Quantum Dots for Quantitative Flow Cytometry Protocol

Overview

Introduction of Quantum Dots for Quantitative Flow Cytometry Protocol

General Considerations

The use of standard calibration beads to define the quantity of fluorophore-tagged ligands or proteins on cells or beads represents an essential element of quantitative flow cytometry that has enabled the direct comparison of inter-laboratory data as well as quality control in clinical flow cytometry. Commercial standards, such as Quantum^TM FITC MESF beads, comprise a mixture of five distinct populations of beads, where the subsets are functionalized with discrete titers of fluorescein conjugates. The equivalence of fluorescence radiance of each bead population in the mix is correlated to the fluorescence of soluble fluorescein. Thus, the surface occupancy of fluorescein on the bead standards is known as the Molecules of Equivalent Soluble Fluorophores (MESF), where the MESF value is equal to the known number of molecules in solution. We have recently described a simple method of producing fluorescence calibration beads that bear discrete and defined quantities of quantum dots and demonstrated their applicability as standard calibration beads useful for multiple color calibration. The measurement model is based on a simple ratiometric formula, Eq. 1, for comparing the number of fluorophores of Samples 1 and 2. Sample 1 is an analyte capable of emitting radiation of intensity I1 that is equivalent to a single unit of Sample 2; Sample 2 is a calibration standard emitting radiation of intensity I2. For simplicity, both samples must be excited at the same wavelength and emit in the same spectral region as defined by the same bandpass filter. Equation 1 is derived from spectroscopic parameters: extinction coefficient measured at the excitation wavelength (ε_λ), quantum yield (ϕ), and the BP transmittance (%T) of sample emission. It enables one to determine the number density (ri) of sample fluorophores.

Quantum Dots for Quantitative Flow Cytometry

In Eq. 1, e can be measured from absorption spectra for most probes. The quantum yields ϕ_i can be determined by the end-user relative to known textbook standards.

The photophysical properties of qdots are not clearly defined like the intrinsic physical constants of molecular systems. Quantum dots are semiconductor nanoparticles with tunable optical properties that are strongly dependent on size. The spectroscopic properties of dots are dependent on their size dispersion, shape, and surface defects. Subtle differences in preparation can lead to batch-to-batch variation in basic spectroscopic properties, such as quantum yield. To the best of our knowledge, the commercial dots are not spectroscopically standardized nor are they well characterized in the product literature for them to be used "as made". In this chapter, we refer to the previous characterization of a pair of CdSe quantum dots with emission bands centered at 525 nm and 585 nm. The lot numbers are cited to emphasize the notion that the spectroscopic characteristics of these dots are strictly relevant to samples from the indicated lots. Our previous study of these samples showed an anomalous dependence of emission yield on excitation wavelength that was more pronounced in QD525 dots than in the QD585 quantum dots. This was attributed to size heterogeneity and surface defects in the QD525, consistent with characteristics previously described in the literature. It is therefore important to characterize new batches of dots before they can be used as calibration standards. Two lots of subsequently purchased QD585 dots were shown to have quantum yield values of 0.12 and 0.3, respectively.

Molecular Assembly of Qdot Fluorescence Calibration Beads

Qdots can be photo-excited at any wavelength that can be matched to a fluorophore of choice. Therefore, calibration standards can be assembled for any fluorophore by using qdots whose emission spectra overlap with the target fluorophore. The approach is based on the display of variable and discrete site densities of streptavidin-functionalized qdots on biotin functionalized beads. Streptavidin-functionalized quantum dots are decorated by up to eight streptavidin tetrameric units per dot. It is therefore difficult to control the valency with which each dot binds to biotin-functionalized bead. The problem of unregulated multivalency has been shown to significantly limit the effort to consistently produce discrete molecular assemblies of qdots and beads at each titer of qdots. We have previously shown that an intervening layer of site directed biotin-labeled antibodies attached to beads circumvents the problem of multivalency by limiting the capture of a single dot to a single antibody in a reproducible and theoretically predictable manner. The molecular assembly on beads has been well characterized by us in terms of dissociation constants at each junction module. We have shown that titration of streptavidin-functionalized qdot to antibody bearing beads can yield discrete quantities of bound qdots in a manner consistent with the weakest equilibrium binding constant between the qdots and the antibody.

Principles and Practice: Applicability of Qdot Calibration Beads to Quantitative Flow Cytometry

Although calibration standards, such as Quantum™ FITC MESF beads, are strictly applicable to fluorescein-based probes, it is useful to present a simple correction factor that extends their applicability to other probes operating in the same spectral range. These include most fluorescein conjugates that tend to have lower quantum yields than the parent molecule where simple derivitization of fluorescein typically attenuates the yield by at least 18% (e.g., fluorescein biotin) or greater (e.g., 80% for FITC antibody conjugates (f/p ratio 4:1)). The correction factor (cf) that is derived from Eq. 1 is shown in Eq. 2.

Quantum Dots for Quantitative Flow Cytometry

Figure 1. Summary of the molecular assembly of quantum dot fluorescence calibration beads on streptavidin-coated beads.

Materials of Quantum Dots for Quantitative Flow Cytometry Protocol

Determination of the Relative Emission Yields of Quantum Dots

Streptavidin coated quantum dots QD525 and QD585, fluorescein,and rhodamine B.
Quartz cuvettes for absorption and fluorescence measurements.
Probes are solubilized in any of the working buffers listed below, in dilute solutions where the optical density is kept below 0.1 at the highest absorption wavelength.
Phosphate-buffered saline (PBS).
Tris buffer: 10 or 25 mM Tris–HCl, 150 mM NaCl, pH 7.5.
HHB buffer: 30 mM HEPES, 110 mM NaCl, 10 mM KCl,1 mM MgCl2, 10 mM glucose, pH 7.4.

Preparation of Qdot-Labeled Microbeads

Streptavidin-coated polystyrene particles (6.7 mm in diameter,0.5% w/v).
Streptavidin-coated quantum dots QD525 and QD585.
FLAG peptide (DYKDDDDK), bioM2 anti-FLAG antibody. BioM2 antibody contains significant quantities of free biotin and requires purification prior to use (see Note 1).
Biotinylated FLAG (FLAG^bio) and FITC-conjugated FLAG peptides.
PBS, Tris, and HHB buffers.
Tris and HHB buffers supplemented with 0.1% bovine serum albumin (BSA).
Soluble streptavidin; soluble biotin.

Equipment

Spectrophotometer such as Hitachi model U-3270.
Spectrofluorometer such as QuantaMaster™ Model QM-4/2005.
Flow cytometer such as BD FACScan.

Methods of Quantum Dots for Quantitative Flow Cytometry Protocol

Determination of the Relative Emission Yields of Quantum Dots

This protocol describes the measurement of relative emission quantum yields of quantum dots. The quantum yield of QD525 is measured relative to fluorescein, and QD585 is measured relative to rhodamine B. These dots are selected because their spectral characteristics are most suitable for analysis on a standard BD FACScan. Several factors, including size heterogeneity and potential impurities, have been shown to cause anomalous dependence of the quantum yield on the excitation wavelength when dots are excited at different wavelengths, in apparent violation of Kasha's rule: that the energy and yield of the emission should be independent of the excitation wavelength. It is therefore useful to collect a photoexcitation spectrum of any new batch of quantum dots to determine characteristic emission behavior over the putative spectral range of analysis. The protocol described in this chapter is applicable to measurements made using instruments comparable to the ones utilized here.

Measure absorption spectra of the probes following standard procedures. The molar absorptivity of all probe solutions should be at or below 0.1. In this limit, errors due to inner filter effects alone are on the order of 10%.
The slitwidths of the excitation and emission monochromators must both be set at 2 nm. Narrow slitwidths allow for better resolution of size dependent emission of poly disperse sized dots.
Use appropriate wavelength cutoff and BP filters for stray light rejection.
It is useful to make measurements of complete emission spectra of probes by off-resonance excitation: at 450 nm for the green probes and 540 nm for the yellow probes. Make sure that the molar absorptivity of the qdots and standard sample are matched at the excitation wavelengths.
Use Eq. 3 to calculate the relative quantum yields (ϕ_s) of QD525 and QD585 using fluorescein and rhodamine B as standards.
Use the integrated intensity of the emission spectra of sample relative to fluorescein (ϕ_ref = 0.93) or rhodamine B (0.31 in water). Is and I_ref are the integrated band intensities. The optical densities (OD) of the sample (s) and reference (ref) are taken at the excitation wavelength. n is the index of refraction of the solvent; 1.32 for water. For qdots, ϕ_S is accurate for the given lot number and excitation wavelength; therefore, lot number and source should always be identified.
Measure the emission intensity of all probes exciting at 488 nm (the primary excitation wavelength in a BD FACScan flow cytometer) and calculate relative quantum yields using Eq. 3.
To collect photoexcitation spectra, set the wavelength for measuring the emission intensity at the peak or along the red edge of the emission spectrum. Collection at the emission peak allows for greater sensitivity while collection along the red edge allows for a broader spectral window.
The excitation spectra of fluorescein and rhodamine B are expected to match their corresponding absorption spectra, provided the photoexcitation spectra have been corrected for the spectral distribution ( L_λ) of the lamp. L_λ depends on the characteristics of the lamp and monochromators. Most modern instruments automatically perform this correction provided it is enabled and properly set up. In cases where a spectrofluorometer instrument lacks the capability to automatically perform a correction of L_λ of excitation spectra, this can be done manually (see Note 2). This is an important consideration as significant distortions can be observed in the uncorrected photoexcitation spectrum.
Photoexcitation spectra of qdots measured relative to standards are useful for determining the quantum yield per unit wavelength of qdots over a desired spectral window.
Measure %T. Overlay the measured emission spectra with the appropriate BP filter and the calculated fraction of the emission band that overlaps with the BP filter is %T.

General Preparation of Qdot-Labeled Microbeads

Preparation of M2 Beads

Conceptual details can be found in the original literature. Much of the work involves the assembly of bioM2 antibodies on beads. The M2 antibodies are critical for achieving an effective monovalent binding interaction. This is an empirical finding from centrifugation assays in which it has been shown that the number of antibody-captured qdots is equal to the site density of antibodies on the same beads. Based on earlier studies, this chapter makes the following assumptions about the stock suspensions of 6.7 mm beads: (1) there are typically 40,000 beads/mL; (2) there are ≈4 million saturable bioM2 binding sites on these beads as determined from centrifugation binding assays of a fluorescently labeled FLAG peptide. Nevertheless, the reader should always verify particle counts by using a hemocytometer or particle counter.

Use a pipette to extract the desired volume of beads from a vigorously shaken stock suspension. Centrifuge beads at 14,000 × g and resuspend in desired buffer (Tris or HHB) containing 0.1% BSA in a minimum volume of 100 mL. To minimize nonspecific binding, the beads must be vortexed gently for 30 min to allow 0.1% BSA to coat beads.
Step a in Fig. 1a. Add bioM2 antibody at the desired stoichiometry: ≥2 bioM2:1 total streptavidin sites on all beads (see Note 1). Vortex or shake for an hour at room temperature.
Centrifuge and wash beads twice to remove free bioM2. The two footed binding assures that the bioM2 is tightly bound to beads.
Step b in Fig. 1a. Commercially available biotinylated M2 antibodies have about three to four biotin conjugates on the Fc domain of the antibody according to manufacturer's data sheet. It is likely that two of the biotin groups are cis-bound to the streptavidin on the beads. To prevent the extra biotins from subsequent binding to streptavidin-functionalized quantum dots, add a stoichiometric excess of soluble streptavidin to M2 beads, and allow to vortex mildly at room temperature for 15 min. Wash the streptavidin-saturated M2 beads once in buffer containing 0.1% BSA and centrifuge at 14,000 × g for 2 min.
After the removal of the supernatant in step b, add 80 mL of 1 mM soluble biotin to the bead pellet, and spin down immediately after brief vortexing. Wash five times by repeated centrifugation and removal of supernatant, and finally resuspend in 100 mL of buffer. Soluble biotin is added in order to saturate excess biotin-binding sites on the bound streptavidin to prevent undesired binding of biotinylated FLAG peptide used to tether qdots to bioM2.
Depending on the stock of bioM2 beads, aliquot samples into 25 mL volumes (or 1X10⁶ beads) for long-term storage (see Note 3).

Preparation of Quantum Dot Calibration Beads

An aliquot of 1 × 10⁶ bioM2 beads is first saturated with 100 nM biotinylated FLAG peptide at room temperature for 1 h under mild vortexing, and washed two times in Tris–BSA buffer by repeated centrifugation and removal of supernatant. Resuspend in 100 mL of buffer. It is important to note, that washes of the beads does not confer significant loss, by dissociation, of the peptide as one might expect. In cells or beads, where the density of receptors is relatively high (e.g., >1 × 10⁵ receptors/cell), the binding and dissociation of ligands is believed to occur in the "collisional limit" regime. In this limit, the probability of dissociated ligand rebinding to an adjacent receptor is much higher in terms of normal bimolecular mass action kinetics, where the ligand would normally diffuse away. Therefore, the eight million peptide binding sites on these beads are expected to show anomalous low loss of ligand.
Add 100nM streptavidin-coated Qdots (final concentration) to the beads, and incubate at room temperature for an hour. The qdots bind to the biotin end of the FLAG peptide. Wash the beads, resuspend them in buffer, and analyze them at the flow cytometer. The attachment of qdots and beads is governed by the bivalent FLAG/M2 interaction Kd = 2nM.

Discrete Labeling of Quantum Dots Calibration Beads from Mass Action Considerations

This protocol describes the application of mass action to the preparation of beads of known surface coverage. The quantity of bound qdots can be determined from a simple mass action model as shown in Eq. 4a where Q or [Q]₀, and Ab or [Ab]0 represent the initial concentrations of qdots and bioM2 sites on beads, respectively. [Q]_free is unbound qdots at equilibrium. v is the volume of buffer in which beads are suspended, n is the number of beads, and A is Avogadro's number.

Quantum Dots for Quantitative Flow Cytometry

In keeping with standard practice of generating fluorescence calibration beads, it is desirable to decorate four distinct populations of beads with surface concentrations of quantum dots that span a wide range of surface occupancy. For convenience, here we describe the synthesis of a four-component set of fluorescence calibration bead populations decorated with discrete qdots surface densities ranging from 6 × 10⁴ to ≈ 2 million. Using 1 × 10⁶ bioM2 beads in 100 mL volumes yields 66.0nM Ab. Use Eq. 4b to establish the desired surface densities. The reader might find it useful to realize that, based on mass action considerations, the limiting reagent [Q]₀ is depleted by Ab, and in the concentration range of interest, [Q · Ab] ≈ [Q]0. The reader can show that the error in this assumption is within 10% when [Q]₀ ≤ 0.7[Ab]₀.

To establish a good distribution in surface occupancy of qdots, one might consider to space [Q]0 titer intervals at half log, i.e., [Q]0 = 1, 3, 10, 30 nM to yield [Q · Ab] = 0.97, 2.9, 9.6, 28.0 nM corresponding to 5.8 × 10⁴, 1.75 × 10⁵, 5.84 × 10⁵, 8.76 × 10⁵, 1.75 × 10⁶ qdot/bead.

Add 25 mL of bioM2 beads (1 × 10⁶ beads in total) to 25 mL of 20 mM FLAGbio and 50 mL of Tris buffer, mildly vortex the mixture at room temperature for 1 h.
Wash the beads twice in Tris–BSA buffer by centrifugation and resuspension (at max. speed, 14,500 × g).
Resuspend the beads in freshly made (see Note 3) qdot solution in Tris–BSA buffer, mildly vortex at room temperature for at least an hour (see Note 3).
Centrifuge the QD-M2 calibration beads at max speed for 90 s and remove 60 mL of the supernatant. Check beads on flow cytometer.
Discrete histograms of narrow distribution are a good indicator of a well-regulated assembly (see Note 3). Plot mean channel fluorescence (MCF) versus Qdot/Bead. A linear plot (R2 = 0.99) is a good indicator of a well prepared set of calibration standards.
Store the beads at 4°C and use for up to 1 month (see Note 3).

Standardization of Flow Cytometry Data with Fluorescence Calibration Beads: Correction for Spectral Mismatch, Quantum Yield Between Sample and Standard Calibration Beads

It is useful to demonstrate the applicability of qdot calibration beads relative to Quantum™ FITC MESF calibration beads. This is achieved through Eq. 2, which establishes a correction factor that is used to account for the differences in the measurement parameters, e, ϕ, and %T, between any given probe and the calibration standard.

Quantitative Measurement of Qdot-Labeled Virus Particles Bound to Cells Using Flow Cytometry and TEM

Virus particles are amenable for tagging with qdots. Once labeled, the degree of labeling can be resolved by transmission electron microscopy (TEM). In this way, flow cytometry can be used to quantify bound viruses. Human papillomaviruses (HPVs) are etiological agents of a number of benign and malignant tumors of the skin and mucosa. We have recently analyzed the binding of QD585-labeled HPV particles to A431 cells by flow cytometry. QD585 calibration beads were used to measure the total number of qdots per cell ≈ 12,000 QD585 per cell, using Eq. 1. In previously unpublished data from that study, TEM was used to estimate the number of qdots per virus particle.

Notes of Quantum Dots for Quantitative Flow Cytometry

Commercial bioM2 is contaminated with high levels of free biotin, which is very likely to block binding of bioM2 to beads. This is removed by using a 100 kDa cutoff Microcon Centrifugal Filter Device.
(a) Assuming a starting material of 500 mL of anti-FLAG bioM2 antibody (in 50% glycerin, 1 mg/mL). Add to centrifugal device and centrifuge in cold room at 3,000 × g for 25 min (or till there are<50 mL).
(b) Wash the antibody in 25 mM (or 50 mM) Tris buffer 3× by repeating Note 1a. It should be noted that glycerin is very hard to filter through, therefore longer centrifugation time and/or additional washing steps might be needed.
(c) Resuspend the purified and concentrated bioM2 in 100 mL of 25 mM Tris buffer (5 mg bioM2/mL), make 20 mL aliquots and keep frozen in −20°C freezer till needed.
(d) Assuming 4 × 10¹⁵ molecules of bioM2/mg: each aliquot of bioM2 (0.1 mg) can coat 25 million 6.7 mm streptavidin bead (1 × 10¹⁴ biotin binding sites) with 2× excess of bioM2.
Correcting excitation spectra for spectral distribution of the lamp. Uneven features in spectra are due to the uneven spectral distribution (L_λ) of the lamp. To extract (Lλ) easily:
(a) Overlay the normalized absorption spectra and the excitation spectra. Because emission almost always originates from the lowest vibrational level of the first excited state, the emission yield is independent of excitation wavelength for most organic fluorophores. Therefore, one expects a resemblance between the normalized absorption and photoexcitation spectra of fluorescein. The excitation lineshape of fluorescein retains the same basic features that mirror the absorption spectrum except for structural artifacts introduced by the uneven spectral distribution of the Xenon arc lamp.
(b) Divide the normalized excitation spectrum column with the normalized absorption spectrum column. The result, L_λ.
(c) Divide all of the uncorrected excitation spectra columns with L_λ. The corrected data.
Qdots and bioM2 antibodies have limited shelf lives. Although Invitrogen Qdots are suspended in a proprietary buffer environment that confers fluorescence, when qdots are diluted in a regular buffer, they gradually lose fluorescence intensity over time. It is therefore important to maintain qdots in this buffer as long as possible. If calibration beads are not stored in this buffer, then it is important to frequently check the fluorescence intensity of the beads. This can be done by comparing the fluorescence intensity of dots left in regular buffer with a fluorophore of known concentration over time. The reader might also consider making minimal quantities of Qdot beads. In this way, long-term storage of qdots outside their native buffer is avoided. BioM2 antibodies also lose activity–30% over 6 months. The activity of bioM2 beads can be assayed using FITC-labeled FLAG peptide and commercial (Quantum™ FITC MESF) standard calibration beads to check on the integrity of the bioM2 beads over time. The typical assay involves periodic checking of the surface density of FITC-labeled FLAG peptide on bioM2 beads using Quantum™ FITC MESF beads or a centrifugation assay. The narrow distribution of histograms associated with each titer of qdots mixed with bioM2 beads can be used as an indicator of the quality of calibration beads. Broad histograms associated with freshly made calibration beads usually indicate: (a) Prevalence of nonspecific binding. This can be checked by running a negative control, i.e., adding qdots to bioM2 beads without the biotinylated FLAG peptide; (b) Heterogeneous binding of qdots due to loss of bioM2 activity.