From: Ed Luther (edluther@comcast.net)
Date: Sat Nov 22 2003 - 19:14:37 EST
Hello Stephen and responders to this thread. It seems to me that you are trying to obtain morphological data (implying some degree of resolution) from a flow system (a zero resolution system). Techniques such as FRET are available, but they are technically challenging. Why not try an imaging based approach? I agree with your time estimate for doing confocal microscopy, basically ruling that out. Howard, in a forthcoming follow up message proposes sorting and then visualization (tedious) and also the new "imaging in flow" cytometers, which he feels do not have a high enough throughput. An alternative laser scanning cytometry approach could be considered. This is essentially flow cytometry analytical principles applied to cells attached to either microscope slides, or microtiter plates, chamber slides, etc. The quantitative aspects of flow cytometry are kept intact, but the technology is laser scan imaging based, so that morphological information is included in the automated analysis, as well as the ability to visualize any of the cells from the analysis. Laser scanning cytometry is well suited for cell cycle analysis. A developed example is an in vitro micronucleus assay that parallels what you are trying to accomplish. The cells are segmented based on protein fluorescence, nuclei are sub-segmented within the cells. This technique calculates the number of nuclei within the segmented cells using an image processing watershed algorithm, quantifies the constituents (such as DNA content or area, immunofluoresence markers, etc. ) of the nuclei within a cell, restricts analysis to cells with proper DNA ploidy, and then identifies cells that contain micronuclei. The same techniques could be modified to tell you if a cell is a heterokaryote, and then if it has one or two nuclei. The 1:10000 event of interest rate that your numbers translate to is well within our throughput capabilities. If you wish to learn more about this approach please feel free to email me. Ed Luther Principal Scientist Strategic Research Development CompuCyte Corporation ----- Original Message ----- From: "S. Sullivan" <sjs93@cam.ac.uk> To: cyto-inbox Sent: Thursday, November 20, 2003 5:53 AM Subject: Re: [Discriminate fused cells and fused nuclei] > > > > ---------- > > From: S. Sullivan[SMTP:SJS93@CAM.AC.UK] > > Sent: Thursday, November 20, 2003 5:53:44 AM > > To: Cytometry Mailing List > > Subject: Re: [Discriminate fused cells and fused nuclei] > > Auto forwarded by a Rule > > > Nick, Howard et al., > > I am fusing murine ES cells with somatic cells....on average 10% of the > cells fuse. Using a selection regieme, I know at least 1 cell hybrid is > formed per 1000 heterokaryons. > > Nick's question regarding the formation of cell hybrids from > heterokaryons....the heterokaryons need to undergo a mitotic event to form > hybrids...DNA replicates, the nuclear membrane is broken down and the > chromosomes divided and two tetraplod hybrid cells are formed from the > heterokaryon. The cell hybrids are genotypically stable (tetraploid) over > many passages. > > I am not concerned with the nuclear protein make up becoming homogenised > over time as I will be sorting very shortly after fusion. > > So when I am looking for hybrid cells, I will have an excess of > heterokaryons....using nuclear specific dyes and confocal microscopy would > take me forever. > > Please note that the system does not have to sort hybrids from > heterokaryons with 100% efficiency. Even if I could reduce the > heterokaryon:cell hybrid ratio by a factor of 10 this would be very useful. > > Stephen
This archive was generated by hypermail 2.1.6 : Thu Jan 01 2004 - 17:44:04 EST
