Home > FCS Classroom > Assay Development > Choosing an FCS Assay Type > Autocorrelation vs. Cross-correlation Assays

Autocorrelation vs. Cross-correlation Assays

FCS Assays can be designed with one fluorescent molecule in autocorrelation mode (mass-dependent), or with two fluorescent molecules in cross-correlation mode (mass-independent), giving you the flexibility to use a broad range of dyes and study the interaction of molecules with similar molecular weights.

The QuantumXpert FCS Spectrometer has been specifically designed for solution-based FCS assays, and offers the flexibility of two excitation wavelengths (491 nm and 532 nm) and three emission detection channels.

The easiest biological interactions to monitor with FCS are binding interactions. FCS can be used to detect both direct binding and competition binding interactions, which can be used to quantify an analyte of interest.

This section briefly describes the properties of the two correlation assay types when measuring both forms of binding interactions and critical factors to consider when designing and implementing FCS measurements:

• Properties of Autocorrelation Assays
• Properties of Cross-correlation Assays

Properties of Autocorrelation Assaysback to top

General Properties of Autocorrelation Assaysback to top

The QuantumXpert can be used to obtain autocorrelation functions in three distinct emission channels. The recommended dyes for these three channels are:

• Channel A (510 nm to 525 nm): Alexa-488, GFP, Quantum Dots 525
• Channel B (550 nm to 575 nm): Alexa-532, R6G, Quantum Dots 565
• Channel C (590 to 710 nm): Cy3.5, Cy3.5b, Alexa-568, Texas Red, Quantum Dots 605

These fluorescent dyes have been selected because of:

• Minimal crosstalk between detection channels
• Resistance to irreversible photobleaching
• Resistance to triplet state excitation or intersystem crossing
• Availability in reactive form to label probes

Autocorrelation measurements provide information about diffusion time and aggregation state of the fluorescent molecules being studied. Binding is typically measured by monitoring the formation of bound complexes which result in a shift to slower diffusion times and decrease in average number of particles in the detection volume.

Critical Autocorrelation Assay Design Factorsback to top

• Quality of Fluorescent Probes: It is critical that the probes are purified to minimize free dye. Free dye in excess of 3-5% of total fluorescence will affect your ability to quantitatively analyze autocorrelation measurements.
• Molecular Weight of the Complex: It is critcal to consider the difference in molecular weight of the dye-labeled probe relative to the complex that is formed upon binding. In general, the molecular weight should be increased 3 to 5 fold when the freely diffusing probe binds to the target. In the event that the target and probe are of similar molecular weight, it is also possible to immobilize the probe on a larger particle to increase the molecular weight (e.g. carrier protein, bead or Quantum Dot), or to use a target with multiple binding sites for the fluorescently labeled probe.

Additional factors to consider when designing FCS assays can be found at: Optimizing FCS Measurement Conditions

Autocorrelation Assays of Direct Binding Interactionsback to top

Direct binding of a fluorescently labeled probe to a target will provide information about the total concentration of probe binding sites (B_max) and the binding affinity of the probe (K_D, or Equilibrium Binding Dissociation Constant).

where [L] is the concetration of labeled probe, B_max is the total binding sites, and K_D is the equilibrium dissociation binding constant.

Typical Procedure: Autocorrelation Assay of Direct Binding:

1. Determine diffusion time (τ_D) of the fluorescently labeled probe alone (Free Probe).
2. Determine if quantum yield for the labeled probe changes when it binds to the target. If the fluorescence intensity changes for the same probe concentration after the target is added, then a correction factor must be calculated in order to accurately calculate B_max.
3. Choose a single target concentration and incubate with a series of different probe concentrations that span the predicted range of binding affinity.
4. Measure the fluorescence correlation data of each sample with The QuantumXpert FCS Spectrometer. Only autocorrelation functions will be used in the analylsis.
5. Fit the data with FCSXpert software using either a one- or two-component diffusion model (see Choosing Between FCS Fitting Models). In the two-component fit, the fast diffusion time should be the Free Probe, and the slow diffusion time the Probe-Target complex.
6. Plot the Fraction Bound as a function of probe concentration.
   • Fraction Bound is calculated as fraction of slow diffusing component (F2) * the total number of diffusing particles (N) in the detection volume.
7. Both B_max and K_D can then be obtained from this plot by fitting with a hyperbola (saturation binding isotherm).

Autocorrelation Assays of Competition Binding Interactionsback to top

Competition binding assays allow you to determine the binding properties of an unlabeled analyte.

These assays are conducted by pre-incubating the labeled probe and target, then adding a series of unlabeled analyte preparations at different concentrations. The unlabeled probe will displace fluorescently labeled probe from the probe-target complex, providing information on binding affinity (K_i) and the fraction displaced (or fraction of labeled probe remaining in the complex).

Typical Procedure: Autocorrelation Assay of Competition Binding:

1. Determine diffusion time (τ_D) of the fluorescently labeled probe alone (Free Probe).
2. Conduct a Direct Binding Interaction Assay for probe and target, as described above, in order to obtain a K_D for this interaction.
3. Pre-incubate labeled probe and target until equilibrium is reached. This will form the Competition Complex that is used to quantify unlabeled analyte. It is assumed that the unlabeled analyte being characterized is a competitive inhibitor of the labeled probe used to form the Competition Complex.
   • NOTE: In order to achieve maximal sensitivity when binding unlabeled analyte, concentrations should be selected so that ~50% binding is achieved (i.e. near the K_D)
4. Incubate the Competition Complex with a series sample preparations that contain the unlabeled analyte.
5. Measure the fluorescence correlation data of each sample with The QuantumXpert FCS Spectrometer. Only autocorrelation functions will be used in the analylsis.
6. Fit the data with FCSXpert software using either a one- or two-component diffusion model (see Choosing Between FCS Fitting Models). In the two-component fit, the fast diffusion time should be the Free Probe, and the slow diffusion time the Probe-Target complex.
7. Plot Fraction Bound as a function of unlabeled sample concentration.
   • Fraction Bound is calculated as fraction of slow diffusing component (F2) * the total number of diffusing particles (N) in the detection volume.
8. IC₅₀ and Fraction Bound can be obtained by analyzing the Fraction Bound plot with the the equation below: where [I] is the concentration of inhibitor, IC₅₀ is the 50% inhibitory concentration, and p is the Hill Coefficient (slope factor) which should be 1 for simple binding.
9. K_i for the unlabeled analyte can be calculated from IC₅₀ using the Cheng-Prusoff Relationship: where K_D is the equilibrium dissociation binding constant for the labeled probe, [L] is the concentration of the labeled probe, and IC₅₀ is the 50% inhibitory concentration.

Properties of Cross-correlation Assaysback to top

General Properties of Cross-correlation Assaysback to top

The QuantumXpert can be used to obtain three cross-correlation functions. In cross-correlation, two different fluorescent labels with distinct excitation and emission properties are detected in two emission channels. Coincidence of these fluorescent labels on the same diffusing particle results in a cross-correlation signal. The QuantumXpert calculates the cross-correlation functions for:

• Channels AxB
• Channels AxC
• Channels BxC

The QuantumXpert also acquires cross-correlation data for BxA, CxA and CxB, and the FCSXpert Software uses this additional data to improve the counting statistics for the cross-correlation functions reported.

Cross-correlation Assays of Direct Binding Interactionsback to top

Direct binding of a fluorescently labeled probe to a target will provide information about the total concentration of probe binding sites (B_max) and the binding affinity of the probe (K_D, or Equilibrium Binding Dissociation Constant).

where [L] is the concetration of labeled probe, B_max is the total binding sites, and K_D is the equilibrium dissociation binding constant.

Typical Procedure: Cross-correlation Assay of Direct Binding:

1. Determine diffusion times (τ_D) of the two fluorescently labeled particles being studied. This may be a labeled probe and a labeled target, or two labeled probes that recognize and bind to the same target particle.
2. Determine if quantum yield for either of the labeled probes changes when it binds to the target. If the fluorescence intensity changes for the same probe concentration after target is added, then a correction factor must be calculated in order to accurately calculate B_max.
3. Choose a single labeled target concentration and incubate with a series of different labeled probe concentrations that span the predicted range of binding affinity.
4. Measure the fluorescence correlation data of each sample with The QuantumXpert FCS Spectrometer. Both autocorrelation and cross-correlation functions will be used in the analysis.
5. Fit the data with FCSXpert software using either a one- or two-component diffusion model (see Choosing Between FCS Fitting Models). In the two-component fit, the fast diffusion time should be the Free Probe, and the slow diffusion time the Probe-Target complex.
6. Plot fraction bound as a function of probe concentration.
   • Fraction Bound is the number of diffusing particles reported in the cross-correlation analysis panel.
7. Both B_max and K_D can then be obtained from this plot by fitting to a saturation binding isotherm.

Cross-correlation Assays of Competition Binding Interactionsback to top

Competition binding assays allow one to determine the binding properties of an unlabeled analyte.

These assays are conducted by pre-incubating the labeled probe and target, then adding a series of unlabeled analyte preparations at different concentrations. The unlabeled probe will displace fluorescently labeled probe from the probe-target complex and will provide information on binding affinity (K_i) and the fraction displaced (or fraction of labeled probe remaining in the complex).

Cross-correlation assays of competition binding can be developed using the principles outlined above.