic 64 65 60 78 54 50 64 68 5 Plastic vial CL(n) Hemocyanin Natural 32 37 52 53 37 30 45 50 6 Plastic vial CL(s) Synthetic 40 38 73 78 38 42 85 79 7 Plastic vial CL(n)-PC(n) Natural 38 40 44 54 38 40 44 54 8 Plastic vial CL(s)-PC(s) Synthetic 44 50 65 84 44 50 65 84 7 Plastic vial CL(n) Apolipoprotein Natural 45 44 50 56 46 44 53 54 8 Plastic vial CL(s) H. human Synthetic 65 60 75 84 65 60 77 89 9 Plastic vial CL(n)-PC(n) Natural 55 59 46 57 50 49 48 60 10 CL(s)-PC(s) Synthetic 70 64 86 95 64 66 90 92 11 Plastic vial CL(n) Lysozime Natural 34 31 55 45 34 22 36 40 12 Plastic vial CL(s) Synthetic 45 61 74 45 44 45 70 56 13 Plastic vial CL(n)-PC(n) Natural 32 36 65 32 30 30 66 45 14 Plastic vial CL(s)-PC(s) Synthetic 55 68 74 55 50 47 70 53 15 Plastic vial CL(n) Myoglobin Natural 34 30 45 42 33 45 67 70 16 Plastic vial CL(s) Synthetic 60 49 90 95 52 45 88 94 17 Plastic vial CL(n)-PC(n) Natural 40 39 51 50 44 34 75 75 18 Plastic vial CL(s)-PC(s) Synthetic 63 46 88 92 50 41 80 93 *Stability was evaluated as a ratio of corresponding LC peak area of the conjugate with a right averaged mass to the one of degraded products, in %.

[0078] Storage and Analysis of Cardiolipin-Protein Conjugates

[0079] 1. Solution of conjugate 1-18 in DMSO-0.2 M carbonate buffer (concentration10 mg/mL).

[0080] 2. Solution of conjugate 1-18 in DMSO-0.2 M carbonate buffer (concentration10 mg/mL) with additional 1-thioglycerine (final concentration1 mM).

[0081] 3. Solution of 1-18 in DMSO-0.2 M carbonate buffer (concentration10 mg/mL) with additional 1-thioglycerine (final concentration5 mM).

[0082] 4. Solution of 1-18 in DMSO-0.2 M carbonate buffer (concentration10 mg/mL) with additional 1-thioglycerine (final concentration5 mM) and SDS (final concentration0.01%).

[0083] Results are seen in table 1 above.

Example 4Synthetic Approach to Protein Phospholipid Conjugates Using Triazole-Amide Linkers

[0084] In this example a new CuAAC click approach for the attachment of synthetic PE (phosphoethanolamine) antigen to clinically relevant 2GPI and PT is demonstrated. In spite of structural differences, both CL and PE are immunogenic and proved to be useful in diagnostics of anti-phospholipids (aPLs). The click procedure has advantages of high yields and purity of the products with improved chemical stability compared to oxidized phospholipids. Moreover, the new procedure can be directly employed for similar quantities of different phospholipids and proteins. In order to compare with previously reported analogues, a PE-BSA conjugate was prepared as well (Scheme VI below).

##STR00017## ##STR00018## ##STR00019##

[0085] First, we functionalized each protein with water soluble alkyne activated ester alkyne group and then subjected it to CuAAC click reaction with commercially available PE azide in a molar ratio 1:25. After simple precipitation, we obtained the desired conjugated 5-7 in good yield and purity (Scheme 1; yields 24 80%; full conversion as determined by gel electrophoresis) The present inventor has demonstrated CuAAC chemistry as an improved method for preparation of novel phospholipid-protein conjugates with a high potential for the diagnostics of autoimmune diseases. The resulting phospholipid-protein conjugates show high binding affinity and specificity for the autoimmune antibodies.

[0086] FIG. 3 shows a representative scheme of enzyme-linked immunosorbent assay (ELISA) for detection of autoimmune antibodies against antigens of the present invention: phospholipids, proteins and phospholipid-protein conjugates. P=protein, L=linker, R=phospholipid residue, TMB=3,3,5,5-tetramethylbenzidine, HPR=horseradish peroxidase.

[0087] Conjugates and control antigens we used in the enzyme-linked immunosorbent assay (ELISA), applying series of disease-associated or human normal plasmas (HNP, n=10; FIG. 3, Table 2). The disease-associated samples contained high levels of antibodies against phospholipids (aPLs), a-2-GPIs; control samples to assess cross-reactivity contained autoantibodies to single-stranded and double-stranded DNA.

TABLE-US-00002 TABLE 2 Results of ELISA assay using diverse antigens: cardiolipin, prothrombin, 2-glycoprotein, phosphoethanolamine azide and synthetic conjugates 5-7 of scheme VI.* Absorbance at 450 nm: Analyte HNP Antigen a-PL a-2GPI a-ssDNA a-dsDNA (n = 10) CL 1.97 0.90 0.90 0.63 0.45 2GPI 0.75 1.01 0.60 0.32 0.22 CL:2GPI.sup. 0.98 0.80 0.55 0.61 0.52 PT 0.45 0.31 0.43 0.45 0.23 BSA 0.31 0.25 0.34 0.28 0.24 PE azide 1.03 0.91 0.80 0.54 0.33 5 (BSA-PE) 1.44 0.56 1.44 1.21 0.43 6 (PT-PE) 1.05 0.44 1.01 0.72 0.45 7 (2GPI-PE) 1.02 1.15 0.5 0.4 0.19 *a-PL, a-2GPI, a-ssDNA and a-dsDNA = human plasma tested highly positive against phospholipids; 2GPI, single-stranded and double-stranded DNA, respectively. HNP = human normal plasma; averaged absorbance for n patients is presented ( 0.20). .sup.2GPI (0.001%) was added to cardiolipin under blocking conditions resulting in non-covalent binding. CL = cardiolipin. PT = prothrombin. Each sample was measured in the duplicate with resulting deviation in absorbance 0.20.

[0088] High cross-reactivity of aPLs with other antigens such as ssDNA and dsDNA is an obstacle for their utility in studies and diagnostics of autoimmune diseases. Improved specificity of aPL binding was achieved by covalent cross-linking of biologically complementary molecules such as PE with PT and PE with 2-GPI. The IgG ELISA experiments show that this has been achieved for PE-2-GPI conjugate (absorbance 1.02 vs. 0.47-0.54 when incubated with APL, a-ssDNA and a-dsDNA, respectively) (Table 2).

[0089] The reproducibility of ELISA tests and stability of antigens upon storage in solution at 20 C. was tested (FIG. 4). The latter was done by TLC and gel electrophoresis. As appears from FIG. 4 conjugates 5-7 showed superior reproducibility than individual phospholipids and oxidized cardiolipin conjugates (97-98% vs. 83-89%, respectively). Stability upon storage in solution was increased up to 6 months at 20 C. vs. 1.5-2 months for oxidized CL analogues. This implies that high purity and immunogenicity of the novel molecules has a positive effect on their diagnostic performance, which renders them applicable for studying various autoimmune conditions.

Example 5New Strategy for the Preparation of Cardiolipin Azide Reagent and Click Conjugation to Proteins

[0090] The rationale beyond new synthesis was to use synthetic, pure and well-characterized lipids and to decrease steric hindrance that arises upon bioconjugation to proteins. The latter was done using extended linker between the cardiolipin and azide moiety (see FIG. 5).

[0091] 5.1 Synthesis of Cardiolipin-Azide 10 of FIG. 5

[0092] A solution of cardiolipin (CL) 1 (10 mg, 6.66 mol) in 1.5 mL tBuOH 100 L was mixed with NaHCO.sub.3 (10 mg in 100 L), NaIO.sub.4 (30 mg in 200 L water) and KMnO.sub.4 (10 mg in 200 L water). The reaction was flashed with argon for 3 min and afterwards kept in a microwave reactor (15V) for 1 h at 45 C. After the starting material was no longer detectable by TLC, the reaction was quenched by adding 150 mg Na.sub.2SO.sub.3. The mixture was afterwards acidified with 5% HCl to pH 3.0 and washed twice with t-BuOH. t-BuOH fraction was dried over Na.sub.2SO.sub.4 and evaporated in vacuo. Yield 86%; R.sub.f 0.37 (chloroform:methanol:water 3:1.5:0.2, v/v/v), HRMS-ESI nm/z 1406.84772 ([M+Na]+, C.sub.72H.sub.130Na.sub.2O.sub.19P.sub.2 calcd 1406.84766).

[0093] To a solution of oxidized cardiolipin (5 mg) in 1 mL DMSO 4 mg succinimide ester in 20 L DMSO and 8 L N,N-diisopropycarbodiimide were added. After keeping the reaction for 5 h at room temperature TLC showed complete conversion of the starting material. Intermediate NHS-cardiolipin was further reacted with 3-azidopropan-1-amine (1.1 eq., 0.35 mg) at room temperature in dark over 12 h. The product was used in further steps without purification. R, 0.55 (chloroform:methanol:water 3:0.5:0.2, v/v/v), HRMS-ESI m/z: 1488.91209 ([M+Na]+, C.sub.75H.sub.136N.sub.4Na.sub.2O.sub.18P.sub.2 calcd 1488.91203).

[0094] 5.2 Click Reactions

[0095] Click reactions were performed as previously described herein. The products were initially purified by gel filtration using Zeba spin desalting columns (Life Technologies), and afterwards precipitated from cold acetone (20 C.). The resulting conjugates were washed twice with cold acetone, dried in vacuo and analyzed by mass spectrometry and gel electrophoresis. Final yields of products based on the absorbance at 280 nm: 88% (6, scheme VI), 92% (7, scheme VI), 89% (7, FIG. 5), 84% (8, FIG. 5).

[0096] Chemical composition of product antigens 7-8 (FIG. 5) was verified by gel electrophoresis and MALDI MS spectra. According to MS data, cardiolipin residues were attached to beta2GPI and PT, respectively. This is significantly higher than for a previous approach using direct coupling of NHS-cardiolipin derivative to proteins (1-2 residues), and might be a result of reduced steric hindrance of the cardiolipin reagent.

Example 6Superior Specificity and Stability of New Molecules Compared to Controls

[0097] As mentioned above, low stability and high cross-reactivity are major issues of currently applied phospholipid antigens. As a result, detected antibodies are not specific to an autoimmune disease and the tests have low to no value for diagnostics. We prepared new antigens which contain covalently linked phospholipid and proteins which mimic biologically relevant autoimmune antigens. To validate their stability and specificity vs. controls, we carried out ELISA assays with well characterized human monoclonal antibodies toward HIV-1 antigens and healthy controls.

[0098] 6.1. ELISA assays, reproducibility analysis and testing the stability of antigens were carried out as described before (Molecules 2015, 20, 10253-10263). Briefly, the ELISA assay includes two incubation steps followed by enzyme-promoted signal generation. First incubation involves binding of target antibodies within the sample to immobilized antigen on the surface of a microplate. At the second step, the bound antibody interacts with the secondary anti-antibody (e.g. anti-human IgG or IgM), covalently labeled with an enzyme such as horseradish peroxidase (HRP). In the presence of H.sub.2O.sub.2 HRP catalyzes conversion of non-colored organic substrate TMB to oxidized TMB(ox), which has a distinctive absorbance signal at 450 nm (A.sub.450). Thus, by measuring A.sub.450 one can quantify titers of particular antibodies towards immobilized antigen.

[0099] Initially, background signal of each sample was tested to be below A.sub.450 0.06 using non-coated microplates upon blocking at the same conditions as for the original experiment. To determine equilibrium time for the aPLP-antigen binding, pre-coated microplates were subjected to incubation with plasma samples in dilution 1:200 over a time course of 40 min-3 hrs. Secondary incubation was always carried out for 1 h at room temperature using 1:35.000 dilution of corresponding HRP-conjugate. Results of this assay are presented in the Supporting Information.

[0100] Linear range for each antigen was determined by testing series of control dilutions (HNP, HCL in dilutions 1:50 to 1:2000). According to the results plasma dilutions 1:100-1:750 were within the linear range of assay for each antigen (R.sup.2>0.97). For the controls and patient samples, weak positive (+/) and positive (+) signal were determined as 2- and 3-fold signal, respectively, above the mean value for a healthy control cohort (n=16).

[0101] 6.2. Specificity Assessment (Compound Numbers are as in Scheme VI Above or FIG. 5 as Noted for Each Compound)

[0102] The results of IgG class ELISA for novel antigens and controls obtained using monoclonal antibodies towards HIV-1 antigens and healthy sera samples are shown in Table 1. We observed the highest level of unspecific binding for non-covalent complexes of PLs and proteins 1+8 and 1+3 (scheme VI, mixed antigen control). In contrast, cross-linked conjugates (6-7, scheme VI) and (7-8 FIG. 5) showed only low to no non-specific binding in similar experiments (for example, Table 3, data for (8, FIG. 5) compared to 1+8 and 1+3, Scheme VI. Conjugates 6-7 (Scheme VI) and demonstrated a weak binding to anti-gp41, whereas binding to anti-p24 and B12 antibodies completely disappeared upon PL-protein conjugation. High reactivity on HIV-1 specific monoclonal antibodies and healthy control sera observed for mixed PLP controls could result from the high heterogeneity of non-covalent structures leading to unspecific interactions with polyclonal antibodies. In turn, cross-reactivity of phosphoethanolamine containing antigens 6-7 (scheme VI) might be caused by the presence of PEG linker. Alternative linker design for the attachment of phosphoethanolamine to proteins is an objective of our on-going research and will be published in the near future. Finally, the most specific antigens 7-8 of FIG. 5) were further applied in the studies of human samples described below.

TABLE-US-00003 TABLE 3 Analysis of binding specificity for cross-linked antigens and controls.numbering as in Scheme VI unless otherwise stated.* Binding human Response in monoclonal antibodies: healthy controls Antigen nr./details* a-p24 a-gp41 B12 (% of patients) 6/PE.sup.PEG-PT + +/(12%) 7/PE.sup.PEG-2GPI +/ +(6%), +/(12%) 7, FIG. 5)/CL-PT +/ +/(6%) 8, FIG. 5)/CL-2GPI .sup.(100%) 8/2GPI +/ +/(25%) 3/PT +/ +/ +/(25%) 1 + 8/CL + -2GPI +/ +/ +/ +(44%), +/(25%) 1 + 3/CL + PT +/ +/ + +(25%), +/(25%) 8 + RN.sub.3/PE.sup.PEG + 2GPI +/ +/ + +(25%), +/(12%) 3 + RN.sub.3/PE.sup.PEG + 2GPI +/ +/ + +(25%), +/(25%) *Weak positive (+/) and positive (+) signal were defined as 2- and 3-fold absorbance signals, respectively, above the mean value for a healthy control group. PE.sup.PEG, PT, CL, 2GPI are PEGylated phosphoethanolamine, prothrombin, cardiolipin and 2-glycoprotein I, respectively. Covalent cross-linking and non-covalent mixing of the antigens is defined as hyphen and plus, respectively.

[0103] 6.3. Reproducibility and Chemical Stability of Antigens 7-8 of FIG. 5 upon Diverse Storage Conditions

[0104] Experiments were carried out as described herein and showed that ELISA assays using 7-8 of FIG. 5 were highly reproducible with absorbance values varying within a range of 0.5-0.8 of mean titer for each independent analysis. Antigens 7-8 and cardiolipin were also analyzed by gel electrophoresis, ELISA and MALDI MS after storage in dark glass vials at 20 C.-+4 C., and being immobilized on microtiter plates at +4 C. Antigens 7-8 were stable for up to 3 months at +4 C. in glass vials and on the surface of microtiter plates, and up to 1 year in glass vials at 20 C. Remarkably, cardiolipin which is used in current kits degraded after approx. 3 weeks of storage at +4 C. in a glass vial or on the microtiter plate, and after 3 months of storage in a vial at 20 C.

Example 7Autoantibody Detection in Patient Samples Using New Antigens

[0105] The initial ELISA assay was done manually using microtiter plates coated with antigens 6-7 of Scheme VI and 7-8 of FIG. 5 or control antigens. First, pre-coated microplates were analyzed using commercially available polyclonal controls (human normal plasma HNP and human plasma containing antibodies to cardiolipin, HCL, Immunovision). Plasma titration experiments showed low response of IgM antibodies to the dilution, whereas IgG antibodies showed reduced signal upon dilution with a high degree of linearity (R2>0.95; Supporting Information). Therefore, in this work we focused on the IgG assay. In addition, repeatability of assays and PLP stability were studied as described before and showed superior levels compared to controls in agreement with our previous data.

[0106] Next, studies of the SUH cohort (27 patients diagnosed with pediatric SLE) under the developed ELISA conditions identified five and seven aPLP positive patients (19% and 26%) when antigens 7 and 8 (FIG. 5) were applied. This was in contrast with nine and three positive patients (33% and 11%) detected using beta2GPI and mixed antigen cardiolipin+beta2GPI, respectively. Robustness and clinical relevance of the new antigens was further evaluated by the analysis of blood samples from adult patients obtained from Odense University Hospital, Denmark (OUH). The assay was performed using similar plate coating, incubation conditions, controls and cut-off values as for SUH cohort, although in a fully automated regimen. The results were compared to commercial anti-cardiolipin and anti-beta2GPI tests (Euro Diagnostica), which were performed for OUH patients under similar conditions.

[0107] Summarized ELISA results for polyclonal control antibodies, healthy control sera and two patient cohorts are shown in FIG. 6 (5 and 6 correspond to compound 6 and 7 respectively of Scheme VI). Negative and positive controls (HNP and HCL, Immunovision) are shown as a x and a star, respectively. As can be seen, the background signal of healthy controls was highest for cardiolipin 1, PT and mixture of phosphoethanolamine with PT. SUH samples showed higher aPLP titers than the OUH cohort, especially when antigen 8, FIG. 5 and mixed antigen cardiolipin+beta2GPI were applied. Notably, aPLP positive samples indicated by compounds 7-8, FIG. 5 dramatically varied from those indicated by separated antigens. Group analysis using ANOVA confirmed a statistically significant difference between the groups within SUH and OUH cohorts with a p value of 0.01.

[0108] Finally, we analyzed medical records for the patients with respect to medication use. The data were divided into groups containing one or more medications and subjected to statistical analysis in Stata. Using two independent tests (Student's t-test and Fisher test), we confirmed no correlation between a particular medication or a group of medications and elevated aPLP titers.

Example 8Correlation Between Disease Activity and aPLP Titers for Antigen 8 of FIG. 5

[0109] For statistical analysis, the following clinical parameters of each patient were considered: age (at onset and at sample date), gender, ethnicity, race, diagnosis, disease activity (at onset and at sample date), clinical manifestations, patient's complaints, and treatment history. Differences were analyzed for the antigens and controls reported in Table 1. using Student's t-test and ordinary least squares (OLS) analysis in Stata. Groups were compared for difference in means of antibody titers using ANOVA. A p-value of less than 0.05 was considered significant for each correlation.

[0110] Students's t-test indicated a correlation between Smith positivity and elevated aPLP titers using exclusively antigen 8 (FIG. 7).

[0111] To further evaluate statistical difference between observed clinical parameters and aPLP titers, we performed ordinary least squares (OLS) analysis using Stata. The difference in aPLP titers for Smith negative and Smith positive groups was found to be statistically significant with p value of 0.002 (FIG. 8).

[0112] Notably, no correlation with clinical parameters was observed for antigen 7 or other antigens and controls (p>0.09). This implies uniqueness of the properties by antigen 8 and its potential in prognose of highly active autoimmune conditions which is indicated by Smith positivity.

CONCLUSIONS

[0113] In this work, we introduce novel synthetic phospholipid-protein antigens, cardiolipin-beta2GPI and cardiolipin-PT, and prove their utility in diagnostics of autoantibodies. Novel antigens reported herein are prepared by convenient CuAAC click chemistry approach in high yields and purity (example 5). Tests of specificity against human monoclonal antibodies prove our hypothesis of improved specificity as a result of conjugating biologically complementary molecules in a regioselective fashion (example 6). Next, we show the utility of the prepared PLPs in enzyme-linked immunosorbent assay (ELISA) using patient samples and healthy controls (example 7). Finally, we prove correlation of the observed aPLPs with multiple clinical parameters and verify the causal effect of high disease activity on the elevated aPLP titers (example 8).

[0114] Our findings indicate higher prognostic potential of cross-linked PLPs than a simple mixture of its components or separate tests on each of them which are currently available on the market [37]. Moreover, the prepared antigens 7-8 (of FIG. 5) showed higher reproducibility and stability than currently available PL antigens and have advantage of straightforward synthetic route, high chemical homogeneity and decreased cross-reactivity on nonspecific human antibodies (Table 3).

REFERENCES

[0115] M. G. Paulick, A. R. Wise, M. B. Forstner, J. T. Groves, C. R. Bertozzi. J. Am. Chem. Soc. 2007, 129, 11543. [0116] WO 2012/134925 A1 [0117] H. C. Hang, J. P. Wilson, G. Charron. Acc. Chem. Res. 2011, 44, 699. [0118] J. A. Precher, C. R. Bertozzy. Nat. Chem. Biol. 2005, 1, 13. [0119] M. Chassignol, Y. Aubert, V. Roig, U. Asseline. Nucleos. Nucleot. Nucl. Acids 2007, 26, 1669. [0120] A. M. Rouquette, C. Desgruelles. Lupus 2006, 15, 403, and references cited therein. [0121] A. Castro. H. Wang. WO 2007/061793 A2, PCT/US2006/044572. [0122] J. S. Yadav et al. Tetrahedron Lett. 1996, 37, 6603; [0123] K. Kasireddy et al. Bioorg. Chem. 2005, 33, 345. [0124] Molecules 2015, 20, 10253-10263