Kinetics on the Octet Systems: What Lies Beneath the Curves

Pui Seto Scientist ForteBio, A Division of Pall Life Sciences

Basic Kinetics: What can the sensorgram tell us? Background on BLI technology An ideal sensorgram Recognizing non-ideal behavior More complex models Using more complex models to optimize kinetics assay

Label-free, real-time analysis

• A layer of molecules attached to the tip of an optic fiber creates an interference pattern at the detector.

BioLayer Interferometry (BLI)

• A layer of molecules attached to the tip of an optic fiber creates an interference pattern at the detector.

• Any change in the number of molecules bound causes a measured shift in the pattern

BioLayer Interferometry (BLI)

Octet Systems and Features

LOD (direct): (multi-step):

LOD (direct): (multi-step): Sub-ng/mL Sub-ng/mL

Inside the Octet “Classic” System

ForteBio Biosensors and Assay Kits Biosensor Biosensor Functionality

AHQ AMQ FAB ProA ProG ProL SA HIS NTA GST AMC AHC AR2G APS SSA

Anti-Human IgG Fc Anti-Murine IgG Fv Anti-Human Fab CH1 Protein A Protein G Protein L Streptavidin Anti-Penta-HIS Ni-Tris NTA Anti-GST Anti-Murine IgG Fc Capture Anti-Human IgG Fc Capture Amine Reactive 2nd Gen Aminopropylsilane Super Streptavidin

Kits/Methods Dip and Read Immunogenicity Dip and Read Residual Protein A Dip and Read HCP Custom Assay

Custom biosensors can be made to specifications by Fortebio

Kinetics Application: Drug Discovery Process

Research: Target ID

& Validation

Lead Characterization & Optimization

10’s -1000’s 100s 10K – 100s

# of Samples per run

Primary Screening

100K – 3M

Antibody Applications: 1) Antigen generation - identify high producers 2) Serum titering 3) Screening for antigen specific antibodies 4) Quantitation of hybridoma supernatants 5) Off-rate ranking 6) Epitope binding and domain mapping 7) Apparent affinity measurement KD 8) Antibody sandwich pair identification 9) Assay development 10) Speed up humanization workflow 11) Identify high producers for manufacturing

Secondary Screening & Hit Validation

>150 Daltons

BLI Platform Capabilities

Quantitation • Direct • Sandwich • ELISA • mg/mL to sub-ng/mL

Kinetics • Label-free ka, kd, KD • Proteins • Peptides, Oligos • Small molecules

Kinetics: What can a sensorgram tell us?

• Before we launch into simple kinetics theory, we should keep in

mind that: • Curve shapes should only be described by the rate constants and

the analyte concentration • If there are any other influences at play, then the kinetics

constants which you calculate can be meaningless

• Don’t over interpret data

An Ideal Sensorgram A ‘classic’ 1:1 binding curve

A + B A B k a

d k

Both the association and dissociation phases follow a path described by a single exponential function

Res

pons

e

1:1 Binding

Dissociation Association

What are Req, Rmax and KD?

• Most curves, if the binding is left for long enough, will reach a point where the rates of association and dissociation are the same. This is the equilibrium binding level (Req)

• There is a fixed amount of ligand on the sensor surface, so there must be a maximum possible amount of sample binding at equilibrium. This is the saturating or maximum binding level (Rmax)

• The affinity constant KD is defined by the ratio of rate constants kd / ka

The relationship between Req, Rmax and KD

The simple 1:1 Interaction: A + B A B k a

d k

%ag

e of

Rm

ax

Time

100

50

Saturation (Rmax) is achieved if conc = 100x KD

Equilibration (Req) at 50% saturation if conc = KD

Detection limit around 0.1x KD

The relationship between Req, Rmax and KD

[ ] [ ][ ] a

dD k

kAB

BAK =⋅

=

The simple 1:1 Interaction: A + B A B k a

d k

%ag

e of

Rm

ax

Time

100

50

Saturation (Rmax) is achieved if conc = 100x KD

Equilibration (Req) at 50% saturation if conc = KD

Detection limit around 0.1x KD

ka = M-1s-1 kd = s-1

A Closer Look at Affinity Constant KD

[ ] [ ][ ] a

dD k

kAB

BAK =⋅

=

This is a VERY important relationship:

Relationship shows that the Affinity (KD) is equal to the ratio of the rate constants

So, ANY set of rate constants which have the same ratio will yield the same KD

=== nM1kkK

a

dD

For example: 10-3

106 10-6

103 etc

KD and Rate Constants This sensorgram shows the same concentration of 3 different analytes binding to the same immobilized ligand. They have the same KD but very different rate constants, and would look identical in an end-point assay such as ELISA

Time

Res

pons

e

[ ] [ ][ ] a

dD k

kAB

BAK =⋅

=

Dissociation is a simple decay process, and is independent of sample concentration

kd % complex dissociated per

second

Time to 50% dissociation

1 100 0.69 s 0.1 10 6.93 s

1 x 10-2 1 69.3 s 1 x 10-3 0.1 11.55 min 1 x 10-4 0.01 1.93 h 1 x 10-5 0.001 19.25 h 1 x 10-6 0.0001 8 days

Slow off-rates need a long time to fit accurately!

Recognizing Non-ideal Behavior

Recognizing Non-ideal Behavior

Time

Res

pons

e Non ideal behavior arises where the interaction proceeds through multiple binding sites.

Often a slower phase, never reaching equilibrium

Often has portion which never dissociates

1:1 Binding

Heterogeneous Binding

Here, the curves show more than one phase, both in the association and dissociation curves. Basically, what this means is that there are more than one binding event going on.

The Most Likely Cause of Heterogeneity

Heterogeneity can come from the original samples or from the way that the ligand has been immobilised. Direct coupling methods like amine coupling will usually tend to give rise to heterogeneity due to their potentially random orientation. Capture approaches are better for kinetics, since they use specific orientation. Heterogeneous binding can also be observed if concentrations well above the KD are used, where non-specific binding can become more significant.

Mass transport (MT) Limited Binding

More difficult to recognise than Heterogeneity The curves are slower than expected and can be limited by the speed at which the analyte diffuses to the surface.

Asolution + B A B k a

d k

k m Asurface

-k m

To reduce MT effects, it is best to consider the binding as a ‘supply and demand’ process. If demand for binding outweighs supply to the surface (by diffusion), then it will be that diffusion which defines the observed rates. So, to minimise MT effects, limit consumption and speed up supply by: Reducing the level of immobilised ligand (whilst still maintaining the required sensitivity) and increase the stir rate of the sample plate

Solutions For Correcting Non-ideal Binding

• Heterogeneity • Reduce sample concentration range and use capture approach

• Concentrations far above KD often unmask heterogeneous, non-specific sites

• Use capture approach for kinetics experiment, e.g., AHC, or AMC biosensor

• Try using a new lot of ligand or analyte

• Mass Transport fit • Try to reduce sample depletion and increase sample supply

• Increase ‘shake speed’ • Decrease ligand level

Complex Binding System

More Complex Reaction Schemes

• Here, deviation from 1:1 binding is a function of the type of interaction, rather than some experimental artifact . The 2 most common complex reaction schemes are: • Heterogeneous ligand • Bivalent analyte

• Let’s look at these on the next few of slides……..

Heterogeneous Ligand

Assumes 2 independent ligand binding sites

A + B1 A B1 k a1

d1 k

A + B2 A B2 k a2

d2 k

B1 B2

A A

This is an example of ‘parallel’ interactions, where the formation of AB1 and AB2 proceed independently of each other,

Octet software can calculate both KDs

Bivalent Analyte

Assumes the bivalent analyte can form the ’bridged’ AB2 complex. This causes a slower dissociation than expected and is purely an artifact of the surface interaction.

A + B A B k a1

d1 k

AB + B A B2 k a2

d2 k

Formation of AB complex

Formation of AB2 complex

This is an example of a ‘linked’ interaction, where the formation of AB2 cannot proceed before the formation of AB, and AB cannot

dissociate before the dissociation of AB2 Also, Octet software will calculate KD1 but not KD2 due to

unknown concentration of surface AB complex

Use Capture Approach To Deal With Bivalent Analyte

Avidity issue, typically seen with antibody as analyte, slow off-rate observed

Antibody capture approach can be used for bivalent analyte, avoiding avidity

Y Target molecule

Bivalent molecule

Anti Human capture sensor surface

Sensor surface

Target molecule

Bivalent molecule Y

Complex Curves 1:1 Homogeneous Curves

Summary • It’s the shape of sensorgrams with respect to time that give

us the kinetic information • Curves should only be influenced by ka, kd and concentration • An understanding of the relationships between ka, kd, Req,

Rmax and KD can help understand the processes and design our kinetics assays

• Using the more complex models available in the Octet analysis software can help troubleshoot more complex curves and optimize the design of kinetics experiments

• ForteBio offers different chemistry biosensors for off-the-shelf use

• Octet can use up to 16 biosensors for scouting experiments, and each sensor is independent

Contact Info

• Pui Seto • ForteBio - A Division of Pall Life Sciences, Pall Corp • Email: Pui_Seto@Pall.com • office phone: +1.650.289.6861 • fortebio_support@pall.com or call 888-OCTET-75 • www.fortebio.com