Journal of Immunological Methods, 67 (1984) 37-51 Elsevier
37
JIM 02940
A Flow Cytometric Technique for Quantitation of B-Cell Activation 1 Joseph G. Garner *, Robert
B. C o l v i n **
and Robert T. Schooley
,.2
* Infectious Disease Unit of the Medical Service and ** Immunopathology Unit of the Department of Pathology, Massachusetts General Hospital, and Harvard Medical School, Boston, MA 02114, U.S.A.
(Received 5 May 1983, accepted 19 September 1983)
A flow cytometric method for enumeration of intracytoplasmic antibody-containing cells was developed. Flow cytometry was found to yield results comparable to traditional fluorescence microscopy in cells stimulated to produce intracytoplasmic antibody by either pokeweed mitogen or Epstein-Barr virus. This technique offers several advantages over traditional fluorescence microscopy, including objective measurement of fluorescence intensity and rapid analysis of large numbers of cells. Key words: intracytoplasmic antibody - flow cytometry - pokeweed mitogen - Epstein - Barr virus
Introduction
Polyclonal B-cell activators "have been used extensively in investigations of the control of immunoglobulin synthesis (Waldmann and Broder, 1982). Several parameters of B-cell activation have been monitored, each with characteristic advantages and disadvantages. These include determination of [3H]thymidine incorporation (Greaves et al., 1974), demonstration of intracytoplasmic antibody by immunofluorescence (Hijams et al., 1969), enumeration of immunoglobulin-secreting cells (Eby et al., 1975) and measurement of immunoglobulin concentration in culture supernates (Waldmann et al., 1974). Flow cytometry is an increasingly accessible technology which offers several _advantages over traditional fluorescence microscopy, including rapid enumeration of large numbers of cells, objective determination of fluorescence intensity, and graphic as well as numerical display of data (Horan and Wheeless, 1977; Loken and Stall, 1982). In addition to its widespread use in surface antigen analysis, flow cytometry has been used to detect both intracytoplasmic (Zeile, 1980) and intranuclear mole1 Supported in part by grants from the Leukemia Society and the National Institutes of Health (AI 17057, HL 18646). 2 Address reprint requests to: Dr. Robert T. Schooley, Infectious Disease Unit, Massachusetts General Hospital, Boston, MA 02114, U.S.A. 0022-1759/84/$03.00 © 1984 Elsevier Science Publishers B.V.
3~ cules (Traganos et al., 1977). We have developed a flow cytometric system for quantitating polyclonal B-cell activation by detection of intracytoplasmic antibody.
Materials and Methods
Lymphocyte preparation and pokeweed mitogen activation Human peripheral blood was obtained from normal volunteers and anticoagulated with preservative-free heparin (10 U / m l ) . The mononuclear cell fraction was isolated by centrifugation over a Ficoll-Hypaque gradient. Pokeweed mitogen (PWM)-activated cells were generated by resuspending the peripheral blood mononuclear cells (PBMCs) at a concentration of ] × 106/ml in RPM1 1640 supplemented with penicillin (250 U / m l ) , streptomycin (250/~g/ml), 10 mM Hepes buffer, 10% heat-inactivated type AB serum and PWM (GIBCO, Grand Island, NY) at a final dilution of either 1:50 or 1:100. One milliliter of this cell suspension was placed in replicate 12 mm × 75 mm clear plastic tubes and incubated at 37°C in a humid atmosphere containing 5% CO 2 for 5-6 days. Control PBMCs were treated identically except for the omission of PWM.
Epstein-Barr virus (EB V)-transformed cell lines The EBV-transformed lymphoblastoid cell lines P~HR-1 and B95-8 were maintained in RPMI 1640 supplemented with penicillin, streptomycin, Hepes buffer and 10% fetal calf serum (FCS). Additional cell lines were established by transformation of adult T-cell depleted mononuclear cells with EBV derived from the B95-8 cell line.
Human myeloma cell line The RHL-4 human myeloma cell line was maintained in RPMI 1640 supplemented as described for EBV-transformed cell lines.
Intracytoplasmic antibody staining for fluorescence microscopy Cells were washed in phosphate-buffered saline (PBS), suspended at a concentration of 1 X 106/ml in PBS with 5% FCS and cytocentrifuged onto microscope slides. After air drying, cells were fixed in a 1 : 1 methanol : acetone solution at 4°C for 30 min, air dried and stained with a 1 : 20 dilution of an FITC-conjugated polyclonal goat F(ab') z anti-human immunoglobulin antiserum (TAGO, Burlingame, CA) for 30 rain at 37°C in a humidified atmosphere. Slides were then washed twice in PBS, counterstained with 1% methylene blue and mounted in Bacto FA mounting fluid (Difco, Detroit, MI). At least 300 cells from each specimen were counted under oil using a Zeiss Standard fluorescence microscope.
Intracytoplasmic antibody staining for flow cytometry Cells were washed in PBS, suspended at a concentration of 1 x 106/ml and pelleted. The supernate was removed and the cells resuspended in 1 ml of chilled PBS containing 2% paraformaldehyde with 0.25% saponin (Fisher, Fair Lawn, N J),
39 100 ~ g / m l of L-a-lysophosphatidylcholine (lysolecithin; Sigma, St. Louis, MO) or 0.01% Triton X-100 (Sigma, St. Louis, MO). Following incubation at 4°C for 5 min, cells were washed twice with 5 ml of cold PBS. Next, 0.075 ml of a 1 : 40 dilution of the goat F(ab') 2 anti-human immunoglobulin antiserum was added to the pellet. In some experiments, FITC-conjugated human albumin (Cappel, Cochranville, PA) in a 1:40 dilution was used instead of goat anti-human immunoglobulin antiserum. Cells were agitated briefly and incubated at 4°C for 30 min. They were then washed twice in 5 ml of PBS, resuspended in 1 ml of PBS with 2% FCS and stored at 4°C in the dark until use. PWM-stimulated cells often formed clusters of 2-5 cells during the fixing and staining procedure. These cell clusters were disrupted by repetitive pipetting with a P1000 Gilson Pipettman pipetter (Rainen, Woburn, MA) immediately prior to analysis by flow cytometry. Less than 2% of the cells remained aggregated and almost all of the remaining clusters were dyads. Paraformaldehyde, saponin and lysolecithin were dissolved in PBS and stored at - 2 0 ° C for less than 1 month prior to use. Triton X-100 was prepared on the day of use.
Flow cytometry Flow cytometric studies were performed with an Ortho Spectrum III flow cytometer (Ortho Diagnostic Systems, Westwood, MA), using the 488 mu band of a 50 mW argon laser. Cells for analysis were selected by gating on forward and right angle scattered light as described by Hoffman et al. (1980). Fluorescence was standardized with glutaraldehyde-fixed human erythrocytes; the mean channel for these erythrocytes was less than 5 at the gain used for intracytoplasmic antibody detection. Histograms were stored and processed on an Apple II microcomputer. In general, 1500-3000 cells were counted for each determination; no determination with less than 500 cells was used.
Results
P WM-stimulated preparations Representative scatter plots (cytograms) from unstimulated and PWM-stimulated PBMCs appear in Fig. 1. The unstimulated cells appear on the cytogram as a relatively circumscribed cluster similar to that produced by fresh lymphocyte preparations. Due to their larger size and altered internal structure, PWM-stimulated cells appear not only in the region occupied by unstimulated cells, but also in a l~road band extending to areas of higher forward and right angle light scatter. The trigger region (the area of the cytogram in which the fluorescence of cells is recorded) was set to encompass the region of both the unstimulated cell cluster and the PWM-treated cell clusters. In practice this meant that all of the cytogram except the portion below the unstimulated cell cluster, an area which contains residual platelets and cell fragments in fresh PBMC preparations (Hoffman et al., 1980), was included within the trigger region. Fig. 2 is a flow cytometry histogram of reactivity of PWM-stimulated and unstimulated PBMCs with the goat anti-human immunoglobulin F(ab')2 prepara-
:i t:
UJ
,,=,
l: i: i:
FO U)
0 CO 0
rr
n< nO LL.
n-
i
ii
,9
:~iiiii::
::!
i
RIGHT ANGLE S C A T T E R
RIGHT ANGLE S C A T T E R
Fig. 1. Cytograms of PBMCs after 5 days in culture in the absence (left) or presence (right) of PWM. In the left cytogram the majority of cells are unstimulated lymphocytes and appear in a circumscribed cluster in the lower left-hand corner of the cytogram. In the right cytogram there are many more cells appearing in areas of greater forward and right angle light scatter. As noted in the text, less than 5% of cells in the preparations are multiple cell aggregates. Thus. cells with greater forward and right angle light scatter are predominantly single cells.
it." U.I El
=E
:3 Z
_d
U.I 0
A
i ~I' k i
10
t
, "~ "t" ," .i i , ,.i,l t
40
80
T
n,., ILl El :3 Z ..I .J iii (3
B
~
120
i ~r-'T--T-"~r---f-~r--r---r-r--[ 160 200 240 INTENSITY
FLUORESCENCE
-I I
II
I0 T T
|
I
40
I
I
I
I
80
I
I
I
I
I
120
FLUORESCENCE
I
I
I
I
I
I
160
I
200
I
I
I
I
I I
24.0
INTENSITY
Fig. 2. Flow cytometric histograms of PBMCs stained with the goat anti-human immunoglobulin F(ab') 2 preparation after 5 days in culture in the absence (A) or presence (B) of PWM. Histogram A reveals only low levels of fluorescence in a few cells. Histogram B shows an increase in cells containing intracytoplasmic antibody. T marks the threshold fluorescence above which a cell is considered to be intracytoplasmic antibody positive. In this experiment, 1.8% of the unstimulated cells and 11.5% of the PWMstimulated cells are positive. The increment in intracytoplasmic antibody-positive cells due to PWM treatment is therefore 9.7%.
41
tion. Reactivity. with this antiserum was minimal for almost all unstimulated cells. In contrast, 5-208 of cells from PWM-stimulated cultures fluoresced more intensely than cells from unstimulated cultures. When PWM-stimulated and unstimulated cells are stained with FITC-conjugated human albumin there is minimal staining of either cell population (Fig. 3). In 5 such experiments, staining with FITC-conjugated human albumin resulted in less than 1% mean difference in fluorescence between PWM-stimulated and unstimulated cells. In addition, the presence or absence of surface immunoglobulin does not influence results obtained with this assay because the fluorescence gain is set substantially below that used for detection of surface antigens. We routinely used the goat anti-human immunoglobulin preparation to stain both the stimulated and unstimulated cells. Selection of the level of fluorescence above which cells were considered positive was the only subjective aspect of this technique. The fluorescence threshold was set at the beginning of the abrupt increase in unstimulated cell numbers on the histogram (point T in Fig. 2). This resulted in between 0.1 and 3.0% of the unstimulated cells being considered positive, with most unstimulated PBMC specimens containing between 0.2 and 2.0% positive cells. After establishment of the fluorescence threshold for unstimulated cells, PWM-stimulated cells were analyzed.
10
JO
80 .7-TTflr~~..~~p 120 FLUORESCENCE
160
200
i
INTENSITY
--I 240 FLUORESCENCE
INTENSITY
Fig. 3. Flow cytometric histograms of PBMCs stained with FITC-conjugated human albumin after 5 days in culture +I the absence (A) or presence (B) of PWM. In this experiment 1.6% of the cells in A and 1.2% of the cells in B are positive.
42 The percentage of unstimulated cells with fluorescence levels above the threshold was subtracted from the percentage of PWM-treated cells exceeding the threshold, yielding a result which expresses the net increase in intracytoplasmic antibody-containing cells induced by the mitogen. Optimum conditions for preparation of cells for analysis by flow cytometry were determined. Varying concentrations of saponin (0-1.0%) and lysolecithin (0-400 # g / m l ) were tested. Saponin, 0.25%, or lysolecithin, 100 ~g/ml, was found to reliably facilitate entry of the FITC-conjugated F(ab')2 fragment into the cells. Serial dilutions of this antiserum revealed that a 1:20 or 1 : 4 0 dilution provided optimal staining. Some staining occurred if PWM-stimulated cells were treated with 2% paraformaldehyde in PBS prior to staining, but the number of PWM-stimulated cells exhibiting fluorescence was usually increased when either saponin or lysolecithin was used to increase cell membrane permeability. In general, little staining occurred when neither paraformaldehyde or a detergent was used. Incubating the cells in paraformaldehyde/saponin or paraformaldehyde/lysolecithin solution for more than 5 min failed to increase the number of fluorescing cells. Incubation for greater than 60 min usually diminished the number of fluorescing cells. Saponin and lysolecithin yielded roughly equivalent values for intracytoplasmic antibody-positive cells. Another detergent, Triton X-100, 0.01%, has been used previously to increase cell permeability for intracytoplasmic antibody staining (Zeile, 1980). We used this agent in a small number of experiments and found that it gave results similar to those obtained with saponin and lysolecithin. The cytogram and histogram of a specimen were stable for at least 3 days after staining and fixation. Fixation with paraformaldehyde prior to staining lymphocyte surface antigens with monoclonal antibodies has been reported to lead to unacceptably high levels of background fluorescence (Lanier and Warner, 1981). We therefore treated cells with lysolecithin or saponin, followed by incubation with the F(ab') 2 fragment prior to fixation with paraformaldehyde. Although the resulting cytograms and histograms were similar to those obtained using our standard protocol, fixation with paraformaldehyde after staining decreased the number of cells enumerated by the flow cytometer by approximately 50%. In some experiments the cytogram was divided into 2 gating regions, a 'lymphocyte' region encompassing the relatively restricted lymphocyte cluster illustrated in Fig. 1A, and a 'blast' region encompassing the area of increased forward and right angle light scatter shown in Fig. 1B minus the 'lymphocyte' region. Table 1 shows that the few positive unstimulated PBMCs appear exclusively in the 'blast' region, but that positive PWM-stimulated PBMCs appear in both the 'lymphocyte' and the 'blast' regions, although the majority are in the 'blast' region. In another experiment, we stained PWM-stimulated and unstimulated PBMCs with the T-cell monoclonal antibodies OKT3, OKT4 and OKT8; aliquots of cells were also stained for intracytoplasmic antibody. Table II illustrates that while the relative number of OKT3, OKT4- and OKT8-positive cells changed little with PWM activation,, the percent of cells which were intracytoplasmic antibody positive increased 5-fold. In addition, absolute numbers of both T-cells and intracytoplasmic antibody positive B-cells increased dramatically following PWM stimulation.
43 TABLE I CYTOGRAPHIC LOCATION OF CELLS C O N T A I N I N G INTRACYTOPLASMIC ANTIBODY Cytographic region
Lymphocyte + blast Blast only Lymph only
% Intracytoplasmic positive cells
Number of cells a
No PWM
PWM
No PWM
PWM
1.8 5.6 0.0
7.7 8.5 3.1
3 711 1014 2 759
22 561 18930 4177
a Cell counts were not adjusted to 1 × 106/ml prior to staining in this experiment. The variation in cell numbers reflects PWM-induced proliferation.
Comparison of flow cytometry and fluorescence microscopy for detection of intracytoplasmic antibody We compared the results obtained with the traditional fluorescence microscopy method of detecting intracytoplasmic antibody with those obtained using the flow cytometer. Prior to staining cells for use in the flow cytometer, a small aliquot was cytocentrifuged, fixed, stained, and counted by fluorescence microscopy. The remainder of the specimen was prepared as previously described and analyzed by flow cytometry. Fig. 4 displays data from one such experiment. In addition, an aliquot of cells stained for use on the flow cytometer was cytocentrifuged, counter stained, and examined by fluorescence microscopy. Cells prepared for flow cytometry were similar in appearance to those prepared for fluorescence microscopy by the traditional method (Fig. 5). The time course of intracytoplasmic antibody appearance as cletermined by flow cytometry (Fig. 6) was similar to that previously reported for immunoglobulin-secreting cells following PWM stimulation of PBMCs (Fauci et al., 1980).
EB V-transformed lines In order to apply our technique to cells stimulated by a B-cell mitogen other than
TABLE II T-CELL SUBSETS A N D INTRACYTOPLASMIC ANTIBODY F O L L O W I N G PWM STIMULATION a
No PWM PWM
OKT3 + cells
OKT4 + ceils
OKT8 + cells
Intracytoplasmic antibody + cells
84.9 b 88.4
60.3 61.7
22.6 25.7
1.2 6.4
a Note that the sum of OKT3 + cells plus intracytoplasmic antibody + cells is less than 100%. This is due to the presence of unstimulated B-cells and macrophages. b All values are % positive.
44
O U u~ O =E
,/
I,-
/
I--
/ /// / /,,/
2
P E R C E N T POSITIVE BY F L O W C Y T O M E T R Y
Fig. 4. Comparison of the percentage of cells containing intracytoplasmic antibody as determined by fluorescence microscopy and flow cytometry. The open circles represent unstimulated PBMCs; closed circles represent PWM-stimulated PBMCs.
PWM. we examined EBV-transformed cell lines for intracytoplasmic antibody. Fig. 7 contains histograms of the p~HR-1 and V1 cell lines, p~HR-1 has been passaged for over a decade; cell line VI was established by EBV transformation of adult T-cell depleted PBMCs 2 months prior to these experiments. Fig. 8 contains photomicrographs of these 2 cell lines stained to reveal intracytoplasmic antibody via fluorescence microscopy. The p~HR-1 cell line is almost totally intracytoplasmic antibody
TABLE 111 MIXING EXPERIMENT USING AN INTRACYTOPLASMIC ANTIBODY-POSITIVE LINE (CA) A N D AN INTRACYTOPLASMIC A N T I B O D Y - N E G A T I V E LINE (RHL-4) Cell Line
% Intracytoplasmic antibody-positive cells
RHE-4 CA 2 : 1 CA : RHL-4 1 : 1 CA : RHL-4 1:2 ,CA: RHL-4
N.A.
Expected
Measured by flow cytometry 1.2 ~'
N.A.
41.2
27.9 21.2 14.5
24.8 20.5 11.5
" Threshold set arbitrarily and held constant throughout the experiment.
45
Fig. 5. Top: PWM-stimulated cells stained for intracytoplasmic antibody by the traditional method and counterstained with methylene blue. Bottom: PWM-stimulated cells stained for use in the flow cytometer, cytocentrifuged and counterstained with methylene blue.
DAY
FOLLOWING
PWM
Fig. 6. Kinetics of intracytoplasmic flow cytometry.
STIMULATION
antibody
appearance
160 FLUORESCENCE
FLUORESCENCE
following
200
stimulation
240
by PWM as determined
by
’
INTENSITY
INTENSITY
Fig. 7. Flow cytometric histogram of the P,HR-1 (A) and VI (B) cell lines stained with the FITC-conJugated polyclonal goat F(ab’), anti-human immunoglobulin antiserum. If the threshold is set so that 0.9% of the P,HR-1 cells are considered positive for intracytoplasmic antibody. then 53.1% of the cells from cell line VI are positive.
47
Fig. 8. Top: cell line P3HR-1, stained for intracytoplasmic antibody by the traditional method. Bottom: cell line VI, stained for intracytoplasmic antibody by the same technique. Note the correlation between intracytoplasmic antibody positivity in these 2 cell lines as determined by flow cytometry (Fig. 7) and fluorescence microscopy (Fig. 8).
4~
negative and cell line VI possesses many positive cells by both techniques. In general, recently transformed cell lines contain many more intracytoplasmic antibody-positive cells than do cell lines which were established several years ago. Flow cytometry of mixed cell fines We selected 2 cell lines for a mixing experiment. Cell line CA, established by EBV infection of T-cell depleted adult PBMCs, contains many intracytoplasmic antibody-positive cells; cell line RHL-4 contains very few positive cells (Fig. 9). Table III shows the results of an experiment in which these 2 cell lines were mixed prior to staining for analysis by the flow cytometer. The experimental results correlate well with those expected. Discussion
We have shown that detection of intracytoplasmic antibody by flow cytometry is a useful technique for quantitation of polyclonal B-cell activation which has certain
49
Fig. 9. Left: cell line CA, stained for intracytoplasmicantibody by the traditional method. Right: cell line RHL-4, stained for intracytoplasmicantibody by the same technique.
advantages over fluorescence microscopy. The production of intracytoplasmic antibody in response to PWM stimulation is not an all-or-none phenomenon. At any point in time a PWM-stimulated specimen contains cells with a wide range of intracytoplasmic antibody contents. Cells containing small quantities of immunoglobulin are difficult to consistently classify as positive or negative by fluorescence microscopy, thus leading to unavoidable subjectivity (Waldmann and Broder, 1982). In addition, counting cells by fluorescence microscopy is a slow and laborious process. Our flow cytometric method, in contrast, provides an objective and reproducible assessment of the fluorescence intensity of each cell. In addition, the flow cytometer is capable of processing thousands of cells in a few seconds, thus improving reproducibility by greatly increasing the sample size of cells studied. The major disadvantage of this technique is the requirement for approximately 1 x 106 cells from each specimen to be assayed.
5{)
In general, our technique yields results correlating well with those obtained by fluorescence microscopy. One minor difference is that the level of intracytoplasmic antibody positivity in unstimulated cells as measured by flow cytometry (0.1 3,0%) is slightly greater than the range of 0-0.5% previously reported using traditional immunofluorescence (Wu et al., 1973). This disparity is probably due to the flow cytometer's ability to detect levels of intracytoplasmic antibody below the threshold of detection with fluorescence microscopy. Flow cytometry is especially useful in examining recently established EBV-transformed lines since many of the positive cells are only weakly fluorescent and are quite difficult to classify reproducibly by fluorescence microscopy (note the large 'shoulder' region in Fig. 7B in comparison to PWM-stimulated PBMCs in Fig. 2B). Detection of intracytoplasmic antibody by flow cytometry depends on increasing the cytoplasmic membrane permeability of lymphocytes in suspension sufficiently to permit FITC-conjugated anti-immunoglobulin F(ab') 2 molecules to enter the cell without causing cell lysis, extensive cell aggregation or significant extracellular leakage of intracytoplasmic immunoglobulin. We used either lysolecithin or saponin as permeabilizing agents in this system. Lysolecithin's effect on cell membrane permeability has been carefully studied (Miller et al., 1979). We found another detergent, saponin, to be equally effective in our system, as well as less expensive. A third agent~ Triton X-100, has been used previously in a flow cytometric system to permeabilize cells for intracytoplasmic antibody staining of myeloma cells using FITC-conjugated anti-human immunoglobulin antiserum (Zeile, 1980). The mechanism of action of these detergents is not well understood, but probably involves insertion of detergent molecules into the cell membrane lipid bilayer, with enhanced permeability being due to pore formation (Weltzien, 1979). We used paraformaldehyde simultaneously with a detergent in order to fix cells in a state of increased permeability prior to staining. Attempts to fix the cells after detergent treatment and immunofluorescent staining resulted in a 50% reduction in cell number, possibly due to detergent-induced cell lysis, aggregation, or adherence to centrifuge tube walls prior to fixation. In summary, we have determined optimal conditions for detection of intracytoplasmic antibody using flow cytometry and have shown that our technique yields results comparable to those obtained by fluorescence microscopy. Flow cytometry offers significant advantages over fluorescence microscopy including rapid analysis of large numbers of cells and objective determination of fluorescence intensity. This technique can be used alone as a monitor of B-cell activation as manifested by antibody synthesis, or in conjunction with other assays of B-cell activation. It should also be possible to utilize this approach to determine immunoglobulin subclass, light chain type or to detect non-immunoglobulin intracytoplasmic antigens. Finally, combination staining using rhodamine and fluorescein-labeled antibodies should be possible and would permit simultaneous monitoring of 2 intracytoplasmic antigens or, alternatively, a surface and an intracellular antigen.
51
References
l~by, W.C., C.A. Ch0ng. ~. Dray and G.A. M01inar0,1075, J. Immun01.115, 1700. Fauci, A.S., G. Whalen and C. Burch, 1980, Cell. Immunol. 54, 230. Greaves, M., G. Janossy and M. Doenhoff, 1974, J. Exp. Med. 140, 1. Hijams, W., H.R.E. Schuit and F. Klein, 1969, Clin. Exp. Immunol. 4, 457. Hoffman, R.A., P.C. Kung, W.P. Hansen and G. Goldstein, 1980, Proc. Natl. Acad. Sci. U.S.A. 77, 4914. Horan, P.K. and L.L. Wheeless, Jr., 1977, Science 198, 149. Lanier, L.L and N.L. Warner, 1981, J. Immunol. Methods 47, 25. Loken, M.R. and A.M. Stall, 1982, J. Immunol. Methods 50, R85. Miller, M.R., J.J. Castellot, Jr and A.B. Pardee, 1979, Exp. Cell Res. 120, 421. Traganos, F., Z. Darzynkiewicz, T. Sharpless and M.R. Melamed, 1977, J. Histochem. Cytochem. 25, 46. Waldmann, T.A. and S. Broder, 1982, in: Advances in Immunology, Vol. 32, eds. F.J. Dixon and H.G. Kunkel (Academic Press, New York) p. 1. Waldmann, T.A., S. Broder, R.M. Blaese, M. Durra, M. Blackman and W. Strober, 1974, Lancet ii, 609, Weltzien, H.U., 1979, Biochim. Biophys. Acta 559, 259. Wu, L.Y.F., A.R. Lawton and M.D. Cooper, 1973, J. Clin. Invest. 52, 3180. Zeile, G., 1980, Cytometry 1, 37.