 |
 |
|
Polychromatic Flow Cytometry on the FACS Aria: The Emory Vaccine Center Strategy The Emory Vaccine Center has recently acquired a FACS Aria, which is up and running in our BSL-3 laboratory. This document is meant as a brief introduction to the Aria, and is particularly focused upon the development of polychromatic flow cytometry applications on this instrument. This document will evolve and grow, and I welcome any feedback by email. Note that parts of this discussion also apply to our MoFlo cell sorter. Table of Contents
With the introduction of new instruments capable of measuring >8 fluorescence parameters such as the FACS Aria and LSR-II cytometers from Becton Dickinson, and the Cyan from DakoCytomation, the instrument manufacturers have in some sense leaped ahead of the available reagents for these systems. Of course, the reagent manufacturers are working hard to fill the pipeline with reagents that can take advantage of these systems. Check out the links below, and contact your sales reps for more up-to-date information.
It is also possible to prepare your own reagents to take advantage of these instruments. Mario Roederer's web site has all of the protocols, and is the bible in this area. The base configuration of the FACS Aria has 9 fluorescence detectors on three lasers, with the possibility of adding 4 more. The FACS Aria at the VRC is equipped with the full 13 fluorescence detectors, as shown in the diagram below.
The diagram, taken from the BD FACS Aria Users Guide, shows optical filters for the base configuration. Note the absence of optical filters on position C on each of the trigons, and in positions G and H on the octagon; PMTs in these slots—indicated in the lettered circles are also missing in the base configuration. In contrast, the FACS Aria at the Emory Vaccine Center is equipped with PMTs in all positions, though we have had to order filters from a third party (filters are available from both Omega and from Chroma). The two most common conventional dyes that will be useful on the violet laser (the left-most trigon) are Alexa 405 and Alexa 430. Both are directly conjugated to antibodies using methods common to N-hydroxysuccinimidyl esters (NHS). Note that these dyes are almost exactly equivalent to Cascade Blue and Cascade Yellow, respectively.The normalized emission spectra of the Alexa dyes is shown in the figure below. Note that I am not displaying the excitation spectra, but these can be obtained on the BD Spectrum Viewer.
The filters included with the instrument were chose for the features of these dyes, and we see no reason to change them. Alexa 405 is an excellent dye for use with the violet laser; it is reasonably bright, it is not excited by the 488 nm laser, and it requires little compensation in other channels. In contrast, although Alexa 430 works, it has a number of drawbacks: (1) it is only suboptimally excited by the violet laser; (2) it is not very bright; (3) it is significantly excited by the 488 nm laser in addition to the violet laser; (4) it has a very broad emission spectrum and requires extensive compensation against FITC, PE, PE-Texas Red (or PE-Alexa610), and even against PE-Cy5 (or PE-Alexa647). I'm guessing that it would be the first label that I would drop if I didn't need all of the channels of the cytometer for an experiment. If it is used, it should be used on an antibody that binds with high affinity to a highly expressed cell surface protein, such as CD8. Finally, we will be developing applications of Quantum Dots that we think may prove very useful in conjunction with the violet laser. If we are successful, the filter configuration on our violet laser will change significantly. Quantum dots can be synthesized to emit light across the visible spectrum. The current configuration for the blue laser will accomodate labels as in the following table:
We will be attempting to add a sixth color to this panel, as indicated in the following table:
Note that we are replacing the Cy5.5 tandems in the first table with the Cy5 tandem (or Alexa 647 tandem), in order to make more room between it and the PE-Cy7 (or PE-Alexa750) tandem. In between these two, we will attempt to use the PE-Alexa700 tandem. The emission spectra for the PE-Alexa dye tandems are shown in the figure below (note that for the Alexa700 and Alexa750, only the spectra for the small molecule dyes themselves were available).
In addition to buying what we can, we are preparing our own custom tandem dyes, using only a slight modification of the methods found on Mario Roederer's website, for preparation of the PE-Cy5 tandem. Our only modification is that we use "Reaction Buffer B" (100 mM carbonate, pH 8.4) during the modification of the phyobiliprotein with the NHS-derivatives of the Alexa dyes, instead "Reaction Buffer C" (500 mM carbonate, pH 9.0) that Mario uses for coupling of the cyanine dyes, which use different reactive group chemistry. I have not systematically evaluated the performance between the two buffers for the Alexa dye couplings, but we have verified that buffer B works, and we note that pH 8.3-8.4 buffers are often recommended for NHS couplings (for example, as in this PDF file from Molecular Probes). Three more notes about the tandems.
The current configuration for the red laser will accomodate labels as in the following table:
The original publications from Roederer and colleagues (1, 2, 3) on polychromatic flow cytometry used the following three labels run off of a red laser: APC, APC-Cy5.5, and APC-Cy7. We and others have found that it is possible to replace the APC-Cy5.5 tandem with a direct antibody conjugate of Alexa680. Although the excitation spectrum of Alexa680 suggests that it would not be optimally excited by the 633 nm laser, we have found that it provides a reagent with excellent brightness, and therefore we do not believe that it is worth the effort to produce an APC tandem. We've also heard that it is possible to use direct conjugates to Alexa 700, and that the emission spectrum of these reagents might better "fit" between the APC and the APC-Cy7 (or APC-Alexa750) tandem. Setting up the Aria PMT Configuration On an instrument like the FACS Calibur which does not have interchangeable filters, the meaning of the parameter names FL1, FL2, FL3, and FL4 are always the same. On instruments with changeable filters such as the FACS Aria, FACS Vantage, and the MoFlo, such names are often meaningless, as are the detector names like A, B, C.... Continued use of these names will violate the principle that FCS files should contain within them all of the information necessary to interpret the data without cross-reference to a notebook. The FACS DiVA software that runs the Aria has an Instrument Configuration panel that looks like this (sorry for the low resolution):
You can store a large number of named Instrument configurations, as seen in the box at the left hand side. For each configuration, you can turn on 1-3 lasers, which are labeled Blue (488 nm), Red (633 nm), and Violet (405 nm). Each row in the three-column table to the right of the figure corresponds to a different parameter that you wish to measure; in the figure, the parameters are given familiar names such as FITC, PE, APC, etc., corresponding to the commonly used fluorophores. I believe that instead of using the familiar fluorophore names in the parameter field, that it would be much better to label the parameters with the specifications of the bandpass filter directly in front of the relevant detector. Since the same filters might be used on more than one laser (e.g. a 780/60 filter for PE-Cy7 and APC-Cy7), it will be necessary to also to use a shorthand indication of the laser used in the parameter name. The table below provides what might be a typical set of replacements.
In order not to confuse users, it would be good to sneak the fluorophore names back into the FCS files someplace, but where? A good place for this is in the Labels portion of the Experiment Layout dialog box, depicted in the figure below:
Each of the cells in the Experimental Layout dialog box contains two rows. The first row contains the Parameter name, as determined by the Instrument Configuration panel (see above). The second row contains the reagent name, such as CD8 as indicated in the blue-colored cells in the figure above. My suggestion is to use CD8-APC-Cy7 instead of CD8 as in the example above, providing a final cell that looks like this:
In my opinion, this will provide you with FCS files that do a much better job of containing all of the information relevant to the experiment. |
| Site by John Altman | Last Modified: Wednesday, December 24, 2003 |