| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Automation and Analytical Techniques |
1
Cliniques Universitaires Saint-Luc, Université Catholique de Louvain, Clinical Biochemistry Department, Clos-Chapelle-aux-Champs, 30-UCL 30.46, 1200 Brussels, Belgium.
2
Scientific Section of the Medical Staff, Belgium Armed
Forces, Queen Astrid Military Hospital, Rue Bruyn, 2, 1120 Brussels,
Belgium.
a Address correspondence to this author at: Laboratoire de Technologie Moléculaire Appliquée, Service de Biochimie Clinique, Clos-Chapelle-aux-Champs, 30-UCL/30.46, B-1200 Brussels, Belgium. Fax 32-2-764-39-59; e-mail gala{at}sang.ucl.ac.be.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Two commercial reagents, ORTHOPermeaFixTM (OPF; Ortho) and FIX&PERM Cell Permeabilization Kit® (F&P; An Der Grub Bio Research GmbH, Imtec) have recently been claimed to avoid these limitations. They have been shown to be suitable for the detection of many cytoplasmic (CD22, CD3, and myeloperoxidase) or nuclear (terminal deoxynucleotidyl transferase) antigens and can be used for the simultaneous detection of membrane and intracellular antigens (4)(5)(6)(7).
As an alternative, a very promising noncommercial method, based on microwave heating (MWH) treatment, has recently been reported to achieve the same performance with intracellular targets [CD68, proliferating cell nuclear antigen (PCNA), Ki-67] (2).
In this study, we compare OPF, F&P, and MWH, focusing on the detection of cytoplasmic (bcl-2 and gp-170) and nuclear antigens (p53 and rb/p105), all of which are proteins of interest in diagnostic and research areas: the bcl-2 gene product plays a key role in the inhibition of the programmed cell death or apoptosis; the glycoprotein gp-170, encoded by the MDR-1 gene, is the marker of classical multidrug resistance, for which the importance of quality-control procedures and standardization of detection methods have been emphasized recently (8). The latter two proteins, p53 and rb/p105, are the products of the corresponding tumor suppressor genes and are meant to play a key role in the control of the cell cycle. These four antigens, which are detectable by flow cytometry (9)(10)(11)(12)(13)(14)(15)(16)(17), have all been involved, alone or in association, with the occurrence of drug resistance. This makes interesting the use of a simple and standardized method that allows the simultaneous detection of cytoplasmic and intranuclear antigens. A panel of cell lines (CLs) constitutively expressing or lacking expression of those antigens were used as controls.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Positive and negative CLs (listed in Table 1
) were chosen for their differential expression of the four
antigens of interest.
|
p53.
The presence of a wild-type or mutated p53
gene was determined by sequence analysis of the whole open reading
frame in IM9, DOHH2, and DG75. Mutated p53 was found only in
DG75 (J-L. Gala, unpublished results). RAJI and CCRF-CEM both express a
mutated p53 protein (18).
rb.
rb/p105, the retinoblastoma gene product, is not expressed
in the negative control lymphoblastoid line IM9 (19). 6M1
and 6M2, which are IM9 subclones stably transfected with the control
plasmid pZipNeoSV(X), have been shown to lack expression of the protein
and were also included as negative controls (17).
Positive-control CLs expressing the rb gene included 3C3 and 5C3, which
are IM9 subclones transfected with pZipRb, a plasmid containing the rb
gene-coding sequence, and have been shown to express the gene product
(17). The rb-positive myeloid line KG1 was also included
(20).
bcl-2.
The Epstein-Barr virus-negative (group 1) Burkitt
lymphoma CLs DAUDI and DG75 do not express bcl-2 (21). bcl-2
is expressed in the Epstein-Barr virus-positive CL RAJI and the
subclone DG75 LMP, which is DG75 transfected with pSHISLMPd3, a plasmid
containing the latent membrane protein-coding sequence
(21)(22). bcl-2 is also overexpressed in
follicular lymphomas carrying the translocation t(14;18) and has been
found in DOHH2, a CL derived from such a lymphoma (23).
gp-170.
CCRF-CEM, derived from a T-cell acute lymphoblastic
leukemia, does not express gp-170, the marker of MDR-1. VLB8
and VLB100, obtained by culturing CCRF-CEM in progressively
increasing concentrations of vinblastine, express gp-170; however,
expression is higher in VLB100 (24).
antibodies
Monoclonal antibodies (mAbs) recognizing intracellular antigens
included the following: p53 (clone DO7; Dako, Prosan) at a final
dilution of 1/100; bcl-2 (clone 124; Dako, Prosan) at a final dilution
of 1/70; rb (clone G3-245, Imtec) at a final dilution of 1/500; gp-170
(clone JSB-1; PharMingen, Imtec) at a final dilution of 1/60; and
PCNA (clone PC10; Dako, Prosan) at a final dilution of 1/50.
mAbs mapping external epitopes were also used and included the gp-170 monoclonal 4E3 (Imtec) and CD19 (clone HD37; Dako, Prosan) both used undiluted.
Two isotype negative-control antibodies, IgG1 (a gift from H. Bazin, Laboratory of Experimental Immunology, UCL, Brussels, Belgium) and IgG2a (clone DAK-GO5; Dako, Prosan) were used throughout the study.
procedures
Preliminary steps.
For each CL, the presence or absence of the
antigen assessed was confirmed by established methods, namely Western
blot and an immuno-alkaline phosphatase method (APAAP)
(17)(21)(24)(25)(26)(27). Western blot and
APAAP were performed with the same antibodies as for flow cytometry.
The only exception was gp-170 (MDR-1) detection by Western
blotting where C219 (Dako, Prosan) at a final dilution of 1/400 was
used in place of JSB-1, as described previously (24).
Immunoblotting was obtained from 1 x 10 cells,
and each blot was probed overnight at 4 °C with the respective mAb
and incubated for 1 h at room temperature with either
alkaline-phosphatase-conjugated anti-mouse antibody (DAKO, Prosan) for
p53, bcl-2, and gp-170, or horseradish peroxidase-conjugated anti-mouse
antibody (Amersham) for rb.
For p53 and bcl-2, the total cellular protein lysates were prepared by harvesting whole cells in hot Laemmli electrophoresis sample buffer as described (21)(27)(28). After immunoblotting on nitrocellulose membrane (Bio-Rad), the specific binding was visualized with 4-nitroblue tetrazolium chloride and 5-bromo-4-chloro-3-indolyl-phosphate (Sigma Chemical Co.).
For rb, whole cells extracts were separated by overnight electrophoresis in an 8% sodium dodecyl sulfate-polyacrylamide gel and transferred to an Immobilon poly (vinylidene difluoride) membrane (Millipore) as described (17)(26). The BM Chemiluminescence Western blotting kit (Boerhinger Mannheim) was used for detection according to the manufacturer's instructions.
For gp-170, cells were resuspended in 0.5 mL of lysis buffer (0.01 mol/L Tris-HCl, pH 7.4) and run on sodium dodecyl sulfate-polyacrylamide gel as described (24). Immunoblotting and detection were identical to the procedure used for p53 and bcl-2.
The autofluorescence of each CL was measured before further evaluation. The three methods were performed according to the manufacturers' instructions or as described (2). For each antibody, experiments were repeated three times, and each trial with a particular CL was performed in triplicate.
Evaluation of the microwave oven performance.
Although
microwave ovens are simple apparatuses, they can differ by their total
power capacity. To better characterize the performance of our system
and to allow other laboratories to set up an identical procedure, we
measured the time needed to boil 1000, 500, 250, 100, and 10 mL of
distilled water in our microwave oven (Philips M610) at the maximum
power setting (800 W).
The performance of our apparatus was as follows: 1000 mL of distilled water boiled at 12 min 9 s; 500 mL boiled at 6 min 22 s; 250 mL boiled at 3 min 35 s; 100 mL boiled at 1 min 52 s; and 10 mL boiled at 58 s.
fixation/permeabilization methods
OPF.
Cells (10) were washed in 1x
phosphate-buffered saline (PBS; 1.44 g/L KH2
PO4–90 g/L NaCl–7.95 g/L Na2
HPO4), resuspended in the OPF reagent, and then
treated according to the manufacturer's instructions. Briefly,
10 cells were washed in PBS and fixed for 45 min in 1.5 mL
of OPF reagent diluted twice, and then washed with PBS containing 15
g/L bovine serum albumin, 5 mg/L EDTA, and 50 mL/L fetal calf serum.
F&P.
Cells (10) were washed in PBS and
resuspended in 100 µL of reagent A (fixation) for 15 min, and then
incubated in 100 µL of reagent B (permeabilization) and 20 µL of
the primary mAb for 60 min. This prolonged permeabilization/incubation
step was chosen after comparing a 15-min incubation as recommended by
the manufacturer to 30 min and 60 min. In our hands, a 60-min
permeabilization/incubation step allowed optimal immunolabeling without
altered morphology or cell loss.
MWH treatment.
Cells (10) were washed in PBS and
fixed on ice for 30 min in 20 mL/L PFA in PBS. The supernatant was
discarded after centrifugation (500g for 5 min at room
temperature), and cells were resuspended in 10 mL of a 0.01 mol/L
sodium citrate buffer, pH 6.0, containing 5 g/L bovine serum albumin in
a 50-mL propylene tube. The cell suspension was heated in a microwave
oven at the maximum power setting (800 W), reaching a temperature
between 90 and 100 °C. Heating times of 30 and 60 s were
compared for each marker. Cells were then chilled on ice for 10 min.
After centrifugation (500g for 5 min at room temperature),
the supernatant was discarded, and the cells were washed in PBS. The
requirement for ice during fixation and after heating was assessed with
the p53 monoclonal and corresponding control CLs. Results obtained with
and without ice after PFA fixation and/or heating were compared.
immunolabeling
For each of the above techniques, incubation was performed with
the mouse primary mAb (100 µL) at the appropriate dilution for 60
min. This step was followed by incubation with a fluorescein
isothiocyanate-conjugated goat anti-mouse IgG antibody (100 µL) for
30 min in the dark. Cells were resuspended in 200 µL of ISOTON II for
analysis. Cells were washed in PBS before and after each incubation
step. Incubation with the various antibodies was performed at room
temperature.
complementary analysis
Separate set of preliminary experiments.
Recommended a few
years ago for the detection of intracytoplasmic antigens
(1)(29)(30), saponin was shown to
detect the intranuclear antigen PCNA poorly (2).
Accordingly, saponin (20 mL/L PFA for 30 min plus 1 g/L saponin for 5
min) was compared with OPF for the detection of both the intranuclear
antigen p53 and the intracytoplasmic antigen bcl-2. For p53, RAJI and
CEM were used as positive CLs, and IM9 served as negative control. For
bcl-2, RAJI and DG75 LMP were used as positive CLs, and DAUDI served as
negative control.
Additional controls.
Peripheral blood samples collected from
healthy blood donors were used as negative controls for all the
intracellular antigens.
To assess the reproducibility of the study by Lan et al. (2), we repeated the published experiments using the PCNA mAb and the U937 monocytic CL.
Antigenic preservation of external epitopes after MWH was assessed on lymphocytes collected from patients presenting with a B-cell chronic lymphocytic leukemia as well as on the gp-170 VLB100 CL, using the CD19 and 4E3 antibodies, respectively.
The influence of delayed immunostaining on cellular morphology, immunoreactivity, and cell scatter criteria was evaluated for the three methods, using RAJI cells and p53 mAb. Cells fixed, permeabilized, and immunostained immediately were compared with cells kept at 4 °C and immunostained 3 days and 1 week later, respectively.
flow cytometry
All samples were run on a flow cytometer (FACScan, Becton
Dickinson) equipped with an argon ion laser (488 nm). Forward and right
angle light scatter signals were collected for all samples, along with
fluorescein isothiocyanate fluorescence measured through a 543 nm
bandpass filter. The percentage of positive cells was measured from a
cutoff set using an isotype-matched nonspecific control antibody. The
mean channel fluorescence (MCF) was measured for a total of 5000 cells.
statistical analysis
Graphical representation.
Bland and Altman
(31)(32) and, more recently, Petersen et al.
(33) emphasized the importance of graphical interpretation
of analytical data when comparing a reference and a field method. They
both advocated the use of difference plots. However, current data do
not allow a strict application of these concepts. Data recorded with
distinct intracytoplasmic and intranuclear antigens on distinct CLs
cannot be pooled together. Moreover, an optimal cytometric method must
be defined as the method giving the lowest MCF value with CLs defined
as negative controls while giving the highest MCF value for CLs
expressing the antigen of interest. The statistical agreement of the
MWH technique was assessed using SD and a graphical approach adapted to
the current situation. We considered the MCF values recorded for the
negative-control CLs, the cytoplasmic antigens, and the intranuclear
antigens separately.
Choice of the reference value.
OPF and F&P have both been
proposed as standard methods (3). There is no consensus yet
on which method is the best; hence we have no way of considering a
unique reference value. According to Bland and Altman (31),
the mean of the two measurements is the best estimate we have when we
do not know the true value. Considering the MCF differences existing
between OPF and F&P, the MCF average of these two commercial techniques
was therefore chosen as the reference value. To determine if the MWH
method showed MCF values comparable with this chosen reference value,
we took into account the Bland and Altman (31) statement
that how far apart measurements can be without causing difficulties is
a question of judgment. We considered a MCF difference (threshold)
15% as good agreement. The arbitrary choice of 15% was considered
to be conservative because a 15% difference is smaller than the
average difference between OPF and F&P (which was 18%, with one-half
of the values exceeding the 15% difference). The related questions
focused on the agreement of the methods and on reproducibility.
Statistical analysis was performed on the triplicate results of one representative trial. Slight variations of intranuclear antigenic expression of rb and p53 were indeed observed over time with positive CLs, regardless of the method used, despite the log phase pattern of growth and despite comparable method reproducibility evidenced by SD. Such variations were therefore assumed to be cell cycle-related. Western blot analysis on rb-positive CLs confirmed this hypothesis and excluded a cytometric artifact (17). Because these variations were observed with any of the three methods performed at the same time on the same positive CLs, the trial giving the highest MCF value was chosen as representative.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Reproducibility of PCNA detection.
Our findings confirmed
those of Lan et al. (2) using MWH on the U937 CL and
peripheral blood mononuclear cells with the marker PCNA. PCNA was
detected in the majority of U937 cells (70% positive cells; MCF =
238), whereas peripheral blood mononuclear cells were negative (1.2%
positive cells; MCF = 75). These results allowed us to proceed
further and compare the three fixatives with other antibodies.
Results of the quantification for p53, bcl-2, rb, and gp-170.
Results for the four antigens in the three methods are given in Table 2
. Each commercial method was effective in distinguishing
negative and positive CLs for expression of intracytoplasmic and
intranuclear antigens. Variability of antigen expression was noticed,
however, among the three different methods according to the cell type
and the antigen assessed. Altogether, variability between the
respective methods in this study, calculated as the difference between
the highest and lowest MCF values, was between 9% and 36%. The MWH
method compared favorably with OPF and F&P, especially with
intranuclear antigens, where it ranked consistently first or second in
the MCF intensity. In 5 of 12 antigen detection tests, MWH gave the
highest MCF value. In the remaining experiments, the difference between
the MWH method and the best MCF value was between 6% and 16% (the
only exception being the 30% difference for rb detection in KG1). The
MCF with MWH was also associated with a consistent lower background
observed with isotypic antibodies or when no primary antibody was
added. When compared with OPF, the background for the CLs IM9, RAJI,
CEM, and DG75 was decreased between 17% and 64% with MWH (Table 3
). Variation of antigen expression was also observed between
positive CLs for one particular antibody, especially with the
intranuclear antigens p53 and rb. Nonetheless, the same profile of
expression, i.e., high, low, or intermediate constitutive expression,
was consistently found in the positive CLs, irrespective of the
method used. For each marker, Western blot and APAAP results confirmed
the expected positivity and negativity of the control CLs. Although not
designed for precise quantification of protein expression, Western
blots were in total agreement with the antigenic profile (negative vs
either low, intermediate, or high MCF value) of the various CLs (Fig. 1
).
|
|
|
Technical factors potentially influencing MWH results.
Extending the heating time from 30 s to 60 s did not improve
the results with the MWH method. However, the loss of liquid and
cells observed after heating for 60 s did not occur after 30
s. Performing the MWH method with or without ice during fixation and/or
after heating yielded comparable results with p53 on positive- and
negative-control CLs (data not shown).
Preservation of external epitopes by MWH.
It is worth noting
that no CD19 and 4E3 labeling was observed with the MWH method when
testing B-cell chronic lymphocytic cells and gp-170 subclones,
respectively.
effect on cell morphology and immunoreactivity
The effect on cell morphology was measured by light scatter
signals. The three methods preserved good cellular morphology and
allowed easy discrimination between intact cells and debris, as
illustrated with the p53 mAb and DG75 in Fig. 2
.
|
Cell immunoreactivity, cell scatter properties, and excellent cellular
morphology were obtained with the OPF and F&P techniques for a time
interval as long as 1 week after fixation. However, immunolabeling had
to be performed within 2 to 3 days after fixation with the MWH method.
A 1-week delay caused a decrease in antigen immunodetection and an
alteration of the cellular morphology (Fig. 3
).
|
statistical analysis
Negative-control CLs.
For five of eight MCF measurements, the
difference between MWH and the MCF mean calculated on OPF and F&P
results was <15% (Fig. 4
A). However, regarding rb antigenic expression on 6M1 and IM9,
we noted a higher difference: -38% and +35%, respectively. In both
cases, the difference between OPF and F&P was also >15% (23% and
21%, respectively). Nonetheless, the graphical representation
indicates that the SD of the three methods were comparable.
|
Nuclear antigens rb and p53.
For four of seven measurements,
the results were highly similar to the mean OPF and F&P (-10%, -2%,
+5%, and +11%; Fig. 4B
). In the other three measurements, analysis of
p53 expression in LMP, DG75, and CEM cells showed a noticeably greater
MCF with MWH (29%, 30%, and 22%, respectively). Altogether, MCF
values tended to be higher (but not significantly) for MWH. SD were
comparable for the three methods.
Cytoplasmic antigens bcl-2 and gp-170.
For four of five
measurements, results were comparable (variations between 0% and 11%;
Fig. 4C
). There was a slight trend towards lower MCF values with MWH.
However, the MCF value of the CL VLB100 with MWH was much greater than
the MCF mean of the other two methods (47%). Reproducibility of the
three methods was comparable, as shown by the SD.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In a diagnostic laboratory, the choice of a particular technique is increasingly influenced by considerations such as versatility, rapidity, involvement of potentially harmful reagents, cost, and application to a wide range of antigens. In this context, OPF, F&P, and MWH are easy to perform, require approximately the same amount of time to complete (~3 h 30 min) and have been shown to be suitable for a wide range of antibodies. None of them are particularly harmful. Because their computer-aided design improved optimization and standardization of the labeling procedure, OPF and F&P have recently been proposed as standards for the detection of intracellular antigens (3). The cost of OPF and F&P, however, may be prohibitive for routine diagnostic use. As an alternative, MWH requires only PFA and a citrate buffer, making it a much cheaper method. Chilling cells on ice during fixation and after heating, which is a somewhat cumbersome step described in the initial protocol (2), does not appear necessary. A microwave oven, required for MWH, is a very common and inexpensive piece of equipment.
Regarding the economic aspects, we also ran experiments with PFA and saponin, an inexpensive method recommended for detecting intracytoplasmic antigens when commercial fixatives were not yet available (1)(29)(30). However, MCF values recorded with this method were, in our hands, substantially lower (>30% decrease) than with OPF, an observation that is in accordance with the comparison between Tween 20 and OPF (6) and with the poor detection of PCNA with saponin (2). Such lower performance associated with a lack of consensus on optimal PFA and saponin concentrations, a lack of consensus on the optimal time for fixation and permeabilization, and finally, the reported lack of detection of some intracytoplasmic antigens with saponin (37)(38) prompted us not to pursue the comparison between saponin and other fixatives further.
Regarding versatility, cell morphology preservation and antigenic retention were carefully examined. Morphology is a critical point directly affecting the quality of the results. This is a necessary feature to ensure accurate flow cytometric measurements. Whereas permeabilization methods using detergents have been reported to cause cellular damage (1)(2)(3), morphological characteristics of the cells appeared to be equally preserved with OPF, F&P, or MWH. These methods fixed and permeabilized the cell membrane without major damage to the cell. The light scatter patterns provided clear discrimination between entire cells and cellular debris. Antigenic retention clearly distinguished OPF and F&P from MWH. As already experienced by Lan et al. (2) with MHC class I antigen, we confirmed that the other membrane antigens do not withstand microwave treatment, making MWH unsuitable for such detection. This may, in part, explain the consistent lower background obtained with this method when assessing isotypic antibodies. Each of the three methods appeared effective for the detection of the four antigens on the respective CLs. Flow cytometry results confirmed the preliminary assessment by immunocytochemistry and Western blot analyses of the expected phenotypes of the cells. For each antibody, the method comparison showed a good agreement between MCF values. Likewise, analysis of SD indicated that MWH reproducibility is comparable to OPF and F&P. The SD tended to be smaller for the MWH method than for OPF and F&P (SD = 7.2, 8.5, and 9.5, respectively), but this was not significant. Slight variations of MCF values recorded for each antibody and CL according to the method used might reflect a difference in accessibility or preservation of the antigens. Variations between OPF and F&P methods have also been reported with other antigens (5)(7), with both studies concluding that F&P was more specific and sensitive. Current data show that MWH is as effective as OPF and F&P for detecting intracellular proteins. In several instances, MWH was one of the best methods for the detection of nuclear antigens. MWH therefore appears particularly appropriate for studying variations of nuclear protein content related to cell cycle regulation, a problem that has an increasing clinical impact. The mechanism by which MWH permits intracellular antigen detection while preserving good cellular morphology is not well understood. It has been shown that heat denaturation of native protein secondary structure can be prevented by the cross-linking fixative formaldehyde (2). This could also be related to the combined effects of PFA and microwaves. It seems that successful detection with MWH depends on the nature and localization of antigens.
As reported for OPF (4), fixation and permeabilization may permit a delay before the immunostaining and cytometric analysis. The length of this delay would be of particular interest to small laboratories lacking facilities for flow cytometry. This would allow them to accumulate and send patients samples to reference centers. In the current study, it was found that immunostaining and cytometric analysis could be delayed for up to 1 week following fixation with OPF or F&P and at least 3 days for MWH without adverse effects. Although this delay is shorter for MWH, it would allow routine analysis in common daily practice. It means also that samples that arrived on Friday afternoon could be analyzed on the next Monday.
In conclusion, our data indicate that OPF, F&P, and MWH methods are suitable for the analysis of intracellular antigen expression by flow cytometry. Although slight differences were observed according to cell type and the antigen assessed, each method provides effective fixation and membrane permeabilization, allowing immunologic interaction while preserving cellular morphology. They offer good reproducibility and can be used in routine flow cytometric protocols. In several instances, especially with intranuclear antigens, MWH compared favorably with the commercial methods while being considerably cheaper. This can be of importance in clinical laboratories as well as research settings. Loss of external epitopes appears to be a genuine limitation of the MWH assay, but this seems interestingly associated with lower background staining. Besides nuclear protein analysis, MWH could be used reliably in a timesaving manner to establish appropriate conditions for quantification of intracellular proteins before performing other time-consuming and/or expensive assays (39). According to the present results, the relevance of the MWH method in the clinical setting deserves additional study.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), F&P
(
), and OPF (
).