SEARCH

SEARCH BY CITATION

Keywords:

  • leukocytes;
  • flow cytometry;
  • paraformaldehyde;
  • immunofluorescence;
  • fixation

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

Background:

Immunophenotyping of blood leukocytes often involves fixation with paraformaldehyde prior to cytometry analysis. However, the influence of cell type and marker specificity on the stability of fluorescence intensity after fixation has not been well studied.

Methods:

Human whole blood was stained using a panel of fluorescein isothiocyanate-labeled antibodies to surface markers. Unfixed and fixed samples were analyzed by flow cytometry at 0, 2, 4, 6, 24, 48, and 96 h after staining. Fluorescence measurements were converted to molecules of equivalent soluble fluorochrome for comparison.

Results:

Fixation caused a significant decrease in both forward and side scatter at 48 h which required gating adjustments to achieve resolution of cell populations. The autofluorescence increased progressively in fixed samples (ninefold at 96 h for monocytes). Variable decreases in marker-associated fluorescence became apparent after correction for autofluorescence. The magnitude of the decrease at 96 h varied with cell type and marker, from 5% for CD32 on monocytes to 39% for CD16 on neutrophils.

Conclusion:

The change in fluorescence intensity following staining and fixation of leukocytes varies with cell type and surface marker. Fluorescence stability should be determined for each cell type and marker used, and the confounding effects of fixation on cell autofluorescence should be considered. © 2007 International Society for Analytical Cytology.

Immunostaining of leukocyte surface antigens is an important research tool in immunology. Typical laboratory protocols often involve fixing with 1% paraformaldehyde (PFA) after staining, followed by storage of samples for various periods of time prior to analysis on the cytometer. Such protocols require knowledge of how fluorescence intensity varies over time. This is especially true with experiments involving low marker expression or small changes in fluorescence intensity.

Before determining changes in fluorescence intensity of stained and fixed samples over time, there are two potential effects of fixation on fluorescence intensity that need to be considered. First, does fixation and storage of samples alter the light scatter properties of the cells sufficient to affect resolution of cell populations, and second, does fixation alter cell autofluorescence?

There have been conflicting reports on the effects of poststaining fixation on light scatter properties of cells. Several authors (1–4) report that light scatter properties of lymphocytes remained stable for up to 1 week after fixation in 1–3% PFA. Lal et al. (5), on the other hand, reported that fixation significantly altered the lymphocyte scatter pattern. However, Lal et al. (5) and Lewis (6) report that scatter remains unchanged if the fixative is washed out and samples are stored in buffer. It is widely appreciated that PFA fixation increases cell autofluorescence (1, 2, 6–9). However, we are unaware of any quantitative attempts to determine the effects of fixation-induced changes in autofluorescence on measurement of human leukocyte surface markers.

Most previous studies of the effects of poststaining fixation on FITC fluorescence intensity have focused on lymphocytes, with varied results. Lifson et al. (3) reported that expression of CD3, 4, and 8 was stable for 1 week after fixation. Lanier and Warner (4) reported a slight drop in the fluorescence intensity of a variety of markers after 1 week but relative expression (% positive) remained unchanged. Denny et al. (10) examined antibody binding capacity via fluorescence intensity at 0, 24, and 48 h after fixation with varying results; CD3 and 19 increased at 24 h, CD4, 8, and 16 did not, and CD3, 4, and 19 all increased significantly at 48 h. Lal et al. (5), using a fix-wash technique, also reported variable stability depending on the marker; for example, CD4, 8, and 19 were stable, and CD3 intensity decreased.

There have been limited studies of the effects of fixation and storage on marker expression using cells other than lymphocytes. Lanier and Warner (4) reported stability for human peripheral blood cells but showed no data. Denny et al. (10) reported that the fluorescence intensity of CD4 on monocytes increased at 48 h. Hu et al. (11) reported increases in P-selectin expression on platelets as early as 1-h post fixation. Macey and McCarthy (12) and McCarthy et al. (13) using prestaining fixation with very short PFA exposures, reported significant decreases in fluorescence intensity on all classes of leukocytes for CD18 and 62L epitopes, and lesser decreases for CD11b. None of these studies reported correction for changes in autofluorescence, although the Macey and coworkers papers (12, 13) indicate increases in isotype intensities which may reflect increased autofluorescence.

Because published data on fluorescence stability after fixation are limited and often conflicting, we sought to determine changes in fluorescence intensity up to 96 h after fixation, using a panel of antibodies to surface markers on blood leukocytes, with consideration of fixation effects on cell autofluorescence. We hypothesized that stability after fixation would vary with cell type and marker antibody.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

Subjects

This study was approved by the Research Subjects Review Board at the University of Rochester Medical Center, and informed consent was obtained. Healthy, never-smoking subjects (aged 35 ± 13 years) were recruited from the local population. Blood was drawn from an antecubital vein into sterile 10-ml vacutainer tubes containing sodium heparin (Becton Dickinson, Franklin Lakes, NJ) and immediately processed.

Immunostaining

Whole blood was aliquoted into titered antibody and kept in the dark for 15 min at room temperature. Samples were mixed once, mid-way through the incubation. The antibodies tested (Table 1) were all fluorescein isothiocyanate (FITC) conjugates with the exception of anti-CD45, which was a peridinin chlorophyll (PerCP) conjugate. The anti-CD45 was added to all samples for gating purposes only. Autofluorescence was measured in the FL1/FITC channel (530 nm) using samples stained only with anti-CD45 PerCP.

Table 1. Surface Marker Antibodies Used
AntibodyCloneManufacturer
CD11a25.3.1Beckman Coulter, Fullerton, CA
CD11bICRF44Ancell Laboratories, Bayport, MN
CD163G8Beckman Coulter, Fullerton, CA
CD18L130BD Biosciences, San Jose, CA
CD32L130Serotec, Raleigh, NC
CD452D1BD Biosciences, San Jose, CA
CD5484H10Beckman Coulter, Fullerton, CA
CD62LDREG56BD Biosciences, San Jose, CA
CD6410.1BD Biosciences, San Jose, CA
IgG1X40BD Biosciences, San Jose, CA

After staining, 2 ml of a 1:10 dilution of lysing solution (BD FACS™, BD Biosciences, San Jose, CA) was added to each tube. Samples were mixed and allowed to lyse for 10 min in the dark, then centrifuged at room temperature for 5 min at 300g with no brake. The supernatant was aspirated and the pellet resuspended in 2 ml of Dulbecco's phosphate buffered saline (PBS), without calcium or magnesium, containing 0.1% sodium azide. Samples were mixed and centrifuged for 5 min at 200g with no brake. The supernatant was aspirated. If samples were to be read unfixed, 0.5 ml of PBS/azide was added. Fixative (0.5 ml of 1% PFA) was added to the remainder of the samples. Fixative was prepared by diluting 20% electron microscopy-grade PFA (Electron Microscopy Sciences, Hatfield, PA) with PBS/azide. All buffers and fixatives were sterile filtered, refrigerated, and used within 1 week of preparation.

Additional experiments were performed without azide, and with washing out PFA after fixation, to determine the role of these materials on fluorescence stability.

Samples were read on a FACScan™ flow cytometer (BD Biosciences, San Jose, CA) using settings optimized for blood leukocytes, at 0, 2, 4, 6, 24, 48, and 96 h after fixation. In addition, three sets of samples were also run with no fixation and read immediately after processing (0 h). Quantum high and low level FITC beads (Bangs Laboratories, Fishers, IN) were run on the cytometer along with each set of samples. The pH of the samples was measured at each time point using a Model Φ350 pH meter with a Futura Gel Long Electrode (Beckman Coulter, Fullerton, CA). The pH remained unchanged over the time course and thus was not a factor.

Analysis of Cytometry Data

Data were analyzed using CellQuest™ software (BD Biosciences, San Jose, CA). Cell populations were gated using their positions on a CD45 vs. SSC plot. The geometric mean intensity was converted to molecules of equivalent soluble fluorochrome (MESF) using a calibration curve generated with the quantum beads for each run as detailed previously (14). Correction for changes in autofluorescence was done using MESF. Data were compared using Student's t tests, with significance characterized by P < 0.05.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

To assess the fluorescence stability of the various markers, it was necessary to determine the effects of fixation on cell light scatter properties and gating, and to determine the effects of fixation-induced changes in autofluorescence.

In general, we found that FSC decreased after fixation, becoming significant at 48 h and remaining at that level through 96 h. Similar losses were seen in all three cell types [lymphocytes (11.4 ± 4)%, monocytes (13.1 ± 3)%, neutrophils (12.6 ± 3)%, n = 9]. SSC also decreased but the extent varied more with cell type [lymphocytes (10.7 ± 3)%, monocytes (16.5 ± 3)%, and neutrophils (11.6 ± 2)%]. These changes did not affect resolution of the cell populations using CD45 vs. SSC, although changes in SSC required adjustments in gate placement.

Next, we determined the contribution of autofluorescence. Unstained samples read immediately after fixation showed increased fluorescence intensity compared with unfixed samples (Table 2). When the fluorescence intensity of the stained samples was corrected for changes in autofluorescence, there was no significant difference between fixed and unfixed samples read immediately after processing. The autofluorescence of the fixed samples continued to increase with time in all three cell types (Fig. 1). After 96 h, the increase from the readings at 0 h was greatest in monocytes [(910 ± 114)%] followed by lymphocytes [(735 ± 142)%] and neutrophils [(467 ± 62)%].

thumbnail image

Figure 1. Changes in leukocyte autofluorescence after fixation. (AC) Changes in fluorescence intensity for three cell types from a typical subject. Filled area denotes cells immediately after fixation; dark solid line, 48 h after fixation; light solid line, 96 h after fixation. (DF) Mean (±SD, n = 17) fluorescence intensity expressed as MESF for each of the three cell types.

Download figure to PowerPoint

Table 2. Autofluorescence of Unfixed and Fixed Samplesa
Cell typeUnfixed samples (MESF)bFixed samples (MESF)Difference (%)P value
  • a

    Both samples were read immediately after processing.

  • b

    Values are geometric mean (n = 3) converted to molecules of equivalent soluble fluorochrome.

Lymphocyte771 ± 61892 ± 13218 ± 30.001
Monocyte1,151 ± 851,352 ± 11916 ± 70.001
Neutrophil1,377 ± 1611,483 ± 1816 ± 30.027

It was therefore necessary to correct for autofluorescence at all time points before judging the stability of the fluorescent markers. Once corrected, however, the fluorescence intensities for the markers showing positive expression showed varied patterns of stability (Fig. 2). In some cases changes were small and intensity remained relatively stable for 24 h, in some there was a steady decline over time, and in others the decline reversed after 48–96 h. Table 3 provides the individual data, with significant changes indicated.

thumbnail image

Figure 2. Cell markers showing varying patterns of change in fluorescence intensity after fixation (mean ± SD, n = 17). (A) CD18 expression on neutrophils, (B) CD11a expression on lymphocytes, (C) CD16 expression on neutrophils, and (D) CD62L expression on lymphocytes. All data have been corrected for autofluorescence.

Download figure to PowerPoint

Table 3. Change in Fluorescence of Surface Markers with Storage After Fixationa
Hours after fixation0246244896
  • a

    Values are geometric mean ± SD converted to molecules of equivalent soluble fluorochrome × 104, after correction for autofluorescence.

  • *

    P < 0.05,

  • **

    P < 0.01 compared to 0 h.

Lymphocytes
CD11a6.4 ± 1.06.0 ± 0.86**5.6 ± 0.75**5.8 ± 0.66**5.4 ± 0.83**5.4 ± 0.75**5.0 ± 0.72**
CD184.6 ± 0.804.2 ± 0.78*4.0 ± 0.74**3.9 ± 0.62**3.8 ± 0.49**3.8 ± 0.58**3.6 ± 0.72**
CD62L2.0 ± 0.721.9 ± 0.761.9 ± 0.761.9 ± 0.881.8 ± 0.862.3 ± 0.903.4 ± 0.87**
Monocytes
CD11a8.3 ± 1.67.9 ± 1.57.6 ± 1.4**7.5 ± 1.57.3 ± 1.5**7.0 ± 2.2*7.1 ± 2.4**
CD11b2.9 ± 1.22.7 ± 1.12.5 ± 1.1*2.2 ± 0.92**2.3 ± 0.81**2.5 ± 1.1**2.6 ± 0.97
CD188.3 ± 2.28.2 ± 2.27.6 ± 2.1*7.2 ± 2.1**7.3 ± 1.8**7.4 ± 2.7*7.7 ± 3.0*
CD3211 ± 1.511 ± 1.810 ± 1.69.9 ± 1.210 ± 1.210 ± 2.19.6 ± 1.9*
CD540.89 ± 0.460.85 ± 0.49*0.82 ± 0.51**0.80 ± 0.58**0.71 ± 0.50**0.84 ± 0.600.88 ± 0.88
CD62L5.5 ± 1.55.8 ± 1.55.3 ± 1.5*5.4 ± 1.45.2 ± 1.4*5.2 ± 1.25.7 ± 0.69
CD644.4 ± 2.64.5 ± 2.64.0 ± 2.64.3 ± 2.84.0 ± 2.34.5 ± 3.34.6 ± 3.2
Neutrophils
CD11a3.7 ± 0.393.5 ± 0.26**3.3 ± 0.35**3.3 ± 0.33**3.2 ± 0.26**3.0 ± 0.39**2.9 ± 0.39**
CD11b3.4 ± 1.33.3 ± 1.53.2 ± 1.43.1 ± 1.33.0 ± 1.1*3.2 ± 1.23.3 ± 1.4
CD1658 ± 1956 ± 2253 ± 20*45 ± 13**44 ± 18**35 ± 12**39 ± 25**
CD184.5 ± 0.944.4 ± 0.994.1 ± 1.0*3.9 ± 0.89*4.1 ± 0.77*4.2 ± 1.14.1 ± 1.1
CD3213 ± 1.313 ± 1.512 ± 1.312 ± 1.513 ± 1.312 ± 0.5312 ± 1.4
CD62L10 ± 2.29.7 ± 2.49.3 ± 2.3*9.3 ± 2.3*9.0 ± 2.5**9.0 ± 2.99.2 ± 2.5

Two possible causes for this loss of fluorescence intensity were explored: the presence of azide in the fixative, and the continued presence of fixative during storage. Samples fixed with 1% PFA in PBS without azide showed no significant differences from the original protocol (data not shown). The removal of fixative after 30 min and resuspension in PBS/azide blunted the increase in autofluorescence (Fig. 3), but did not affect the changes seen after correction for autofluorescence. Typical results for CD11a are illustrated in Figure 4. Removal of the fixative resulted in decreased light scatter over time with neutrophils being most affected (Fig. 5). The downward shift in side scatter resulted in the neutrophil region overlapping both monocyte and lymphocyte regions, making gating difficult. Similar results were obtained when 1% bovine serum albumin was added to the resuspension buffer (data not shown).

thumbnail image

Figure 3. Comparison of changes in autofluorescence in three cell types [(A) lymphocytes, (B) monocytes, and (C) neutrophils] seen after removal of fixative compared with standard samples left in fixative (mean ± SD, n = 3). Black circles denote samples resuspended in PBS/azide, and open squares denote samples left in fixative.

Download figure to PowerPoint

thumbnail image

Figure 4. Typical decrease in marker fluorescence of anti-CD11a (mean ± SD, n = 3). (A) Samples stored in fixative, (B) fixative removed and samples stored in PBS/azide. All data have been corrected for autofluorescence.

Download figure to PowerPoint

thumbnail image

Figure 5. Typical changes in the scatter pattern of the neutrophil region (R5) with fixative washed out immediately after fixation. (A) Immediately after washout of fixative, (B) 24 h after washout of fixative, and (C) 96 h after washout of fixative. The examples shown are from a single subject's blood, stained for anti-CD11a, fixed, washed, and resuspended in PBS/azide.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

The fluorescence stability of several FITC-labeled cell surface markers routinely used for phenotypic analysis of human peripheral blood was investigated, along with possible contributions from changes in light scatter and autofluorescence. Both fixed and unfixed cells were studied and changes over time were quantified using standardized fluorescence beads.

Samples left in PFA showed decreases in both FSC and SSC which necessitated gate adjustments but did not hinder resolution. However, in samples washed of PFA and stored in buffer, the large decrease in neutrophil side scatter made gating impossible even at 24 h.

Our study confirms that fixation increases autofluorescence, and, to the authors' knowledge, provides the first quantitative data on this effect. The cause of this increase remains unclear. Eliminating azide from the fixative buffer had no effect. Cell shrinkage can occur with prolonged aldehyde fixation (8). The progressive decrease in FSC, a measure of cell size, from 0 to 48 h after fixation would lend support to cell shrinkage as a contributor. However, FSC was unchanged from 48 to 96 h after fixation, while autofluorescence significantly increased during this interval, suggesting that change in cell size is not the whole answer. The reaction of PFA with a variety of free amine groups (15) may produce fluorescent products. It is possible that the continued presence of PFA furthers this process over time. Lewis (6) reported that washing out the fixative and storing the sample in buffer eliminates the autofluorescence increase. In our hands, washing out the fixative reduced but did not eliminate the increase in autofluorescence.

After correcting for autofluorescence, the changes in fluorescence intensity following fixation were variable, and depended on the cell type and marker. For example, CD62L-related fluorescence on lymphocytes remained stable through 48 h, and then increased at 96 h. CD62L fluorescence on neutrophils decreased significantly at 4 h and continued to decrease thereafter. A number of markers showed decreases through 24 h and then small increases at 48 or 96 h. This pattern of change is similar to that reported by Lal et al. (5) for lymphocytes, and by Denny et al. (10) for CD4 on monocytes.

There are a variety of possible causes for such a loss of marker expression: leakage from lysosomal granules in disrupted neutrophils, cleavage of the fluorochrome and/or antibody complex, internalization of the complex, or quenching of the fluorochrome by fixation products.

Leakage from disrupted neutrophil lysosomes, if it occurred, would likely overwhelm buffer capacity, resulting in acidification. However, the sample pH in this study remained constant, making such disruption unlikely. Furthermore, in their work with lymphocytes, Lal et al. (5) showed that removing neutrophils had no effect on the reported changes.

Cleavage of the antibody or antigen–antibody complex should be prevented by fixation. It is possible that some endogenous proteases might still be active, but Lal et al. (5) reported that inhibiting proteases had no effect. However, since aldehyde fixatives form methylene bridge cross-links between proteins which deform the tertiary structure (13), the continued action of PFA might cause sufficient steric strain to disrupt the fluorochrome-antibody–antigen complex. Several authors (11–13, 16) have shown that PFA decreases antigenicity of some markers when used before staining, either by reacting with the epitopes themselves or with moieties which maintain the structure of the epitope. This might explain the progressive nature of the decreases. Oddly enough, cleavage might also explain the later increases in fluorescence intensity which appeared, for example, with CD54 expression on monocytes. Lal et al. (5) theorized that shed fluorochrome might adsorb nonspecifically to cells creating spurious expression.

Internalization of the antigen–antibody complex might occur because of changes in membrane integrity. Prolonged fixation has been shown to increase membrane permeability (12, 13, 16). Changes in the membrane integrity might also contribute to the decrease in SSC reported here.

Another cause of decrease in fluorescence intensity could be energy transfer. West et al. (17) reported that aldehyde fixation caused quenching of membrane bound probes. Deka et al. (18) showed that self-quenching of FITC can occur if fluorochromes are in close proximity. The apparent decrease in cell size as indicated by the FSC changes, as well as steric transformations bringing susceptible groups into closer proximity, could increase quenching. Interestingly, in our study, some of the largest decreases were seen in neutrophil expressions of CD16—a complex with numerous sites on the membrane of the smallest of the three cell types.

In summary, PFA fixation of human blood altered light scattering properties of leukocytes, necessitating some gate adjustments, but did not affect resolution of the cell populations. Fixation significantly increased the autofluorescence of leukocytes, as much as 910% in monocytes at 96 h. Once corrected for autofluorescence, immunofluorescence intensity generally decreased with time, to varying degrees depending on cell type and surface marker. These variable changes in fluorescence intensity following fixation should be considered in immunofluorescence experiments.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED
  • 1
    Stewart CC,Stewart SJ. Cell preparation for the identification of leukocytes. Methods Cell Biol 2001; 63: 217251.
  • 2
    Givan AL. Flow Cytometry First Principles, 2nd ed. New York: Wiley-Liss; 2004. 296 p.
  • 3
    Lifson JD,Sasaki DT,Engleman EG. Utility of formaldehyde fixation for flow cytometry and inactivation of the AIDS associated retrovirus. J Immunol Methods 1986; 86: 143149.
  • 4
    Lanier LL,Warner NL. Paraformaldehyde fixation of hematopoietic cells for quantitative flow cytometry (FACS) analysis. J Immunol Methods 1981; 47: 2530.
  • 5
    Lal RB,Edison LJ,Chused TM. Fixation and long-term storage of human lymphocytes for surface marker analysis by flow cytometry. Cytometry 1988; 9: 213219.
  • 6
    Lewis DE. Cytochemistry I: Cell surface immunofluorescence. In: BauerKD, DuqueRE, ShankeyTV, editors. Clinical Flow Cytometry, Principles and Application. Baltimore: Williams and Wilkins; 1993. Chapter 9, pp 143156.
  • 7
    Crissman HA,Steinkamp JA. In: MelamedMR, LindmoT, MendelsohnML, editors. Flow Cytometry and Sorting, 2nd ed. New York: Wiley-Liss; 1990. Chapter 13, pp 227247.
  • 8
    Hopwood D. Fixation and fixatives. In: BancroftJD,GambleM, editors. Theory and Practice of Histological Techniques, 5th ed. New York: Churchill Livingstone; 2002. Chapter 5, pp 6384.
  • 9
    Kerr RSC,Kaye AH,Bartlett PF. A quantitative assessment by flow cytometry of the preservation of the Thy-1 antigen by various fixation procedures. Neurosci Lett 1988; 87: 259265.
  • 10
    Denny TN,Stein D,Mui T,Scolpino A,Holland B. Quantitative determination of surface antibody binding capacities of immune subsets present in peripheral blood of healthy adult donors. Cytometry 1996; 26: 265274.
  • 11
    Hu H,Daleskog M,Li N. Influence of fixatives on flow cytometric measurements of platelet P-selectin expression and fibrinogen binding. Thromb Res 2000; 100: 161166.
  • 12
    Macey MG,McCarthy DA. Quantitation of adhesion molecules and other function-associated antigens on human peripheral blood leukocytes. Cytometry 1993; 14: 898908.
  • 13
    McCarthy DA,Macey MG,Cahill MR,Newland AC. Effect of fixation on quantification of the expression of leukocyte function-associated surface antigens. Cytometry 1994; 17: 3949.
  • 14
    Gavras JB,Frampton MW,Ryan DH,Levy PC,Looney RJ,Cox C,Morrow PE,Utell MJ. Expression of membrane antigens on human alveolar macrophages after exposure to nitrogen dioxide. Inhal Toxicol 1994; 6: 633646.
  • 15
    Van Ewijk W,Van Soest PL,Verkerk A,Jongkind JF. Loss of antibody binding to prefixed cells: Fixation parameters for immunocytochemistry. Histochem J 1984; 16: 179193.
  • 16
    Pollice AA,McCoy JP,Shackney SE,Smith CA,Agarwal J,Burholt DR,Janocko LE,Homicek FJ,Singh SG,Hartsock RJ. Sequential paraformaldehyde and methanol fixation for simultaneous flow cytometric analysis of DNA, cell surface proteins, and intracellular proteins. Cytometry 1992; 13: 432444.
  • 17
    West CA,He C,Su M,Swanson SJ,Mentaer SF. Aldehyde fixation of Thio-reactive fluorescent cytoplasmic probes for tracking cell migration. J Histochem Cytochem 2001; 49: 511517.
  • 18
    Deka C,Lehnert BE,Lehnert NM,Jones GM,Sklar LA,Steinkamp JA. Analysis of fluorescence lifetime and quenching of FITC-conjugated antibodies on cells by phase-sensitive flow cytometry. Cytometry 1996; 25: 271279.