Andrew Beernink wrote- >Has anybody sorted virus? Any suggestions re: 1) safety, and ) >labelling/staining techniques? I'm assuming you'd trigger off a >fluorescence signal (Ab label), with a Hoe342 or other xNA "counterstain" to >discount drops with multiple particles. > >This would be done on a MoFlo (488, I70 Spectrum, I302), if it's doable! I recently had an e-mail exchange with Stefan Andreatta <stefan.andreatta@uibk.ac.at>, at the University of Innsbruck, on this topic; he was detecting (and possibly sorting) viruses on a MoFlo, and Matt Ottenberg of Cytomation tells me that Stefan gave a nice presentation on the subject at the European MoFlo Users' meeting a few weeks ago. BUT- My colleagues and I reported (in 1979) detecting scatter signals from individual virions on a specially built flow cytometer with high light collection, very low background, and a flow rate 1/100 of that in a conventional flow cytometer. The observation volume, with a 2 um beam height and a 2 um core diameter, was only a few femtoliters. In theory, we should have been able to detect staining of virus by a few dozen antibody molecules, but there were problems with the label and this was in the days before monoclonals were available, so background fluorescence from the 99+ per cent of antibodies in a polyclonal antibody that did not react with virus would likely have been intolerable. The apparatus would almost certainly have worked with phycoerythrin-labeled monoclonal antibodies; indeed, a much simpler, conventional flow cytometer now in my lab can reliably distinguish lymphocytes bearing just over 100 PE-labeled (1:1) antibodies from background. In fact, the slow flow instruments built at Los Alamos, at Twente University in the Netherlands, and at CalTech, for sizing nucleic acid fragments, should have even higher sensitivity; Rob Habbersett at Los Alamos has detected single molecules of PE. Commercial instruments are a different story. The second paper on detecting individual virions in flow (1999) used Molecular Probes dyes such as SYBR-1, which enhance fluorescence several thousand times on binding to nucleic acid; this made the virions barely detectable in fluorescence on instruments such as the FACScan and FACSCalibur, which probably collect light about six times as efficiently as a MoFlo. Stefan Andreatta, as far as I know, was using similar nucleic acid stains, and, with the proportionally higher excitation power available on the MoFlo, apparently managed to detect virions. However, the particles in this case are undoubtedly bearing at least a few thousand highly fluorescent molecules, whereas it seems unlikely that, without really large scale fluorescence amplification, you could get a signal detectable in a MoFlo (or even a FACScan) from no more than a hundred or so antibodies which you would expect to be bound to a single virion. Hoechst dyes are not a good choice for a nucleic acid stain or counterstain in this context; they only enhance fluorescence a hundredfold or so on binding to DNA, and the fluorescence signal from free dye in solution would likely swamp whatever you got from virus. The most informative paper that I know about to date on detecting virions in flow, again using SYBR type dyes, is Brussaard CPD, Marie D, Bratbak G: Flow cytometric detection of viruses. J Virol Methods 2000; 85:175-182; these authors examined a number of viruses including bacteriophages, HSV, and HIV. Virus sorting, probably with a slow-flow microfluidic sorter (to eliminate aerosol generation problems which, it now seems, apply to mail sorters as well as cell sorters) is one of the projects I envision for my Center for Microbial Cytometry, about which a followup to my mailing earlier this year should appear in the coming weeks. If Andrew can tell us what kind of viruses he wants to sort and for what characteristics (assuming it's not a trade secret), a strategy might emerge. -Howard
This archive was generated by hypermail 2b29 : Sun Jan 05 2003 - 19:01:36 EST