Lecture 4. SPR: immobilization strategies Kinetics

Size: px

Start display at page:

Download "Lecture 4. SPR: immobilization strategies Kinetics"

Alisha Norman
5 years ago
Views:

1 Lecture 4 SPR: immobilization strategies Kinetics

2 Surface plasmon sensor Principle of affinity SP biosensor

The Aim of the Lecture How to make our SPR sensor selective? What type of a signal we expect to measure? How to process it and what information we extract from there.

3 The Aim of the Lecture How to make our SPR sensor selective? What type of a signal we expect to measure? How to process it and what information we extract from there. Quantitative parameters of a binding reaction: concentration of an analyte, reaction rates, affinity etc. specific for a particular sensor type specific for a particular reaction and geometry

4 Immobilization techniques Different applications might require specific binding techniques, however there are general requirements: successful label-free measurement of a specific binding event requires the best possible activity of the ligand conformation and orientation should be close to the in vivo situation non-specific binding should be kept low

Surface modification physisorption: (partial) unfolding of a protein on the surface uncontrolled exchange of molecules between the surface and the solution Self-Assembled Monolayers (SAM) +

5 Surface modification physisorption: (partial) unfolding of a protein on the surface uncontrolled exchange of molecules between the surface and the solution Self-Assembled Monolayers (SAM) + densely packed and stable structure, + flexibility of the end-group chemistry (usually carboxyl but other can be used still fairly rigid doesn t use the whole volume penetrated by the field

6 Surface modification Dextran hydrogel: thin hydrophilic (hydrogel) layer based on highly flexible mainly unbranched dextran molecules Advantages: + highly hydrophilic environment with non-cross-linked structure (keeps molecules in a solution-like state) + increased binding capacity + thickness of the layer can match the penetration depth + can be tailored using various molecular weight (10k 1M) and different chemistry Other chemistries: Polyvinyl alcohol, polyacrylic acid, poly-llysine

7 Immobilisation strategies covalent capture Hydrophobic (supported bilayer)

8 Covalent immobilisation

9 Amine coupling Coupling to carboxylic groups: the most widely used strategy Typical protocol: injection of EDC mixed with NHS in water, ph=4-6 (natural buffering) injection of a protein in acetate buffer ph=4-6. Positively charged protein will be attracted to the negatively charged hydrogel ( pre-concentration ), 1-10min. Protein concentration up to 50 ng/mm 2 (several monolayers) can be obtained Coupling under mild conditions, very few immobilization points, activity is largely preserved (e.g. IgG usually retains of approx. 75% of activity) surface modification with amino groups

10 Amine coupling typical sensogram for amine coupling Proteins with pi>3.5 can be effectively pre-concentrated at ph~5. However acidic proteins will be repelled from the layer

11 Amine coupling micelle mediated immobilization of negatively charged proteins

12 Covalent immobilization Coupling to thiol groups mainly used in the situations when the ligand lacks amino groups or they are located close to the active center can be introduced on the sensor surface or on the molecule very specific disulfide bond can be cleaved by reactive thiols (e.g. mercaptoethanol) coupling to thiol-modified surface coupling to pyridyl disulfide surface

13 Covalent immobilization coupling thiolated binding partner to maleimidemodified sensor surface

14 Covalent immobilization coupling to aldehyde groups based on the generation of aldehyde functionality by oxidation of carbohydrate residues normally not located near the active site, therefore activity is well preserved antibodies are particularly well suited for this type of immobilizaation

15 Capture-based coupling can be indispensable when covalent immobilization might impair activity of the binding site particular molecule should be captured from e.g. cell lysate analyte can be removed requirement: affinity should be high Commonly used pairs: avidin-biotin bond: very robust, as strong as covalent bond His-tagged protein nitrilotriacetic acid (NTA)+Ni 2+. Can be easily broken by a chelating agent e.g. EDTA Antibody antigen

16 included in the vesicles Coupling via Lipid layer Important for targeting membrane protein with drugs (drug screening) targeting lipid membranes (e.g. AMPs) bilayer capture: phospholipids modified with a thiol group histidine modified lipids oligonucleotide modified lipids included in the layer hydrophobic groups attached to CM dextran membrane protein attached to CM dextran

17 Coupling via Lipid layer adsorption of lipid vesicles

18 Coupling via Lipid layer lipid-detergent method

19 Protein modification in solution can be considered: to create functional groups for immobilisation in a particular region of protein to tune pi of the protein

Protein immobilization choice of ph might affect a conformation of protein during binding it might be beneficial to protect the active site of the protein by

20 Protein immobilization choice of ph might affect a conformation of protein during binding it might be beneficial to protect the active site of the protein by having analyte present during immobilisation additional partial EDC/NHS cross-linking might be important to prevent a complex protein from dissociation into monomers

21 Immobilization of other molecules Oligonucleotides: use of biotinylated derivatives, binding at neutral ph EDC/NHS crosslinking of amino-modified nucleotide in the presence of CTAB

22 Biochip Kinetics

23 Modeling Molecular Interaction in SPR The aim: design a model kinetic equation that describes how amount of ligand-analyte depends on time, concentration of analyte and amount of free binding sites left. The signal: ξ = const M A γ 1:1 binding The maximal signal: ξ S = const M β ξ / ξ = γ / β S A

24 General biosensor model Kinetics Reaction kinetics Mass transport

25 Rates of chemical reactions Consider reaction: A+2B -> 3C+D The rates are: d [ D] 1 d[ C] d[ A] 1 d[ B] = = = dt 3 dt dt 2 dt We can define rate of reaction: 0 -> 3C+D-A-2B v d[ ξ ] =, dn dt j = ν dξ j ν j

26 Rate laws Rate of reaction is often proportional to the concentration raised to some power, e.g. b v = k[ A] a [ B] Overall order of the reaction: a+b+... order of the reaction with respect to A: a Reactions of zero order v = k

27 Integrated rate law First order reaction: A->B d[ A] = k[ A] dt ln( A/ A0) = kt [ ] [ ] kt A = A e 0 Half live time kt 1 1 = ln( A0 / A0 ) = ln( ) / 2 = ln 2 Half live time is independent of initial concentration for 1st order reaction

28 Integrated rate law (contd) Second order reaction: A->B d[ A] 2 = k[ A] dt 1 1 = kt [ A] [ A ] 0 [ A] = 1+ [ A ] 0 kt[ A ] 0 Half live time t = 1/ 2 1 ka 0 Half live time varies with the initial concentration for 2nd order reaction

29 First order reaction close to equilibrium Consider reactions: A B At equilibrium forward and reverse rates are the same B A d[ A] d[ B] =, dt dt k[ A] eq = k ' [ B] eq, K = k k ' = [ B] [ A] eq eq Equilibrium constant

30 Pseudo-first order Kinetics For an analyte A (in solution) and a receptor R (immobilized) k a A+ R AR kd Association rate: dγ a = kaα β γ α = dt Dissociation rate: dγ d dt Summing: = k d ( ) [ A] γ dγ = kaα ( β γ) kdγ dt

31 Pseudo-first order Kinetics Let s consider a situation when a high concentration of analyte a 0 is injected into volume V with surface S. At longer time: αv + γs = α V = const 0 dγ S = ka α0 γ ( β γ) kdγ dt V dγ k γ a eq = 0 K = = dt k S d α0 γeq β γ V ( ) eq k a2 <k a1 K=const

32 Pseudo-first order Kinetics dγ S = ka α0 γ ( β γ) kdγ dt V if the concentration of analyte is high: dγ k γ a eq = kaα0 ( β γ) kdγ; K = = dt k α β γ d 0 ( ) eq

33 Pseudo-first order Kinetics Equilibrium analysis (not affected by mass transport) ξ EQ ξ = Kα 0 (1 + Kα ) 0 binding isotherm

34 Other kinetic models Zero-order reaction following initial binding (conformational change in AR complex that blocks dissociation) k k A a1 a2 + R AR AR kd1 kd2 dγ 2 dt dγ1 dt = k γ k γ a2 1 d2 2 = k α ( β γ γ ) k γ k γ + k γ a d1 1 a2 1 d2 2 As the total mass is measured by the sensor ξ / ξ = ( γ + γ )/ β S 1 2 *

35 Other kinetic models Parallel pseudo first-order reactions (e.g. two different receptors or two different analytes) k a1 A R AR kd1 k 2 2 a 2 A + R AR kd 2 β = [ R ] = pβ β = [ R ] = (1 p) β dγ 2 = ka 1α0( β1 γ1) kd1γ1 dt dγ1 = ka2α0( β2 γ2) kd2γ2 dt ξ / ξ = ( γ + γ )/ β S 1 2

36 Other kinetic models Multivalent receptor binding: single receptor binds more than one molecule (e.g. streptavidin, antibodies, triplex formation) k + a1 A R AR kd1 k a 2 AR + A ARA kd 2 β = [ R ] = pβ β = [ R ] = (1 p) β dγ 2 = ka2αγ 0 1 kd2γ2 dt dγ1 = ka 1α 0( β γ1 γ2) kd1γ1 ka2α0γ1+ kd2γ2 dt ξ / ξ = ( γ + 2 γ )/ β S 1 2

37 Thermodynamics in SPR change in Gibbs energy can be found from equilibrium constant: 0 G = RTln K Enthalpy and entropy of the reaction can be found from temperature dependence (van t Hoff equation) G = H T S Activation energy can be found from Arrhenius equation: act k = Pexp( E / RT)

38 Association/Dissociation in an experiment Plateau + Span Plateau Association: Response = Req * (1 - e -kobs * t ) Dissociation: Response = Span * e -kd* t + Plateau

39 SPR case: Mass transfer+reaction

40 Data processing Processing steps: zero response to the base line before analyte injection align all responses so that injection starts at the same point subtract the reference cell response subtract the averaged response to buffer injection

41 Data processing measured measured response: response: receptor receptor reference reference receptor receptor and and reference reference aligned, aligned, base base line line subtracted subtracted reference reference subtracted subtracted 4 injections injections averaged averaged + buffer buffer injection injection buffer buffer injection injection subtracted subtracted data, data, ready ready to to be be analyzed analyzed

Data processing Software: Scrubber2, (Biosensor Tools http://www.cores.

42 Data processing Software: Scrubber2, (Biosensor Tools BiaEvaluation (Biacore AB, Sweden)

43 Data processing Global analysis: k A A+ B AB kd All responses within the data set are fitted to the same values of k A and k D. Chi-squared is calculated for all curves mass transfer limited k k A m A A + B AB km kd conditions to reduce mass transfer effect: low ligand density, high flow rate.

44 Data processing Protein-antibody interaction Best fit using bimolecular model Best fit using mass-transport model

45 Thermodynamics for drug discovery even contribution entropy driven enthalpy driven G = H T S large entropy: affinity increases w. temperature Shuman et al, J.Biomol.Rec.17, p.106 (2004)

to analyse directly (e.g. due to low molecular mass of analyte).

46 How to improve signal for small analytes Sandwich assay Competition or inhibition analysis can be used for interaction that are difficult to analyse directly (e.g. due to low molecular mass of analyte). Competition Inhibition peptide 12p1 inhibition of gp120 YU2 binding to MoAb

47 Other possibilities Qualitative analysis (e.g. screening of several analyte) Determining stoichiometry of a reaction Epitope mapping (checking secondary antibodies) Binding to lipid membranes

48 SPR detection for continous measurements Molecular detection format: continuous detection with weak affinity antibodies. Example: maltose detection using anti-maltose ab.

49 SPR Biosensors for Medical diagnostics diagnostic methods based on monitoring of concentration of disease biomarkers (e.g. PSA) are gaining popularity Desirable: test directly on bodily fluid sufficient throughput continuous monitoring SPR has a potential to meet those requirements

50 SPR Biosensors for Medical diagnostics cancer biomarker (12 approved by FDA) PSA carcinoembrionic antigen (CEA): colorectal& breast cancer cancer antigen CA15-3 (breast cancer) CA125 (ovarian cancer) carbohydrate antigen CA19-9 (pancreas, colon, stomach cancer) alpha-fetoprotein AFP (liver&testicular cancer)

51 SPR Biosensors for Medical diagnostics PSA detection: current clinical assays: ELISA (enzyme-linked immunosorbent), 0.1ng/mL SPR sandwiched assay: 0.15 ng/ml on planar surfaces, 2.4ng/mL on hydrogel 1. mouse monoclonal anti-psa immobilized 2. sample injection 3. amplification with polyclonal rabbit anti-psa 4. amplification with biotinylated anti-rabbit 5. streptavidin coated latex spheres or 4. anti-rabbit coated colloidal gold Why planar surfaces are better for this assay? Besselink et al, Anal. Biochem 333, p. 165 (2004)

52 SPR Biosensors for Medical diagnostics Heart attack markers: Troponin I (TnI), troponin T (TnT), Troponin C (TnC), myoglobin, fatty acid binding protein SPR experiment: 1. biotinylated ctni antibodies immobilized on avidin layer on SAM 2. direct measurement (2.5-40ng/mL) or sandwiched assaywith detection limit 0.25ng/mL and range ng/mL

53 SPR Biosensors for Medical diagnostics Antibody based assays type I diabetes diagnostics using anti-glutamic acid decarboxylase ab (GAD) (GAD immobilized on the surface) detection of antibodies against cholera toxin (cholera toxin immobilized on the surface) hepatitus C anti-adenoviral antibodies Hormone based assays direct detection of hcg (human chorionic gonadotropin, pregnance marker) using anti-hcg immobilized with biotinylated oligos. Detection limit below 0.5ng/mL detection of estrone and estradiol using inhibition format Drug detection coumarin (anticoagulant, 7-OHC): competition assay with 7-OHc on the surface and antibodies injected in a mixture with serum warfarin (anticoagulant) morphine and morphine metabolites

54 SPR for food safety Bacteria: difficult to detect due to large size. Most of the time have to be destroyed by heat, ethanol or detergents E. coli O157:H7 direct detection using immobilized antibodies. Detection limit down to 100 cells/ml detection of PCR products of E.coli genome detection of enterotoxin Salmonella Listeria Campylobacter Jejuni (leading cause of diarrea)

55 SPR for food safety Proteins: secreted by infectious bacteria, toxic in low doses, have low molecular weight 5kDa- 150kDa Staphylococcal enterotoxins direct detection using immobilized antibodies. Detection limit down to 0.5ng/mL with secondary antibody amplification Botulinum neurotoxins Detection limit down to 0.5ng/mL with sandwich assay with polyclonal antibodies.

56 SPR for food safety Low molecular weight compounds large diffusion rate but low molecular weight doesn t produce significant change in refractive index Domoic acid: neurotoxin originated from algae detection using molecularly imprinted polymer. Detection range 2ng/mL 3.3 ug/ml inhibition assay. Detection limit 0.1 ng/ml Mycotoxins (produced by Aspergillus, Penicillum and Fusarium) Detection limit down to 0.5ng/mL with sandwich assay with polyclonal antibodies.

57 Problem Derive equations for parallel pseudo first-order reaction kinetics with a single receptor and two different analytes.

Lecture 7. Surface Reaction Kinetics on a

Lecture 7. Surface Reaction Kinetics on a Lecture 7 Data processing in SPR Surface Reaction Kinetics on a Biochip Surface plasmon sensor Principle of affinity SP biosensor Data processing Processing steps: zero response to the base line before