Immunostaining and fluorescence-activated cell sorting _FACS_ analysis by pengtt


									Supplemental Methods

Immunocytochemical analysis of Hsp90 expression. T24 cells were cultured overnight

on cover slips in complete medium. The coverslips were washed with PBS and

processed either for surface Hsp90 or for total Hsp90 expression. For staining of surface

Hsp90, live cells were incubated with mouse anti-Hsp90 antibody in PBS for one hour at

37o, washed with PBS x3, fixed in 0.4% paraformaldehyde in PBS for 30 min at 37o,

washed with PBS x3, incubated with Alexa 594-conjugated anti-mouse antibody (1:100

dilution, cat. #A11005, Invitrogen/Molecular Probes), washed with PBS x3, incubated

with Hoechst 33342 (1:1000, cat. #H3570, Invitrogen/Molecular Probes) for 5 min,

washed with PBS x3, washed with distilled water x1, and mounted using Molecular

Probes mounting solution (SlowFade Gold, cat. #S36936, Invitrogen/Molecular Probes).

To confirm maintenance of membrane integrity following this fixing protocol, control

experiments were performed using Alexa594-conjugated anti-tubulin antibody. These

experiments demonstrated that no intracellular tubulin was detected using this protocol

(data not shown).

       For analysis of total Hsp90 expression, cells on coverslips were fixed with 0.4%

paraformaldehyde in PBS for 30 min at 37 o, washed x3 with PBS, permeabilized with

0.4% Triton X-100 and 0.1% bovine serum albumin in PBS for 5 min at room

temperature, washed x3 with PBS, incubated with mouse anti-Hsp90 antibody (1:100

dilution) for 1 hr at room temperature and washed x3 with PBS. The cells were then

stained with Alexa 594-conjugated anti-mouse immunoglobulin and counterstained with

Hoechst 33342 as above.

Image analysis. Cells were viewed using a Leica DM IRB fluorescence microscope

equipped with a Z-axis motor (Ludl Electronics, Hawthorne, NY). Stacks of images (11-

15 optical sections at a step size of 0.2 m) were taken with a digital camera

(Hamamatsu) and processed using Openlab Volume Deconvolution software

(PerkinElmer/Improvision, Lexington, MA). A representative set of 9 images of either

cell surface or total Hsp90 were selected for display in Fig. 2. Colocalization analysis of

total Hsp90 (Fig. 2j-r) and the nuclear stain Hoechst 33342 was performed using the

Colocalization software module from Improvision. Statistical analysis of the data

demonstrated less than 1% colocalization in each optical slice (data not shown).

Fluorescence-activated cell sorting [FACS] analysis of cell surface Hsp90. Cell

surface Hsp90 was detected on fixed cells using Hsp90 antibody. Briefly, cells were

cultured in LabTekII chamber slides [Nalge Nunc, Rochester NY], fixed in 10 % [w/v]

formaldehyde in PBS for 2 min, and washed three times in PBS. After blocking with 10

% FCS and 0.1 M L-lysine, cells were incubated overnight at 4 °C with Hsp90 primary

antibody, followed by incubation with Allexa 488-conjugated secondary antibody. Cells

were scraped and resuspended in PBS, and samples were analyzed with a FACSCalibur

instrument [Becton Dickinson, Franklin Lakes NJ]. 10,000 cells were analyzed for each


Analysis of polymerized actin formation and focal adhesion assembly by

immunofluorescence. Cells were serum-starved for 24 h and then detached from culture

plates with PBS containing 2 mM EDTA, washed twice in culture media, and re-

suspended in culture media. While in suspension, cells were treated with Hsp90 inhibitor

for 30 min, and then re-plated onto fibronectin [BD Biosciences] coated dishes, and

further incubated for 4 h. Cells were fixed, permeabilized, and incubated with FAK

antibody [BD Biosciences], Alexa488 conjugated phalloidin [Invitrogen] and DAPI

[Sigma-Aldrich]. Cells were observed with a fluorescence microscope.

Supplemental Figure Legends

Figure S1. a. DMAG-N-oxide does not induce down-regulation of cell surface Hsp90.

Cell surface Hsp90 on T24 cells was visualized by FACS analysis as described in

Supplemental Methods. Exposure to DMAG-N-oxide (1 M) for 24 h did not alter Hsp90

cell surface expression [compare red and black tracings]. The shaded area represents

fluorescence obtained with fluorescently tagged secondary antibody only. b.

Simultaneous incubation with DMAG-N-oxide competes with FITC-GA for binding

to cell surface Hsp90. Cell surface Hsp90 was visualized as in panel a, but DMAG-N-

oxide was added in excess at the same time as FITC-GA.

Figure S2. Live cell incubation with Hsp90 antibody detects cell surface Hsp90. T24

cells grown on coverslips were incubated as described in Supplemental Methods and

were analyzed for either cell surface [a-i], or total Hsp90 [j-r]. Each image represents a

0.2 micron deconvoluted section taken at selected points from top surface to bottom

surface of the cell monolayer [see Supplemental Methods]. Note that surface Hsp90

positivity [yellow] is not apparent on all cells in panels a - i. In contrast, detection of

intracellular Hsp90 reactivity requires cell permeabilization, is far more intense, positive

in all cells, and more uniformly distributed throughout the cell. As discussed in

Supplemental Methdods, < 1% of total Hsp90 [panels j-r] colocalizes with the nuclear

stain DAPI.

Figure S3. FITC-GA is not cell-permeable. The effects of FITC-GA on cell motility

[a] and Akt destabilization [b] were determined as in Methods following 24 h incubation

with T24 cells. Although FITC-GA dose-dependently inhibited cell motility, it had no

impact on Akt protein stability, suggesting that FITC-GA is poorly cell-permeable.

Figure S4. DMAG-N-oxide disrupts ECM-induced focal adhesion and stress fiber

formation. T24 cells plated on fibronectin [FN] in the absence [upper panel] or presence

[lower panel] of DMAG-N-oxide [1 M] were fixed, permeabilized and incubated with

FAK antibody [BD Biosciences; to detect focal adhesions; yellow/orange dots] and with

fluorescently conjugated phalloidin [to detect F-actin polymers; green linear fibers].

Nuclei are falsely colored pink for ease of visualization. Cells were examined by

fluorescence microscope. Note the concomitant disappearance of parallel F-actin stress

fibers and focal adhesions in the presence of DMAG-N-oxide.

Figure S5. Sub-optimal concentrations of DMAG-N-oxide and Hsp90 antibody

result in additive inhibition of cell invasion. The in vitro invasiveness of T24 bladder

carcinoma cells was examined as described in Methods. Effects of DMAG-N-oxide and

Hsp90 antibody, examined separately and together, are displayed as the mean +/-

standard error of 3 experiments.

Figure S6. Chemical structures of the small molecule Hsp90 inhibitors used in this

study. The structures of GA, 17AAG, 17DMAG, DMAG-N-oxide, and FITC-GA are

shown for comparison. All modifications are made to the 17 position of the

benzoquinone moiety of GA.


To top