WO2008111855A1

WO2008111855A1 - Biosensor, surface coating and assay

Info

Publication number: WO2008111855A1
Application number: PCT/NZ2008/000039
Authority: WO
Inventors: John Stanton Mitchell
Original assignee: The New Zealand Institute For Plant And Food Research Limited
Priority date: 2007-03-14
Filing date: 2008-03-03
Publication date: 2008-09-18

Abstract

A biosensor having a ligand binding surface that is a polyalkene having a thickness of less than 100nm and bearing carboxy groups at the surface is disclosed. A ligand may be bound using the carboxy group and the concentration of the ligand or its binding partner measured in a sample, for example using surface plasmon resonance.

Description

BIOSENSOR, SURFACE COATING AND ASSAY

Field of the Invention

The present invention provides a surface coating that can be applied to surfaces for the attachment of molecules to be used in assays such as surface plasmon resonance assays.

Background

Surface plasmon resonance (SPR) is a technology that can be used for monitoring biomolecular reactions and for assays. For a review, see Silin and Plant Trends in Biotechnology 15, 353-359 (1997). The detection of binding relies on an electron charged density wave phenomenon at the surface of a metallic film when light is refracted at the film under particular conditions. The resonance is the result of energy momentum being transformed from incident photons into surface plasmons. Commercial systems are available from Biacore AB and ICX Nomadics.

In SPR, a key component is the sensor. An immobilised molecule on to the sensor surface binds its binding partner, when present. This binding is detected as a drop in the intensity of reflected polarised light at a specific angle of incidence. Particularly, SPR provides a mass detector, detecting binding to the molecule immobilised on the surface. The sensor is generally on a chip. The chip generally consists of a glass surface coated with a thin layer of gold. In the Biacore chips, the gold surface is coated with a carboxymethylated dextran layer. This layer can be linked to and immobilised on the surface. Carboxy groups on the exposed surface can be used to bind ligands. Frequently, the ligand is an antigen (or hapten) and the assay involves the detection of antibody that attaches to the immobilised ligand.

Although the carboxymethylated dextrans have proved to be useful as surface coatings in SPR, they suffer from several disadvantages. Their chemical stability in the face of severe regeneration conditions is less than ideal. They do not provide a thin film polymer coating allowing binding very close to the sensor surface. It is difficult to control the number of functional groups and thus the degree of immobilisation. They are not simple to manufacture.

It is an object of the present invention to provide a biosensor having an improved surface for binding of molecules for use in ligand binding assays, or at least to provide the public with a useful choice. Disclosure of the Invention

The invention provides a biosensor having a ligand binding surface that is a polyalkene having a thickness of less than 100 nm and bearing carboxy groups at the surface.

Preferably, the polyalkene is prepared by spin-coating the surface with a polymerisation mixture comprising at least one alkene bearing a carboxy group. Preferably the mixture comprises at least one compound having two or more alkene groups.

The polyalkene may be prepared by known methods. The polymerisation mixture generally contains a monoalkene and a dialkene and a solvent. At least one of the alkenes bears a carboxy group (for example methacrylic acid) or a group that can be converted to a carboxy group (for example allylamine, a monomer that can be modified to add a carboxylic acid group through a linker. The polymerisation mixture preferably contains an initiator to initiate polymerisation. Preferred initiators include IRGAcure 2022 (bisacylphosphine photoinitiator), benzoin butyl ether (TRIGONAL 15), benzoin, methylbenzoylformate (VICURE 55) and all other benzoin ethers, benzilketals, α-dialkoxyacetophenones, α-hydroxyalkylphenones. Generally the polymerisation is carried out on the surface under nitrogen. UV light may be directed onto the coated surface to promote initiation.

The solvents for the polymerisation mixture may be any suitable solvent. The preferred solvents are polar solvents. Preferably the alkenes and any initiator dissolve readily in the solvent. Preferred solvents include water, ethanol, methanol, dimethylformamide, dimethyl sulphoxide, methylethylketone, acetone, cyclohexane, toluene, dimethoxyethane, dioxane, triethylene glycol dimethyl ether and trichloroethane and mixtures of these. Methanol-water and ethanol-water mixtures and water and triethylene glycol dimethyl ether are preferred solvents.

Preferably, the thickness of the biosensor ligand binding surface is less than 50 nm, more preferably less than 25 nm, most preferably less than 10 nm.

Preferably, the polyalkene is a copolymer of CHR_I=CR₂CO-O-A-CR₃=CHR₄ (Formula I) and CH(Rs)=C(Re)-COOR₇ (Formula II) wherein R,, R₂, R₃, R₄, R₅, and R₆ are independently selected from the group consisting of hydrogen, methyl, and ethyl, and A is an aliphatic chain comprising 1-20 carbon atoms and 0-6 oxygen atoms in the chain wherein any oxygen atoms are not adjacent to each other, and R₇ is hydrogen or an aliphatic chain with 2-20 carbon atoms, bearing a carboxy group, and optionally comprises one or more ester groups or ether groups within the chain.

In preferred embodiments Formula I has the structure CHR₁=CR₂CO-O-A-O-COCR₃=CHR₄.

Preferred polymers include copolymers formed from a subset of molecules of Formula I having the formula CH₂=CR₂-CO-O-(CH₂-CH₂-O)_n-COCR₃=CH₂. A particularly preferred molecule of Formula I is ethylene glycol dimethacrylate.

Preferred polymers include copolymers formed from molecules of Formula II in which R^ is methyl and -R₇ is -(d-C₅-alkyl-O)_n-CO-CH=CH-COOH wherein n=l-10 or -R₇ is -C,_-20- alkyl-CO-CH=CH-COOH.

Especially preferred R₇ groups include those comprising C]_-I0 alkyl or (CH₂CH₂O)_n wherein n=l-4.

Particularly preferred polymers include copolymers where the compound of Formula II is methyacrylic acid or mono-2-(methacryloyloxy)ethyl maleate.

Especially preferred polymers are copolymers of ethylene glycol dimethacrylate and methacrylic acid, and of ethylene glycol dimethacrylate and mono-2-(methacryloyloxy)ethyl maleate. Propenethiol may also be usefully included in either combination. Also especially preferred are various oligoethylene glycol dimethacrylates copolymerised with various lengths of chain in a mono-2-(methacryloyloxy)oligoethyleneglycol maleate.

In another aspect the invention comprises a biosensor having a surface that is a polyalkene having a thickness of less than 100 nm and bearing a ligand at the surface, the ligand being linked to the surface in a group formed from a carboxy group carried by the polyalkene and a group on or linked to the ligand.

In a preferred embodiment the biosensor will be a surface plasmon resonance chip. The coating is applied to a metal coating on a surface of the chip. The metal is usually gold, silver, copper, aluminium, or platinum, especially gold. In a preferred embodiment, the metal is present in a thin layer on an optical chip suitable for use as a surface plasmon resonance biosensor.

The preferred polyalkenes for use in this aspect of the invention are as described for the first aspect of the invention.

The ligand may be fixed to the surface by one or more reactions with a carboxy group on the surface. The reaction or reactions result in an amine, alcohol or thiol of the ligand being fixed to the surface by an amide, ester or thioester link. The ligand is typically a molecule that has a larger binding partner, for example a hapten to which the larger binding partner is capable of binding. The biosensor ligand binding surface may be applied as a coating to an optical chip. The biosensor may be for use in surface plasmon resonance.

A molecule capable of binding a larger binding partner is attached through either a group on or attached to the ligand to the carboxy group of the surface. The carboxy group is often derivatised to form a reactive carboxylating agent to react with the ligand. The chip bearing ligand on its surface can then be used in a conventional instrument for measuring surface plasmon resonance.

It is desirable that the polyalkene coating is thin. If the thickness exceeds 500 nm, the surface plasmon phenomenon cannot be sufficiently detected.

For a flat surface giving good results on surface plasmon resonance, a film of less than 25 nm, preferably less than 10 nm, is preferred. It is preferable that the surface is not hydrophobic. The inclusion of oxyethylene groups in the monomers used to form the film assists in keeping the film hydrophilic. That hydrophilicity assists in the use of the films with aqueous biological samples.

The metal film may be formed using a conventional method, for example, by sputtering, evaporation, ion-plating, or electro-plating. The metal film is applied to a substrate which is generally an optical glass. Other materials transparent to laser beams such as polymethyl methacrylate, polyethylene terephthalate, and polycarbonate can also be used. The surface plasmon resonance phenomenon occurs due to the fact that the intensity of monochromatic light refracted from the border between an optically transparent substance such as glass and a thin metal film layer depend on the refractive index of a sample located on the outgoing side of the metal. The sample can be analysed by measuring the intensity of reflected monochromatic light.

When a binding partner binds to the ligand on the surface coating of the present invention, the refractive index at the surface changes. The binding of large molecules can be measured directly. Alternatively, the concentrations of free ligand in solution can be measured indirectly through the effect on binding the large binding partners to the surface with lowering the amount of binding of the partners to the immobilised ligands attached to the surface at high free ligand concentrations.

One advantage of the present coating is that it is relatively easy to control the surface density of carboxy groups and consequently the immobilised ligands. This can be done by varying the portion of carboxy containing molecules in the mixture used to prepare the polymer.

In another aspect the invention provides an assay wherein the concentration of a ligand is measured in an assay comprising contacting the ligand and a binding partner in a sample with a biosensor having a surface according to the invention having an immobilised ligand and binding of the binding partner to the ligand is measured to allow calculation of the amount of ligand in the sample. Preferably binding of the binding partner to the immobilised ligand is measured by surface plasmon resonance. The ligand may, for example, be a steroid, for example Cortisol immobilised through its 4-position. Generally the amount of the binding partner in this type of assay is known or fixed.

In another aspect of the invention, there is provided a method of assay of a binding partner of a ligand in a sample, comprising contacting the binding partner in a sample with a biosensor having a surface of the invention having immobilised ligand bound to it, and determining the amount of bound ligand, preferably by surface plasmon resonance.

In other aspects the invention may be used for measuring aspects of biomolecular interactions such as kinetics and binding affinities.

In preferred assays the methods are flow through assays where the sample flows past the surface bearing the immobilised ligand. The assay is preferably calibrated using samples comprising known amounts of the analyte. BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows a Spreeta chip in a spin coater mount.

Figure 2 shows a plot of Response (RU) vs. secondary antibody IgG concentration (μg/mL).

Figure 3 shows a rnAb plot of Response (RU) vs. mAb concentration (ng/mL) - with enhancement.

Figure 4 shows a mAb plot of Response (RU) vs. mAb concentration (ng/mL) for replicate two - with enhancement.

Figure 5 shows a mAb plot of Response (RU) vs. mAb concentration (ng/mL) - no enhancement.

Figure 6 shows a Cortisol Buffer assay standard curve - Enhanced with secondary antibody.

Figure 7 shows a mAb Only Assay Curve.

Figure 8 shows comparison plots of total (♦ ) and mAb only (■) responses with non-specific binding subtracted.

Figure 9 shows assay curves for Cortisol using mAb only response (top) and secondary antibody enhanced response (bottom).

Figure 10 shows a response curve for a Cortisol saliva assay with the non-specific binding subtracted.

Figure 11 shows an assay curve for testosterone with subtraction of non-specific binding.

Figure 12 shows a testosterone gold-enhanced assay curve: without subtraction (top) and with subtraction (bottom).

The invention is illustrated by the following non-limiting examples. EXAMPLES

Example 1 - Synthesis of Cortisol-OEG-NH2

1.1 General Experimental Details

The reactions were all performed using analytical grade solvent. The DMF was dried over magnesium sulfate and then stored over molecular sieves. The chloroform and triethylamine were dried over molecular sieves. The Cortisol (Q 3880-000) and testosterone (A6950-000) were obtained from Steraloids (Newport, RI, USA). The di-tertbutyl-dicarbonate and 4,7, 10-triox- 1,13-tridecanediamine were purchased from Fluka Chemie (Buchs, Germany) and triethylamine from BDH (Poole, UK). All other reagents were purchased from Aldrich Chemical Company (Milwaukee, WI, USA) and were used without further purification.

Melting point determinations were done using a Reichert Thermopan instrument. Proton and carbon NMR spectra were acquired using both a 300 MHz Bruker AC300 FT-NMR and a 400 MHz Bruker Avance DRX400 FT-NMR. Chemical shifts (δ) are reported in parts per million (ppm) and coupling constants (J) are reported in Hz. Electrospray mass spectra were acquired using a VG Platform II Electrospray Mass Spectrometer (ES-MS) instrument.

All HPLC analysis was performed by injection of an ~1 mg/ml sample of the compound in question run according to the conditions detailed below for each compound. The injection volume was the maximum that the column could withstand without overloading. The main peak was collected entirely and the volume collected calculated from the times the peak started and ended and the flow rate. This sample was then placed in an HPLC vial to act as the pure standard. The starting ~1 mg/ml solution was diluted to the same concentration as the standard and used as the raw sample. Three replicate injections of 100 μl each of both standard and raw sample were made. The peak integrals were calculated using the computer program and averaged. The percentage purity was determined by:

% Purity = mean raw sample area x 100

mean standard area The column used was a Sphereclone 3 μm C8 15O x 4.6mm and the oven temperature was 4O⁰C and the flow rate lml/min unless otherwise stated.

1.2 Mono-Boc OEG

(Ditert-butyl-dicarbonate- 1 -4,7, 10-trioxa- 1 , 13-tridecane- 1 -amine)

4,7,10-trioxa-l,13-tridecanediamine (PEG-short) (1.9 g) was dissolved in 20 ml of molecular sieves dried methanol in a round bottom flask (50 ml). Di-tert-butyldicarbonate (DBDC) (1 g) was dissolved in dry methanol (10 ml). Triethylamine (1 ml) was added to the vigorously stirred PEG-short solution under septa. The DBDC solution was then added drop- wise under septa to the vigorously stirred solution via syringe. The solution was stirred overnight. The solvent was removed and the sample dried in vacuo. The resultant oil was then column separated by flash chromatography using 4:1 CH₂Cl₂: MeOH, 16:4:0.5 CH₂Cl₂: MeOH: AcOH, 16:4:1 CH₂Cl₂: MeOH: AcOH to yield a clear, colourless oil. Yield: 1.9252g (69%). ES-MS m/z (MeOH): 322 [M+H]⁺.

1.3 Cortisol Epoxide

4,5-Epoxy-( 11 β)- 1 , 17,21 -trihydroxypregna-3 ,20-dione

Cortisol (362.5 mg, 1.0 mmol) was partially dissolved in methanol (13 ml) and ethanol (5 ml) and chilled to O⁰C. Sodium hydroxide solution (10%w/v in distilled water, 1 ml) was added followed by 30% hydrogen peroxide solution (400 μl). The reaction was kept stirring at O⁰C on ice for three hours. The reaction mixture was then raised to room temperature; any remaining solid was filtered off using a sintered glass funnel. The filtrate pH was carefully adjusted to 7.0 using acetic acid and the resulting solution dried in vacuo to yield a clear, colourless oil. This sample was then constituted in distilled water (30 ml) and extracted with 3 x 30ml of ethyl acetate. The organic phase was then washed with 1 x 30 ml of distilled water and the organic phase dried over sodium sulfate. The supernatant was then passed through a bed of calcined alumina (~10g) and the solvent removed and sample dried in vacuo to yield a clear, colourless oil. The product was then column separated using 1 :1 ethyl acetate: M-hexane to yield an analytical sample. Yield: 86.6mg (23%). R_f = 0.36 (1 :1 ethyl acetate: «-hexane). IR (KBr disc): 1450, 1701, 1724, 2369, 2928, 3449cm^"1. ¹H NMR (δ): 1.14 (3H, s, 18-CH₃), 1.36 (3H, s, 19- CH₃), 3.03 and 3.06 (IH, s, 4-H, β and α respectively), 4.30, 4.40 (IH each, d, J- 3.7Hz, 21-H). ¹³C NMR (δ): 15.9 (18-CH₃), 20.0, 21.1, 22.2, 25.8, 28.3, 28.6, 29.0, 29.4, 30.4 (19-CH₃), 32.9, 35.2, 35.3, 40.6, 52.2, 62.8, 62.9, 68.0, 68.6, 206.5, 218.9. ESMS (-40V, MeOH): 363.2 [M+H₂O-H]\ Melting point: 157-160 ⁰C β epimer. 166-169⁰C α epimer. Lit. mp: β 147-148 ⁰C, α 167-168 ⁰C. HPLC: lml/min. 60% MeOH, 100% purity, R, = 4.60 and 4.85min. for the two epimers, λ_max = 204nm.

1.4 Cortisol Acid

(1 lβ)-l l,17,21-Trihydroxypregn-4-ene-3,20-dione-4-ylthio)propanoic acid

Cortisol epoxide (586.8 mg, 1.559 mmol) was dissolved in ethanol (dried over molecular sieves, 5ml). A solution of potassium hydroxide (25%w/v in distilled water, 730 μl) was added to a small flask and stirred whilst 3-mercaptopropionic acid (224 μl) was added. The stirring solution then had the epoxide solution added dropwise and was immediately placed under nitrogen and stirred at room temperature for four hours. Distilled water (30 ml) was added. The aqueous phase was then extracted with diethyl ether (3 x 30 ml) before adjusting the pH of the aqueous phase to 1.5 with IM HCl. The aqueous phase was then extracted with 3 x 30 ml of ethyl acetate. The organic phase was then dried over sodium sulfate and the liquor decanted and solvent removed and sample dried in vacuo. The sample was then column separated using chloroform, 15:1 chloroform: methanol and methanol eluent. The sample was then dried to yield a clear, colourless oil. Yield: 479.9 mg (66%). R_f = 0.42 (5:1 chloroform: methanol). IR (neat): 1108, 1657, 2360, 2920.¹H NMR: δ 0.89 (3H, s, 18-CH₃), 1.21 (IH, t,J= 7.0Hz), 1.47 (3H, s, 19-CH₃), 2.47 (2H, t,J= 7.0Hz, CH₂-COOH), 2.84 (2H, t,J= 7.1Hz, S-CH₂), 3.66 (IH, q,J= 7.0Hz), 4.28 (IH, d,J= 19.4Hz, 21-H), 4.66 (IH, d, J= 19.4Hz, 21-H). ¹³C NMR: δ, 21.4, 22.1,

26.0, 26.2 (S-CH₂), 33.1 (19-CH₃), 35.4, 38.1, 38.4, 39.5, 46.3, 51.7, 53.3, 54.1, 56.1, 60.2, 71.1,

72.1, 93.2, 130.5, 179.6 (carboxylic acid), 182.9 (17-C), 200.8 (20-carbonyl), 216.9 (3-carbonyl). ES-MS (40V, MeOH): 466.1 [M+H]⁺, 488.0 [M+Na]⁺. Mp: 132-136 ⁰C. Lit. mp: 177-178 ⁰C. HPLC: 1 ml/min. 60%v/v methanol. R, = 1.95min. % Purity = 100%.

1.5 Cortisol Ester 2,5-Dioxo-l-pyrrolidinyl (1 lβ)-l l,17,21-trihydroxypregn-4-ene-3,20-dione-4-ylthio)propanoate

Cortisol acid (479.9 mg, 1.029 mmol) was dissolved in dry DMF (4 ml, dried over molecular sieves) and DCC (275.9mg, 1.337mmol, in 1 ml dry DMF) was added dropwise to the stirring steroid solution. This was followed by NHS (153.9 mg, 1.337 mmol, in 1 ml dry DMF) also added dropwise. The reaction was stirred overnight at room temperature in the dark. The white solid formed was then filtered off and washed with dry DMF and the filtrate solvent removed in vacuo. The sample was then column separated using chloroform, 15:1 chloroform: methanol, 10:1 chloroform: methanol to yield a pale yellow semi-solid. Yield: 486.9 mg (84%). R_f = 0.69 (5:1 chloroform: methanol). IR (KBr disc): 1078, 1655, 1736, 2928 cm^"1. ¹H NMR: δ 0.90 (3H, s, 18-CH₃), 1.50 (19-CH₃), 2.64 (2H, t, J= 6.8Hz), 2.83 (2H, t, J= 6.5Hz), 2.88 (4H, d, J= 1.2 Hz, NHS protons), 4.29 (IH, s, broad, 21 -H). ¹³C NMR: δ 16.9 (18-CH₃), 21.8, 23.8, 25.1, 25.8 (S-CH₂), 28.1, 30.6, 31.9, 33.1 (19-CH₃), 33.7, 34.0, 34.4, 39.4, 42.3, 47.7, 48.7, 52.0, 56.4, 68.0, 89.6, 125.6, 158.4, 167.7, 171.0, 179.6 (17-C), 196.4 (20-carbonyl), 206.8 (3-carbonyl). ES-MS: (40V, MeOH) 695.7 [M+ DMF + 2H₂O + Na]⁺. Mp: 139-142 ⁰C. HPLC: 30% methanol, R₁ = 1.86 min, % Purity = 90%.

1.6 Cortisol - OEG - Boc tert-butyl 3-(2-(2-(3-(3-((l lβ)-l 1,17,21 -trihydroxypregn-4-ene-3,20-dione-4- ylthio)propanamido)propoxy)ethoxy)ethoxy)propylcarbamate

Cortisol ester (486.9 mg, 0.864 mmol) was dissolved in dry DMF (3.5 ml, dried over molecular sieves). To the stirring steroid solution, was added mono-Boc OEG (416.0 mg, 1.296 mmol, in 1.25 ml of dry chloroform (dried over molecular sieves)) dropwise, with an additional 2 x 250 μl of dry chloroform used to wash. The stirring solution had dry triethylamine added (750 μl, dried over molecular sieves). The reaction was then stirred at room temperature in the dark for 60 hours. After 12 hours, another 1 ml of dry DMF was added to aid solubility. The reaction was then stopped and solvent removed and sample dried in vacuo before column separation using chloroform, 15:1 chloroform: methanol and 10:1 chloroform: methanol as eluents, yielding an orange oily solid. Yield: 413.6 mg (62%). R_f = 0.32 (10:1 chloroform: methanol). IR (KBr disc) 1707, 2930, 3437 cm^"1. ¹H NMR: δ 0.90 (3H, s, 18-CH₃), 1.43 (9H, s, Boc methyls), 1.50 (3H, s, 19-CH₃), 1.71-1.78 (6H, m, 4H from 0-CH₂-CH₂-CH₂-NH, 2H from steroid fine structure), 2.60 (2H, m, CH₂-COOH), 2.82 (2H, m, CH₂-S), 3.11 (2H, t, J= 6.6Hz, CH₂-CO- NH-CH₂), 3.26 (2H, m, CH₂-NH-CO), 3.50-3.70 (14H, m, 12H from 0-CH₂, 2H from steroid fine structure). ¹³C NMR: δ 16.8 (18-CH₃), 21.5, 22.0, 25.6, 27.7, 27.9, 28.1, 28.3 and 28.6 (O- CH₂-CH₂-CH₂-NH), 29.5 (S-CH₂), 29.8 (CH₂), 30.3, 33.8 (19-CH₃), 34.5, 35.0, 37.9 (C), 42.4 (CH₂), 47.9, 48.1, 48.4, 48.6, 52.2, 52.4, 56.7, 69.0, 69.1, 69.8, 70.1 and 70.3 and 70.6 (CH₂-O), 79.0, 89.6, 126.1, 126.4, 157.3 (Boc terminal amide), 172.7 (steroid terminal amide), 178.9, 196.5 (3-carbonyl), 206.0 (20-carbonyl). ES-MS: m/z (MeOH, 40 V) 385.4 [M+2H]²⁺. Mp: 32- 33⁰C. HPLC: Purity: 99%. MeOH mobile phase, 1 ml/min. R_t = 1.92min, λ_max = 206nm.

1.7 Cortisol OEG-NH₂ N-(3 -(2-(2-(3 -aminopropoxy)ethoxy)ethoxy)propyl)-3 -(- 10,13 -dimethyl-3 -oxo-( 11 β)- 11 , 17,21 - trihydroxypregn-4-ene-3,20-dione-4-ylthio)-4-ylthio)propanamide

Cortisol-OEG-NH₂ (104.3 mg) was dissolved in formic acid (4 mL) and stirred for 4 hours at room temperature in the dark. The solvent was removed in vacuo to yield an orange oil. R_f = 0.13 (10:1 MeOH: Acetic Acid). ES-MS: (MeOH 40V + formic acid) 669.6 [M+H]⁺, 718.1 [M+H₂O+MeOH+H]⁺.

Example 2 - Polymer Synthesis for SPR Chips

2.1 Spreeta Chips

Methacrylate / ethylene glycol polymer coatings were prepared on Spreeta chips according to the following method.

A new Spreeta chip (Texas Instruments, distributed by ICX Nomadics) was removed (using an antistatic cuff) and the gold surface rinsed thoroughly with Analar ethanol. The chip (11) was then mounted in the Spreeta chip mount (12) (see Figure 1) and the mount taped with duct tape to the spin-coater chuck. In a glass vial the following were added in sequence: 1.2 mL of triglyme, 425 μL of ethylene glycol dimethacrylate (EGDMA) and 42.5 μL of methacrylic acid (MAA). The lights were extinguished and 20 μL of bis acyl phosphine initiator (IRGAcure 2022, Ciba Specialty Chemicals) added. The mixture is vortex mixed and then 19.7 μL of polymer solution applied to the flat gold surface of the chip (13). Immediately the spin coater program was started (nitrogen is supplied, compressed air is on and vacuum is on to hold the mount in place). A program allows 30 s of wait to allow shielding of the spin-coater or evacuation of the room. The spin-coater then starts spinning at 2000 rpm for 30 s before stopping. A timer was used to coordinate the UV lamp (400 W, MIFF lamp, Lamp Specialists Auckland, 20 cm from the spin-coater) such that it is switched on 3 s before the spin cycle starts. The UV lamp was then left going for a further 30 min before switching off and recovering the chip. It is important to take care when applying the polymer solution so all of it goes on the gold surface. It is also important to ensure that the chip sits flat in the mount. One Spreeta chip polymer coated using this method had a polymer thickness of 4.5 nm as determined by ellipsometry.

2.2 Polymer coating of BIA core Chips - 10:1 EGDMA : MAA The gold coated glass slides used in the construction of BIAcore chips were prepared as follows. A gold slide (SIA Kit Au BR- 1004-05) was removed from the pack with tweezers held on the edge and rinsed thoroughly with Analar ethanol. The slide was then dried under nitrogen and mounted on the chuck of the spin-coater making sure that it covers the vacuum orifice in the middle and with the gold surface facing upwards. Polymer solution as detailed above was then applied to the chip (23.0 μL) and spin-coating performed as for the Spreeta chip. On visual inspection this coating gave a good thin layer in the middle with thicker layers on the outside.

The polymer coated gold slide was then assembled in a BIAcore chip format according to the kit instructions and the chip then docked and primed with dH₂O in a BIAcore-Q instrument. Some water flushing was required to bring the baseline responses down in the flow cells to within the operating range of the BIAcore through washing off loosely bound material. Flow cell two was run and l-ethyl-3-[3-dimethylamino propyl]carbodiimide hydrochloride (EDC) solution (150 μL) was mixed with N-Hydroxysuccinimide (NHS) solution (150 μL) using a 200 μL mix. 50 μL was then injected at 5 μL/min and gave 86.6 RU of response. The injection was repeated and gave no further increase. Cortisol-OEG-NH₂ was made up to 100 mg/mL in DMF and then diluted to 1 mg/mL in PBS/T pH 9. The amine solution was then injected (100 μL at 5 μL/min). This was repeated for a total of eight injections and gave a total response of 5100.1 RU. The flow cell was then deactivated with ethanolamine (2 x 50 μL, 5 μL/min). Total immobilization response was 4824.9 RU. The surface was then flushed with 50 mM NaOH (15 μL, 5 μL/min) 15 times. This then further reduced the baseline response as loosely bound material was cleaned off the surface. BSA (500 μg/mL, 60 μL, 20 μL/min) was injected and gave -21.2 RU response, so no significant non-specific binding. NB Assume flow rates are 20 μL/min unless otherwise stated. The surface was regenerated (50 mM NaOH, 20 μL - was used in all regenerations). The injections were repeated twice more giving 24.6 RU and 20.2 RU - very low non-specific binding responses given the high protein loading. mAb to Cortisol (1 μg/mL, 60 μL US Biologicals C7904-1 IB) was then injected and gave 372.6 RU of binding. After injection end there was some slight dissociation but most of the binding remained intact. After regeneration the binding was repeated three more times and gave 449.2 RU, 420.7 RU and 41 1.2 RU. mAb (1 μg/mL, 70 μL) was mixed 1 :1 with Cortisol in dH₂O (1 μg/mL, 70 μL) and incubated at room temperature for 5 min before injection of 60 μL. This gave 13.3 RU binding. This was then repeated and gave 5.4 RU of binding - so there is minimal non-specific antibody binding. mAb (500 ng/mL, 60 μL) was injected and gave 167.6 RU.

2.3 Polymer coating ofBIAcorβ Chips - 5:1 EGDMA : MAA

The polymer coating of a BIAcore gold chip was repeated as before but using a 5:1 EGDMA: MAA volume ratio (i.e. 389.6 μL of EGDMA, 77.9 μL MAA). The chip was then docked and primed with dH₂O as before and flushed with water. Flow cell three most closely matched the response at baseline of the previous chip's flow cell two and was thus used in these experiments being washed thoroughly with dH₂O in-situ. The surface was activated as before and gave 210.6 RU of total response. This is 2.48 times that gained with the previous chip, thus possibly indicating that the use of a higher loading of MAA has indeed increased the number of carboxylic acid groups available for immobilisation roughly in proportion to the increase in MAA. The surface was then immobilised with cortisol-OEG-NH₂ as above and gave 5539.8 RU of response. Some of this is likely to be loosely bound material and so it may be more instructive to examine the first three injections which gave 2563.2 RU as compared to 1227.7 RU in the first chip, so once again about twice as much response. The surface was deactivated and cleansed with 50 mM NaOH as before.

BSA binding was repeated and gave 1189.3 RU, 872.2 RU and 711.3 RU. These results are much higher than that obtained with the 10:1 polymer, indicating much stronger non-specific binding response to BSA with a reduction in the EGDMA concentration. mAb was then injected (1 μg/mL, 60 μL, 20 μL/min) and this gave 430.1 RU - so much the same as the earlier polymer at 10:1. Five more replicates were done and gave: 423.8 RU, 417.2 RU, 371.8 RU, 337.1 RU and 362.4 RU. Cortisol in dH₂O (1 μg/mL, 70 μL) was incubated 1 :1 with Cortisol mAb (1 μg/mL, 70 μL) as before. This gave 1.8 RU. This was repeated and gave 6.9 RU, so very little nonspecific binding. mAb (500 ng/mL, 60 μL) was bound and gave 190.0 RU as expected. This was repeated and gave 167.0 RU but there was 14.6 RU residual binding not regenerated so this may be 181.6 RU. This polymer is not demonstrating any greater binding responses than were observed for the previous polymer at 10:1.

2.4 Polymer COA TING ofBIAcore Chips - 20: 1 EGDMA : MAA

A BIAcore gold surface was polymer coated as before but using 20:1 EGDMA : MAA (445.2 μL EGDMA, 23.3 μL MAA) and assembled into a BIAcore chip cartridge as before and docked and primed in dH₂O as before. Despite multiple priming and flushing of the surface at high flow rates, the baseline responses on all the flow cells remained too high for the BIAcore to record. The chip was checked with a laser to make sure that it was correctly oriented in the cartridge, so the polymer coated surface was simply too thick to be used.

2.5 Polymer coating ofBIAcore Chips - Thickness Variation

A new BIAcore gold chip surface was polymer coated as before but using the 10:1 EGDMA : MAA method with 11.5 μL of polymer solution. The polymer appeared visually quite thin and was photographed (not shown). The coating was thinner in the middle than the edges and the polymer coated chip was mounted into a BIAcore chip as before.

The chip was docked and primed in dH₂O as before but only flow cell two could be registered as within range after 3 x priming and dH₂O flushing and cleaning of the glass slide side which had been marked by the spin coater vacuum o-ring. The response of the baseline was high (—19,000 RU) and so the surface was thoroughly flushed with dH₂O at 20 μL/min in-situ.

Another new gold BIAcore chip was polymer coated using the 10:1 EGDMA : MAA ratio but with 46.0 μL of polymer solution. The polymer appeared very thin with good light diffraction and was mounted in the BIAcore chip assembly. The chip was docked and primed repeatedly and the glass surface cleaned thoroughly with lens tissue with and without water, but the responses in all flow cells were too high to register so it appears that the polymer layer is in fact too thick. These last two polymer coated surfaces developed impressions of the BIAcore flow cells in the polymer matrix.

The 1 1.5 μL polymer appears to have thicker coatings than the 23.0 μL polymer indicating that perhaps the lower volume of polymerisation solution is offering less shielding of the area close to the gold surface and hence greater surface bound polymerisation. Higher volumes of polymerisation solution (high excess) take longer to spin off and hence give thicker layers.

A new BIAcore gold surface was polymer coated with 2.3 μL of polymerisation solution as before (10:1 EGDMA : MAA). The polymer had the usual diffraction pattern effect but it is possible that the polymer did not coat the entire surface. The chip was docked and primed with dH₂O as before but despite thorough washing gave baseline responses that were too high and thus out of range.

The 11.5 μL polymer chip was docked and primed and flow cell two flushed overnight with dH₂O before immobilisation with Cortisol-OEG-NH₂ as before. The baseline dropped from 18327.0 RU to 15020.6 RU upon first injection and so was not quite as thick as it first appeared. Immobilisation gave a response of 7729.8 RU and ethanolamine treatment removed 431.2 RU of material. The chip was washed with NaOH as before. BSA non-specific binding was tested as before and gave 3.2 RU, -147.5 RU and 3.0 RU. mAb (1 μg/mL, 60 μL) was injected as before and gave 196.1 RU, i.e. about Vi of what was previously obtained. This is due to a thicker polymer layer indicating that the distance effect is dominant over the functional group binding effect at these thicknesses. The binding was repeated six more times and gave 154.5 RU, 152.6 RU, 96.1 RU, 156.9 RU, 132.2 RU and 83.2 RU. The 1 :1 mAb : Cortisol binding was done as before and gave 60.0 RU and 37.4 RU. mAb binding at 500 ng/mL gave 7.1 RU and 25.5 RU. The surface eventually ended up giving very little mAb binding at all. Based on the non-specific binding values obtained it appears that there may not have been any specific antibody binding at all. The falling responses are not due to the polymer coming off as there is still a thick polymer coating on the surface as evidenced by the baseline responses. It may be that the polymer coating is so thick no mAb binding can be observed and binding responses have just been mAb blocking the surface. Example 3 - Spreeta-Based Surface Plasmon Resonance Sensor Development

3.1 Polymer Coating

Methacrylate / ethylene glycol polymer coatings were prepared on Spreeta chips according to the method of Example 2.1.

3.2 Immobilisation

Cortisol-OEG-Boc (1.25 mg) was dissolved in formic acid (1 mL) and stirred for 4 h in the dark. The formic acid was then removed in vacuo. The residual cortisol-OEG-NH₂ was then made up to 100 mg/mL in DMF based on the free amine. The solution was then diluted to 1 mg/mL in PBS/T pH 9.0 (this is done just before use). The polymer coated Spreeta chip was then mounted in the Spreeta block and the flow cell clamped down using a 50 μm gasket. The system was then tested for leaks, calibrated in air and water and then washed with dH₂O pumping for 8 h at 20 μL/min and then overnight at 5 μL/min.

BIAcore EDC and NHS solutions (70 μL each) were mixed 1 :1 and 50 μL was injected at 5 μL/min followed immediately after completion by another 50 μL pulse. This gave a total response of 278.2 RU. The cortisol-OEG-NH₂ solution was then injected (8 x 100 μL injections) at 5 μL/min. This gave 5751.3 RU of response. NB All responses unless otherwise stated are for the white channel (first channel). The channels are connected in series, so all receive the incoming solutions in the order white, red, green. All flows are 20 μL/min unless otherwise stated. The surface was then deactivated with two pulses of BIAcore ethanolamine (50 μL each at 5 μL/min). The system was then flushed with dH₂O overnight. Once the immobilization process is started one cannot stop until it is completed (usually takes most of an entire day).

3.3 A ntibody Binding

The surface was then exposed to BSA (500 μg/mL, 60 μL) to block and this gave 274.5 RU of response. The surface was then regenerated with glycine pH 2, 20 μL and the process repeated three times more giving bindings of 970 RU, 895.9 RU and 1164.4 RU respectively. The surface then had mAb (1 μg/mL, 4 x 60 μL US Biologicals C7904-1 IB) injected and gave 514 RU. The surface was regenerated as before and the binding repeated and gave 516.5 RU and then again to give 501.2 RU. The surface then had mAb 1 μg/mL 1:1 Cortisol 1 μg/mL 4 x 60 μL injected and gave 7.7 RU. This process was repeated and gave - 227.2 RU. mAb (500 ng/mL, 4 x 60 μL) was then injected and gave 181 RU. This was repeated twice more and gave 209.7 RU and 246.6 RU. The concentration was reduced to 250 ng/mL and this gave 170 RU. Regeneration after this began to prove difficult with clear declines in responses. Various regeneration solutions were tried but only 50 mM NaOH gave proper regeneration.

The running phase was changed to HBS-EP. It was decided to reduce the number of injections to three and so mAb binding at 1 μg/mL 3 x 60 μL was tested three times with 50 mM NaOH 20 μL and gave 194.1 RU, 183.1 RU and 195.2 RU. The non-specific binding was tested by injecting 1 :1 mix of mAb 1 μg/mL: Cortisol 1 μg/mL, 3 x 60 μL giving 105.6 RU, 15.2 RU and 2.2 RU. So clearly we have at least 80 RU of specific binding. mAb bindings at 500 ng/mL, 3 x 60 μL was tested and gave 92.3 RU, 149.1 RU and 157.5 RU. mAb was then injected at 500 ng/mL 3 x 60 μL, flow adjusted to 10 μL/min and then anti-mouse IgG secondary antibody 400 μg/mL, 2 x 60 μL was injected. This gave a total of 1894.8 RU on a mAb binding of 128.5 RU, suggesting 13.7-fold enhancement assuming no non-specific binding. The process was repeated with IgG secondary antibody (Sigma M7023) at 200 μg/mL and gave 1468.7 RU on 330.2 RU mAb binding.

A secondary antibody IgG concentration plot was then prepared with mAb at 500 ng/mL (2x 90 μL) and secondary at 2 x 60 μL and 10 μL/min. concentrations of secondary antibody of 0, 25, 50, 100, 150, 200, 300, 400 μg/mL were tried. Each concentration was done in triplicate. The secondary antibody plot is given below in Figure 2.

This plot indicates that 300 μg/mL is probably the best choice for the secondary antibody concentration. The CVs of this plot average 13.0%, which is too high for an assay curve so this will clearly have to be further reduced as the work proceeds. A mAb binding plot was then prepared using 300 μg/mL secondary antibody and mAb concentrations of 0, 12.5, 25, 50, 100, 250, 350 and 500 ng/mL. This plot is given in Figure 3.

The plot shows clearly the expected decline in response as the primary antibody concentration is reduced but it is not quite linear in this form. The CVs are now averaging 7.3 % on the total response and so have improved but this plot indicates high non-specific binding (~ 1500 RU). The plot looks better when one just takes one replicate of the whole and plots that, Figure 4. Taking the slope of the linear fitted line, to get 100 RU of total specific response would require a final mAb concentration of 50 ng/mL. This should give a detection limit of around 50 pg/mL. The mAb only line is given in Figure 5. By using the ratio of the slopes of the lines one can see that the enhancement is 8.8-fold compared to 10.26-fold in BIAcore at 400 μg/mL secondary antibody concentration. An assay curve for Cortisol was then prepared with mAb concentration at 100 ng/mL (diluted 1 :1 with Cortisol standards to final concentration of 50 ng/mL) and secondary antibody concentration of 200 μg/mL. Cortisol concentrations of 0, 5, 10, 25, 50, 100, 250 pg/mL, 1, 5, 10 ng/mL were used in HBS-EP with 5 min pre-incubations. Each point was done in triplicate. Data was analysed for all three flow cells. Good standard curves were obtained for the total response in flow cell two (red) and the mAb only response in flow cell three (green). These curves are given in Figures 6 and 7. Assay curves could be obtained from flow cell one and for the enhanced and mAb only for the other two but not for the mAb only in flow cell 1. The quality of the other assay curves in terms of shape and regularity was poorer. The CVs for the flow cell one curve averaged 8.5% whilst those for flow cell two averaged 4.6% and those for flow cell three 5.4%.

The assay parameters are given in Tables 1 and 2.

Table 1. Assay Parameters - Enhanced Flow Cell Two

LOD 35 pg/mL

LOD90 20 pg/mL

IC50 160 pg/mL

Dynamic Range 45-520 pg/mL

Sensitivity 228 RU.mL/ng

Table 2. Assay Parameters - mAb Only Flow Cell Three

LOD 19 pg/mL

LOD90 17 pg/mL

IC50 84 pg/mL

Dynamic Range 29-430 pg/mL

Sensitivity 36.8 RU.mL/ng These assays have the necessary detection limit for the measurement of Cortisol in saliva. They are a bit narrower in active range than the BIAcore assay particularly at the higher concentration end. The sensitivity values obtained are excellent but the CVs remain a concern. The enhanced assay parameters have been calculated using the CVs rather than the standard deviations whereas this wasn't necessary for the mAb plot.

Example 4 - BIAcore Polymer SPR Biosensor for Cortisol

4.1 Set-Up

The chip prepared as described in Example 2, was docked and primed. When stored dry the polymer requires at least triple priming to fully rehydrate the polymer and remove all air pockets from within the structure. Flow cell two was used for these experiments without blank flow cell subtraction (BIAcore Q). The program used was:

Quickinject mAb solution (60 μL at 20 μL/min) (antibody to Cortisol, US Biologicals C7904- HB)

Flow to 10 μL/min

Quickinject anti-mouse IgG secondary antibody solution (60 μL) (Sigma M7023).

Flow to 20 μL/min

Regenerate with 1 x 20 μL of 50 mM NaOH, 1 x 20 μL 10 mM glycine pH 2, 1 x 20 μL of 50 mM NaOH

Time points are taken at beginning of first quickinject (baseline), end of first quickinject (mAb binding) and end of second quickinject (secondary antibody IgG). The regeneration is generally effective though sometimes some residual rise in the baseline can occur.

4.2 Secondary antibody binding study

The program was run with mAb set at 1 μg/mL and IgG secondary antibody concentrations of 0, 25, 50, 100, 150, 200, 300 and 400 μg/mL were used, four replicates of each point were done. The results showed that the reduction in signal with decreasing secondary antibody concentration is slight up to 200 μg/mL and steeper after that. The CVs are 4.1% on average (compare with 7.7% for the portable biosensor) so one can see how the BIAcore system has halved the errors. The total signal here is smaller than on the portable Spreeta system though (~5000 RU compared to -9000 RU on the portable Spreeta system).

4.3 Primary Antibody Plot

The program was run with mAb varied (0, 0.025, 0.05, 0.1, 0.25, 0.5, 1 μg/mL) and the IgG secondary antibody fixed at 200 μg/mL. Four replicates of each point were taken. The comparison of mAb only and total response are shown in Figure 8 (corrected for non-specific binding).

The results indicate that the secondary antibody has enhanced the primary antibody response. The enhancement is about 11.9-fold. The CVs average 1.0% of the total response (compare with 2.9% for the portable) so the use of BIAcore has reduced the errors to about 1/3 of those in the portable instrument. We have 1379 RU of specific binding at 1 μg/mL final mAb concentration compared to 1674 RU for the portable instrument, so slightly less. These results show that the limit of detection is about 100 pg/mL.

4.4 Reimmobilisation and surface decay A new flow cell (Fc3) was immobilised following the same method as that applied previously the flow cell was then tested in a BIAcore Q instrument. mAb was bound at 200 ng/mL and gave 83.9 RU. This was then enhanced by IgG secondary antibody (Sigma) at 200 μg/mL and this gave 800.9 RU. This binding was successfully regenerated with two pulses of 10 mM NaOH. This could be used to replace the 50 mM NaOH / 10 mM glycine regeneration sequence. Secondary antibody IgG non-specific binding was tested and gave 353.4 RU, so about 447.5 RU of specific binding. The mAb binding was repeated and gave 81.9 RU and 1106.3 RU on enhancement. The binding was repeated again and gave 106.1 RU and 1342.4 RU. The secondary antibody IgG non-specific binding was retested and gave 444.9 RU. It seems the secondary antibody non-specific response is increasing but so also is the primary antibody response being labelled by the secondary.

The primary antibody binding was repeated at 50 ng/mL and gave 31.8 RU response that was enhanced by the secondary antibody IgG to 354.5 RU. This secondary antibody IgG response is lower than the non-specific response recorded above. The binding was repeated and gave 30.6 RU for primary and 501.9 RU for secondary. The binding was repeated and gave 84.4 RU for primary and 506 RU for secondary. Now on the BIAcore 3000, the mAb (400 ng/mL) was mixed 1 :1 with blank buffer and injected and enhanced as before giving 112.7 RU primary and 1939.4 RU enhanced. The process was then repeated but with 10 ng/mL Cortisol instead of blank buffer. This gave 31.3 RU primary and 817.1 RU enhanced response - so good specific signal. The process was then repeated with mAb at 100 ng/mL instead and gave 24.5 RU and 563.8 RU for the blank and 22.5 RU and 450.2 RU for the 10 ng/mL so 1 13.6 RU of specific binding across the dynamic range. This was then repeated and gave 51.7 RU and 763.4 RU for the blank and 32.9 RU and 652.5RU for the 10 ng/mL, so 110.9 RU of specific binding. The amount of mAb non-specific binding and thus the amount of secondary antibody non-specific binding is increasing likely due to chemical changes on the surface of undetermined cause.

Assay Development - Cortisol in Buffer

An assay was then set up for Cortisol standards at 0, 25, 50, 100, 250, 500, 1000, 2500, 5000, 10000 pg/mL and using mAb at 100 ng/mL (before 1 : 1 mixing) and secondary antibody at 200 μg/mL (Sigma), five replicates. The assay curves are given below in Figure 9.

The assay data are summarised in Table 3.

Table 3. Assay characteristics

mAb Only Enhanced

LOD90 (pg/mL) 58 128 (163)

IC50 (pg/mL) 1410 1295

Dynamic Range (pg/mL) 510-2400 352-2650

Sensitivity (RU.mL/ng) 4.3 34.3

The assay has adequate detection limits for Cortisol detection in saliva and at least for mAb only it has comparable detection limits to those observed for the dextran polymer at the same mAb loading. The dynamic ranges are not bad but could be a little lower to cover low Cortisol saliva concentrations. The slope sensitivity after enhancement is only half that obtained with the dextran assay (first iteration). The reason for such weakness in signal is not known. The results were obtained by using the un-modified flow cell 4 as a reference flow cell for subtraction of the responses. There was a continual updrift in the non-specific binding responses for both the primary and secondary antibodies. This drift has to be subtracted out by the reference flow cell but is clearly undesirable.

Saliva for preparing saliva standards is stripped with activated charcoal by collecting saliva sample (with sugar-free gum) into polypropylene tube, freezing, thawing, centrifuging at 4600 xg for 15 min. and 7.5 mL of supernatant added to 75 mg of activated charcoal and vortex mixed on the maximum setting for 1 min. before shaking at 560 rpm overnight at room temperature. The next day the sample is centrifuged at 4600 xg for 15 min. three times to separate saliva from charcoal. The saliva is then spiked with steroid to produce a high concentration standard which is then diluted as required with the stripped saliva to produce a set of standards.

A new polymer coated flow-channel previously immobilised with Cortisol (5:1 EGDMA:MAA,) was docked and primed and then used for a Cortisol salivary immunoassay using niAb at 100 ng/mL before mixing, secondary antibody Sigma (200 μg/mL) and Cortisol standards in stripped saliva at 0, 0.025, 0.05, 0.1, 0.25, 0.5, 1, 2.5, 5 and 10 ng/mL. The assay curve is given below in Figure 10 and the assay parameters in Table 4.

Table 4. Assay parameters for Cortisol saliva assay

LOD90 38

IC50 350

Dynamic Range 61-633

Slope Sensitivity 133 RU.mL/ng

The assay has the desired sensitivity and sufficient assay signal. The assay curve does not cover the higher Cortisol concentrations (past 1 ng/mL) further indicating the narrow dynamic range of the assay. There are also drift problems associated with changing non-specific binding levels. It may be possible to further reduce drift by applying a more severe regeneration solution. Example 5 - Testosterone BIAcore Polymer studies Synthesis of Testosterone-OEG-NH₂

Testosterone-OEG-NH₂ was synthesized by the method described in PCT/NZ2008/000021 (Example 3), where it is named testosterone-PEG-NH₂. The contents of this document are fully incorporated herein by reference.

5.1 Immobilisation

Two new 10:1 EGDMArMAA chips at 23 μL coating volume were prepared for BIAcore but both were rejected by the instrument as too thick on all four flow channels. A sensor surface previously prepared with 10:1 EGDMA: MAA at 11.5 μL applied volume (Mk4) was docked and flow cell 4 immobilised with testosterone-OEG-NH₂ using the same procedure as for the Cortisol immobilisations. Flow cell 1 was activated and deactivated as before to act as blank.

5.2 Testosterone Buffer Assay

A buffer-based assay for testosterone was formulated using anti-testosterone mAb (US Biological T2950 - 18A) concentration at 400 ng/mL (200 ng/mL final) and Sigma secondary antibody IgG (Sigma M7023) concentration at 200 μg/mL and testosterone standards in running buffer at 0, 50, 100, 250, 500, 1000, 2500, 5000, 10000, 25000 pg/mL. The assay curves with subtraction of non-specific binding are given below in figure 11.

The assay parameters are given below in Table 5.

Table 5. Assay parameters for testosterone buffer assay 1 with and without reference subtraction.

Subtracted Unsubtracted

LOD90 560 148

IC50 4900 493

Dynamic Range 1600-8000 198-4400

Slope Sensitivity 4 RU.mL/ng 55 RU.mL/ng

The assay has a lowest detection limit of 148 pg/mL. 5.3 Testosterone enhancement with Secondary Antibody IgG-GoId

25 nm gold colloid was prepared according to the citrate reduction method and the anti-mouse IgG secondary antibody (Sigma M 7023) was conjugated at 3mg/mL starting concentration (300 μg/mL final concentration) according to JS Mitchell, Y Wu, CJ Cook, and L Main Anal. Biochem. 343, 125-135 (2005).

A testosterone assay curve was then prepared using the 0.4 concentration factor of the gold to enhance and using mAb at 200 ng/mL final concentration and using testosterone concentrations at 0, 0.1, 0.25, 0.5, 1, 2.5, 5, 10, 25 and 50 ng/mL (five replicates of each). The assay curves are given in Figure 8 and the assay parameters in Table 6.

Table 6. Assay Parameters for Testosterone Gold- Enhanced Immunoassay

Subtracted Non-subtracted

LOD90 0.28 1.1

IC50 3.4 4.1

Dynamic Range 0.40-20 1.7-32

Slope Sensitivity 11.6 RU.mL/ng 39.5 RU.mL/ng

The above examples are illustrations of practice of the invention. It will be appreciated by those skilled in the art that the invention can be carried out with numerous modifications and variations. For example the monomers, the haptens, the linkers, the antibodies and the concentrations used may all be varied.

Any discussion of documents, acts, materials, devices or the like that has been included in the present specification is solely for the purpose of providing a context for the present invention. It is not to be taken as an admission that any or all of these matters forms part of the prior art base or were common general knowledge in the field relevant to the present invention as it existed before the priority date.

The term "comprising" as used in this specification means "consisting at least in part of, that is to say when interpreting statements in this specification which include that term, the features, prefaced by that term in each statement, all need to be present but other features can also be present.

Claims

1. A biosensor having a ligand binding surface that is a polyalkene having a thickness of less than 100 nm and bearing carboxy groups at the surface.

2. A biosensor as claimed in claim 1, wherein the polyalkene is prepared by spin-coating the surface with a polymerisation mixture comprising at least one alkene bearing a carboxy group.

3. A biosensor as claimed in claims 2, wherein the mixture comprises at least one compound having two or more alkene groups.

4. A method for preparing a biosensor having a ligand binding surface that is a polyalkene having a thickness of less than lOOnm comprising spin coating to a surface of the biosensor with a polymerisation mixture comprising a monoalkene and a dialkene and a solvent.

5. A method as claimed in claim 4, wherein at least one of the alkenes bears a carboxy group or a group that can be converted to a carboxy group.

6. A method as claimed in claim 5, wherein the polymerisation mixture comprises an alkene bearing a carboxy group.

7. A method as claimed in claim 6, wherein the polymerisation mixture comprises an alkene bearing a carboxy group.

8. A method as claimed in claim 7, wherein the polymerisation mixture comprises methacrylic acid

9. A method as claimed in any one of claims 4-8, wherein the mixture comprises an initiator to initiate polymerisation.

10. A biosensor as claimed in any one of claims 1-4, wherein the thickness of the biosensor ligand binding surface is less than 50 nm.

11. A biosensor as claimed in claim 10, wherein the thickness is less than 25 nm.

12. A biosensor as claimed in claim 11, wherein the thickness is less than 10 nm.

13. A biosensor as claimed in any one of claims 1-4, wherein the polyalkene is a copolymer Of CHR₁=CR₂CO-O-A-CR₃=CHR₄ (Formula I) and CH(Rs)=C(Re)-COOR₇ (Formula II) wherein Ri, R₂, R₃, R₄, R₅, and R₆ are independently selected from the group consisting of hydrogen, methyl, and ethyl, and A is an aliphatic chain comprising 1-20 carbon atoms and 0-6 oxygen atoms in the chain wherein any oxygen atoms are not adjacent to each other, and R₇ is hydrogen or an aliphatic chain with 2-20 carbon atoms, bearing a carboxy group, and optionally comprises one or more ester groups or ether groups within the chain.

14. A biosensor as claimed in claim 13, wherein Formula I is CHR₁=CR₂CO-O-A-O- COCR₃=CHR₄.

15. A biosensor as claimed in claim 14, wherein Formula I is CH₂=CR₂-CO-O-(CH₂-CH₂- O)_n-COCR₃=CH₂.

16. A biosensor as claimed in claim 15, wherein a biosensor molecule of Formula I is ethylene glycol dimethacrylate.

17. A biosensor as claimed in claim 13, wherein in Formula II, R₆ is methyl and R₇ is (C₁-C₅- alkyl-O)_n-CO-CH=CH-COOH wherein n=l-10 or R₇ is -Ci_-20-alkyl CO-CH=CH- COOH.

18. A biosensor as claimed in claim 17, wherein the R₇ groups comprise C_1-I0 alkyl or (CH₂CH₂O)_n wherein n=l-4.

19. A biosensor as claimed in claim 18, wherein Formula II is methyacrylic acid or mono-2- (methacryloyloxy)ethyl maleate.

20. A biosensor as claimed in claim 19, wherein the polyalkene is selected from copolymers of ethylene glycol dimethacrylate and methacrylic acid, of ethylene glycol dimethacrylate and mono-2-(methacryloyloxy)ethyl maleate, of either of these two combinations and also propenethiol, and of oligoethylene glycol dimethacrylates copolymerised with mono-2-(methacryloyloxy)oligoethyleneglycol maleate.

21. A biosensor as claimed in any one of claims 1-4 and 10-20, wherein a biosensor is a surface plasmon resonance chip, and the coating on a metal coating on a surface of the chip.

22. A biosensor as claimed in claim 21, wherein the metal is selected from gold, silver, copper, aluminium, and platinum.

23. A biosensor as claimed in claim 21, wherein the metal is present in a thin layer on an optical chip suitable for use as a surface plasmon resonance biosensor.

24. A biosensor having a surface that is a polyalkene having a thickness of less than 100 run and bearing a ligand at the surface, the ligand being linked to the surface in a group formed from a carboxy group carried by the polyalkene and a group on or linked to the ligand.

25. A method for measuring the concentration of a ligand comprising contacting the ligand and a binding partner in a sample with a biosensor as claimed in claim 24 and measuring binding of the binding partner to the ligand to allow calculation of the amount of ligand in the sample.

26. An assay as claimed in claim 25, wherein binding of the binding partner to the immobilised ligand is measured by surface plasmon resonance.

27. An assay as claimed in claim 25, wherein the ligand is a steroid.

28. A method for measuring the concentration of a binding partner of a ligand in a sample, comprising contacting the binding partner in a sample with a biosensor having a surface of the invention having immobilised ligand bound to it, and determining the amount of bound ligand.

9. A method as claimed in any one of claims 25-28 that is a flow through assay where the sample flows past the surface bearing the immobilised ligand.