discordant results from flow and microscopy

From: Lynn Dustin (dustinl@mail.rockefeller.edu)
Date: Tue Jan 15 2002 - 10:38:29 EST


Hello out there in flow land,

I am trying to help a colleague who is examining expression of a cell
surface (endogenous) marker and a cytosolic (viral protein) marker. She
got different results by flow from those obtained by microscopy.
Expression levels of the viral protein vary from cell to cell in these
clonal populations (for interesting reasons which I am not at liberty to
discuss). By microscopy, she sees two distinct cell populations; the
expression of the cell surface and cytosolic antigens are mutually
exclusive. By flow, we expected to see the same 2 cell populations.
Instead, there was a single staining pattern and most cells appeared to
be stained for both antigens.

We have confirmed that this is not a compensation issue. There does not
appear to be a problem of background staining or a cross-reactive
secondary antibody. We do not get this result with cells (different
clones, derived from the same parental line) that do not express the
viral protein.

We are considering various technical reasons for this discrepancy. If
any of you have experience in this area, your
comments and suggestions are most appreciated.

Possible technical sources of trouble?
1. The cells are adherent. For microscopy, she grew them on coverslips.
For flow, she grew them to the same density on plastic, and removed them
by two different methods: scraping or trypsin. Either method gave us the
same problem. She prefers to scrape the cells to avoid loss of the
surface marker.

2. Both of the antigens are stained indirectly (unlabeled primary,
fluorochrome-conjugated secondary). The antibody combination for the
surface antigen is unchanged. For the cytosolic antigen, a mouse primary
is detected with Texas Red-conjgated 2° for microcopy, but a
PE-conjugated 2° for flow.  We do not have much background staining, as
demonstrated by staining a matched clone that does not express the
target protein.

3. Fixation and permeabilization: for microscopy, the cells were grown
on coverslips, fixed with 3.7% formaldehyde, and stained in the presence
of 0.1% Triton X-100. For flow, the scraped or trypsinized cells were
fixed with 2% paraformaldehyde, then permeabilized, stained, and washed
in the presence of  0.1% saponin. In each case, the protocols were
chosen because they have worked for that particular application in other
experiments.

4. Gating: Cells were filtered before running on the cytometer. We were
able to detect cell debris, single cells and small aggregates (as judged
by forward and side scatter, and checked on the microscope). We gated
separately on the single cells and on the aggregates, with the same
results.

5. Antigen distribution: When examined by microscopy, the cytosolic
viral antigen is localized in discrete regions of the cell. The punctate
distribution may make it easier to see small amounts of antigen. The
membrane antigen, by contrast, is distributed all over the cell surface.

The cytometer is a FACSCalibur set up to use 4 colors (we routinely use
Alexa 488/fluorescein, phycoerythrin, PerCP, and Allophycocyanin at the
same time). The microscope is a Nikon Eclipse TE300 with filters for
fluorescein or Alexa 488, Texas Red, and Hoechst.

Sorry for the long-winded letter. Any ideas on how to sort this out?
Many, many thanks in advance.
**********************************************
Lynn B. Dustin, Ph.D.
Center for the Study of Hepatitis C
Rockefeller University
Box 64
1230  York Ave.
New York, New York 10021
Phone: 212-327-7067
FAX: 212-327-7048
email: dustinl@mail.rockefeller.edu



This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:59:20 EST