Isolation of human mononuclear cell subsets by counterflow centrifugal elutriation (CCE)

CELLULAR

IMMUNOLOGY

85, 373-383 (1984)

Isolation of Human Mononuclear Ceil Subsets by Counterflow Centrifugal Elutriation (CCE) I. Characterization of B-Lymphocyte-, T-Lymphocyte-, and Monocyte-Enriched Fractions by Flow Cytometric Analysis

LARRYM. CHRISTINE

WAHL,"* ILDYM.KATONA,RONALDL.WILDER, C. WINTER, BOULOS HAFUOUI, IRWIN SCHER, AND SHARON

M. WAHL

*Laboratory of Microbiology and Immunology, National Institute of Dental Research, and The Arthritis and Rheumatism Branch, National Institute of Arthritis, Diabetes, Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland 20205; The Department of Medicine and Pediatrics, Uniformed Services University of the Health Services, and Infectious Disease Program Center, Naval Medical Research Institute, Bethesda, Maryland 20814=,’ Received October 4, I983; accepted December 13. 1983 Rapid separation of large numbers of human peripheral blood mononuclear celIs into fractions enriched for B lymphocytes, T lymphocytes, or monocytes was accomplished by counterflow centrifugal elutriation (CCE). The first fraction contained 98% of the platelets. Ten additional fractions containing subpopulations of mononuclear cells were collected by sequential increases in the flow rate while maintaining a constant centrifuge speed. Analysis of the fractions using monoclonal antibodies revealed that fraction 2, which was free of esterase-positive monocytes, was highly enriched for B cells. T lymphocytes (OKT3+) were the predominent cell type found in fraction 4. No enrichment for T-lymphocyte-helper (OKT4+) or -suppressor (OKT8+) subpopulations was observed in the lymphocyte containing fractions. Three fractions (7-9), highly enriched for esterase-positive cells, were predominately OKMl+ monocytes with no evidence of selective separation of monocyte subpopulations. Thus, cell fractions enriched for B cells, T cells, and monocytes could be obtained, by utilizing CCE, in large enough quantities to enable analysis of their functional properties. Of particular interest was the ability to separate small, resting B lymphocytes from monocytes.

INTRODUCTION Rapid separation of large numbers of mononuclear cells from human peripheral blood into fractions which are comprised predominately of B cells, T cells, or monocytes is of considerable importance in the analysis of the functional roles of these cells. Several techniques which have previously been utilized to enrich for these populations include adherence to immunoadsorbent surfaces (l-4) nylon-wool columns (5), or a variety of other substrates, and rosetting with sheep erythrocytes (6, 7) or sheep ’ To whom correspondence should be addressed. ’ The opinions and assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Defense (and/or the Department of the Navy, Army, or Air Force). 3 This work was supported in part by Naval Medical Research Institute, Independent Research Protocol MR0000101.1288. 373 0008-8749/84 $3.00 Copyright 0 1984 by Academic R’es, Inc. All rights of nprochnion in any form reserved.

374

WAHL

ET

AL.

erythrocytes coated with antibody and complement (6). However, these methods may initiate cellular activation, alter cellular function, or change the threshold level of activation (8-10). Furthermore, it has been extremely difficult to separate B lymphocytes and monocytes because of their common cell surface receptors and adhesive properties. We have, therefore, adapted the method of counterflow centrifugal elutriation (CCE), a technique which separates cells by size and density, to the separation of these cells. Utilization of counterflow centrifugation for cell separation was first proposed by Lindahl (11) in 1948. Subsequently a counterstreaming centrifuge was designed to concentrate yeast and horse eosinophilic leukocytes (12). McEwen et al. ( 13) made counterllow centrifugation generally available by designing rotors for CCE which could be used with standard preparative centrifuges and, using this equipment, were able to recover 99% of the leukocytes from malaria-infected monkey blood (14). Further design changes in the separation chambers of the rotor (15-17) enabled the isolation of small numbers of cells from lo-20 ml of peripheral blood. This led to the use of CCE for separation of peripheral blood granulocytes (18-23), monocytes (24-30), and lymphocytes (25-27). We have adapted CCE, which has been used primarily for obtaining monocytes (24-30), to the separation of lymphocyte subpopulations and monocytes. The subpopulations were characterized by labeling with affinity-purified and monoclonal antibodies and analyzed on the fluorescence-activated cell sorter (FACS). Because of the large cell numbers that can be separated by CCE (1 to 2 X 109) with excellent recovery (75-95%), functional studies of these isolated subpopulations can be performed (3 1). MATERIALS

AND

METHODS

Cell preparation. Human peripheral blood leukocytes were obtained from 400 ml of whole blood (NIH Blood Bank) or from leukapheresis. The blood was diluted with phosphate-buffered saline (PBS) to 5 X lo6 leukocytes/ml and 35-ml aliquots of this suspension were layered on 20 ml of Ficoll-Paque (Pharmacia Fine Chemicals, Piscataway, N.J.) in 50-ml centrifuge tubes. Following centrifugation at 400g for 30 min, the mononuclear cells at the interface were collected, diluted with PBS, and centrifuged. The cells were resuspended in cold ammonium chloride lysing buffer (0.14 M) (NIH media unit) for 7 min, diluted in PBS, and centrifuged. The cell pellets were resuspended in Dulbecco’s modified Eagle’s medium (DMEM), without Ca*’ or Mg*+ (NIH media unit), containing 100 U/ml of penicillin, 100 pg/ml of streptomycin, 20 mM Hepes, and 1% BSA for CCE. Counterflow centrifugal elutriution. Separation of the peripheral blood mononuclear cells on the basis of size and density was achieved by utilizing the Beckman elutriation system (18, 23, 24). This system consists of a J-21 Beckman Centrifuge, a JE-6 elutriator rotor, a stroboscope (Beckman Instruments, Inc., Palo Alto, Calif.), a Masterflex pump (7553) and pump head (7014) (Cole-Parmer Instrument Co., Chicago, Ill.). The JE-6 rotor contains an inlet line through which the cells enter the rotor, a 4.5-ml horizontal chamber in which the cells collect, and a bypass chamber on the opposite side of the rotor which is connected to an outlet line. Prior to loading the cells on the elutriator, 400 ml of 70% ethanol were pumped through the system for sterilization, followed by 500-700 ml of PBS and 300 ml of the CCE medium. The mononuclear cells were loaded and collected under sterile conditions in a culture hood.

SEPARATION

OF

HUMAN

MONONUCLEAR

CELLS

BY

CCE

315

Mononuclear ceils ( 1 X 10’) in 50 ml (2 X 10’ cells in 100 ml if a second horizontal chamber was utilized in place of the bypass chamber) were pumped directly into the chamber(s) at a flow rate of 6 ml/min with a rotor speed of 1960 f 10 t-pm. This initial flow rate allowed for stratification of the mononuclear cells in the horizontal chamber(s) according to density and size, with more than 97% of the mononuclear cells being retained in the chamber while greater than 98% of the platelets were eluted. Subsequent to complete loading of the mononuclear cells, the medium flow was directed via a three-way valve through a closed 50-ml tube with inlet and outlet lines, which served to dampen pump pulsations. The mononuclear cells were then eluted by sequential increases in the flow rate. At each established flow rate, 165 ml of medium representing one fraction was collected. The fractions were centrifuged and the cell pellets from each fraction were resuspended in DMEM. Aliquots were counted and their size profiled on a ZBI Coulter counter and channelyzer (Coulter Electronics Inc., Hialeah, Fla.). Esterase and d@rential staining. Cells to be stained were suspended in DMEM containing 50% fetal calf serum (2 X lo6 cells/ml) and 200 ~1 of these samples was centrifuged in a cytospin centrifuge (Model SCA-003 1; Shandon Southern Products LTD, Runcom, Cheshire, England). The cells were classified by morphology utilizing Giemsa-Wright stain and by nonspecific esterase staining (32). Zmmunojluorescence staining for cell surface antigens. Monoclonal antibodies to a common surface antigen on human peripheral blood T cells (OKT3), to the helper/ inducer T-cell subpopulation (OKT4), to the suppressor/cytotoxic T-cell population (OKT8), and to a subset of monocytes and null cells (OKMl) were obtained from Ortho Pharmaceutical Corporation (Raritan, N.J.). Monoclonal antibodies to human HLA-DR Ia-like antigens were a product of Becton-Dickenson, (Sunnyvale, Calif.). The monoclonal reagents were used in indirect immunofluorescence with fluoresceinconjugated goat F(ab’)z anti-mouse IgG (Tago, Burlingame, Calif.). To identify surface immunoglobulin-positive lymphocytes, cells were stained directly with fluoresceinconjugated affinity-purified goat F(ab’)z anti-IgD, anti-IgM, and anti-IgG (Tago). For each sample the appropriate antibody control was run to determine nonspecific staining. The control antibodies were from the same source (i.e., either hybridoma or goat), from the same class, and of similar fluorescein/protein ratios and concentrations as the experimental antibody, but directed against an irrelevant antigen. The cells stained with experimental or control antibodies were analyzed for surface fluorescence with a fluorescence activated cell sorter (FACS II, Becton-Dickenson Electronics Laboratory, Mountain View, Calif.). The relative frequency of cells expressing the various cell surface antigens was calculated (33) following computer subtraction of the control from the experimental fluorescence profiles. Fq receptor determination. The percentage of cells bearing surface IgG in the monocyte-enriched fractions was determined by labeling with fluorescein-conjugated F(ab’)2 goat anti-human IgG (Tago). To determine whether the surface IgG (sIgG) was intrinsic or cytophilic, the cells were treated with a pH 2.5 acetate buffer, containing 0.085 M NaCl and 0.005 M KCl, for 1 min after which the pH was restored to 7.4 with 0.1 M Tris buffer containing 1% BSA. This procedure removed more than 90% of the sIgG, indicating that the sIgG was cytophilic and Fcr receptor associated. The acid-treated cells were then incubated with saturating concentrations of rabbit IgG anti-DNP dimers (kindly provided by Dr. Paul Plotz, NIH, Bethesda, Md.) for 30 min and subsequently labeled with fluorescein-conjugated F(ab’*) goat anti-rabbit IgG

376

WAHL

ET AL.

(Cappel Laboratories, Cochranville, Pa.). The binding of IgG dimers was saturable and was inhibited by heat-aggregated human IgG (B. Haraoui, unpublished data). Cells stained with fluorescein-labeled F(ab’)z goat anti-rabbit IgG without prior incubation with rabbit IgG dimers served as controls. The cells were analyzed with the FACS as described above. RESULTS Distribution and size of lymphocytes in eluted fractions. Subsequent to the loading of the peripheral blood mononuclear cells into the elutriator chamber(s), the cells were separated into fractions by sequentially increasing the flow rate to force a given population of cells out of the chamber while maintaining a constant centrifuge speed. A specific flow rate was established for each fraction (Table 1) and 165 ml of medium was collected to ensure removal of the majority of cells that could be eluted at that setting. The cell type(s) collected in each fraction was identified by differential and nonspecific esterase staining (Table 1). In preliminary experiments, it was determined that optimal separation was obtained by advancing the flow rate in 11 increments and thereby collecting 11 pools or fractions of cells. Size profiles of the 11 fractions

TABLE I Flow Rate, Percentage Recovery, Cell-Size Distribution, and Percentage of Cell Types of Elutriation Fractions of Human Peripheral Blood Cells”

Fraction

Flow rate (ml/min)

UNF 1C 2 3 4 5 6 7 8 9 10 11

6.od 7.0 7.7 8.6 9.5 10.6 11.6 13.0 15.3 19.3 19.3

W’M

Cell volume (pm’) and Percentage’ of cell type Percentage of total cells recovered Lymphocytes Monocytes Granulocytes 2.5’ 6.8 14.5 21.4 15.0 11.4 9.4 5.9 5.0 3.2 1.9

(68)

148’(lcQ 166 (100) 180 (>99) 202 (95) 215 (60)

(8) (10)

(12) 195 (23) 195 (23)

(31) 0) (0) (cl) (5) 414’(40) 414 (91) 414 (89) 411 (84) 410 (64) 413 (38)

- 200-280’

(1) (0) (0) (0) (0) (0) (1) (1) (4) (13) (39)

’ The data are representative of 20 elutriator separations of human peripheral blood mononuclear cells. b The percentages (shown in parentheses) of lymphocytes, monocytes, and granulocytes in each fraction were determined by morphology and esterase staining. Fraction 1 was composed primarily of platelets with some lymphocyte contamination. ’ A total of 165 ml was collected for each fraction. The last fraction (11) was collected by stopping the centrifuge to wash out the cells in the chambers. d The various elutriation fractions were collected at the indicated flow rates at a constant rotor speed of 1960 + 10 rpm. ’ The percentage of recovered cells found in each fraction. Recovery of the cells loaded ranged between 70 and 95%. ‘Cell volume in pm’ was determined on a ZBI Coulter counter and channelyzer which was calibrated with 10.1 pm beads.

SEPARATION

OF HUMAN

MONONUCLEAR

CELLS BY CCE

317

1m7.003m403mm7ol CELL

VOLUME

,$I

FIG. 1. Cell size profiles of a representative counterflow centrifugal elutriation of human peripheral blood mononuclear cell fractions (Fl-1 1), as determined on a ZBI Coulter counter and channelyzer.

and the cells associated with these pools are shown in Fig. 1. The volume (pm3) of the cells in each of these fractions is shown in Table 1. The first fraction was collected at a flow rate of 6 ml/min (Table 1) and contained >98% of the platelets. The ability to concentrate the platelets into a single fraction essentially eliminated their appearance in subsequent fractions. Minimal numbers of mononuclear cells were recovered in fraction (F) 1. As determined both by differential staining and by the absence of esterase-positive cells, the second fraction (F2), which was eluted at a flow rate of 7 ml/min, contained approximately 7% of the total eluted cells, 100% of which were lymphocytes. Fractions 3 and 4 were also almost exclusively lymphocytes. Monocytes began to be detected in subsequent fractions. There was a sequential increase in lymphocyte volume from F2 ( 148 pm3) to F6 (2 15 pm3). Analysis of cell suflace antigens in lymphocyte-enriched fractions. The mononuclear cells in elutriator fractions 2-9 were analyzed for cell surface antigens with specific antibodies to identify which cell types and subpopulations were represented in each

378

WAHL

ET AL.

TABLE 2 Percent of Mononuclear Cells in the Elutriation Fractions Stained by Specific Antibodies Antibody specificity Fraction UNF” 2 3 4 5 6 I 8 9

OKT3+ 556 39 66 76 70 54 a 6 5

OKT4+

OKT8+

42 32 49 52 52 38 5 3 3

16 10 17 22 20 17 3 2 2

SIgD+

10 46 19 3 3 3 2 2 1

SIgM’

sIgG+

la’

0KM1+

16 48 19 4 6 15 15 13 13

22 5 5 5 5 26 81 82 79

23 45 20 9 8 22 40 41 36

25 6 5 7 14 32 82 83 78

DUnfmctionated mononuclear cells and the cells in F2-9 obtained by elutriation were stained indirectly with OKT3, OKT4, OKTS, OKMI, or control antibodies and counterstained with fluorescein-labeled goat anti-mouse IgG or stained directly with fluorescein-labeled anti-Ia, anti-h@, anti-IgM, and anti-IgG and were analyzed for surface fluorescence by FACS. b Percentages of positively stained cells are indicated. The data represent the mean percentage of cells stained in the elutriation fractions obtained from 20 separate experiments.

fraction. Fractions 2-4 which were the most enriched for lymphocytes as determined by morphology and the absence of esterase positive cells were also found to bear lymphocyte-specific markers (Table 2 and Figs. 2 and 3). These data demonstrate

FIG. 2. Fluorescence profiles of cells from F2. Cells were stained directly with anti-IgD, anti-&M, antiIgG, and anti-Ia, and indirectly with antibodiesagainst OKT3 and OKMI and were analyzed on the FACS (- - -). Appropriate control antibody staining of the celk is shown in each panel (----).

SEPARATION OKT3

OF HUMAN

MONONUCLEAR OKT4

CELLS BY CCE 7)

379

OKT8

FIG. 3. Fluorescence profiles of cells from F4. Cells were stained indirectly with antibodies against OKT3, OKT4,OKT8, and OKM 1, and directly with anti+@ and anti-Ia antibodies (- - -) or with the appropriate control antibody (-).

that not only can the lymphocytes be separated from the monocytes, but also that B cells and T cells can be greatly enriched in separate fractions by this technique. Fraction 2 contained approximately 50% B cells as demonstrated by the fluorescence profiles (Fig. 2) and percentage of cells (Table 2) stained by anti-&D, anti-IgM, and anti-Ia, which are characteristic of resting B cells. Although some variation occurred in separating cells from different individuals, the localization of B cells into fraction 2 was consistent as is evident in Table 2 which represents 20 different expeiments. The cells in F2 were the smallest lymphocytes (Table 1). Non-B cells in F2 were mainly OKT3+ cells and some OKMl’ cells. Since the cells in this fraction were esterase negative and cell sorting experiments of these OKMl+ cells indicated that they were also sIgG+, these OKMl” cells probably represent null cells (34). Additionally, the fluorescence intensity of these OKMl+ cells was less than that of monocyteenriched fractions. Thus, an enriched population of B cells which was monocyte depleted was obtained by CCE. The percentage of B cells in F3 decreased to approximately 20%, with a corresponding increase in OKT3+ cells (Table 2). Fractions 4 and 5 contained primarily OKT3+ cells. Few B cells were detected in these fractions as evidenced by the significant decrease in sIgD, sIgM, and Ia staining (Table 2). Fraction 4 contained < 1% monocytes as determined by esterase staining. The OKM 1+ cells detected are of low-intensity staining and likely represent null cells. Approximately 5% esterase-positive cells were detected in F5. Since the OKMl’ antibodies and those reactive with B-cell antigens stain a total of only lo-20% of the cells in F4 and 5, it was apparent that the 70-76% of the cells stained by OKT3 did not account for all the T cells. Examination of the lymphocyte subsets using monoclonal antibodies revealed that no preferential enrichment for either OKT4+ or OKT8+ T

380

WAHL

ET AL.

lymphocytes was observed in F2-6 (Table 2). The 2:l ratio of OKT4+ to OKT8+ was consistent in F2-6. Distribution and size of monocytes in eluted fractions. Monocytes in the elutriated fractions were identified by morphology and esterase staining. Fractions 1-3 were monocyte free, while F4 contained ~1% monocytes (Table 1). Monocytes increased to 5% in F5 and to 40% in F6. Fractions 7-9 were enriched for monocytes (84-91%). The decrease in monocytes in FlO and 11 was accompanied by an increase in large lymphocytes and also granulocytes, with the latter comprising 39% of the cells in F 11. Size determinations for the cells in each fraction revealed that the monocytes were separated on the basis of density since the volume of these cells was essentially the same (414 pm3) in F6-11. Analysis of cell surface antigens and Fey receptors on the cells of the monocyteenriched fractions. Since F7-9 were predominantly monocytes, these cells were characterized for surface antigens commonly associated with monocytes. As shown in Table 2, and in the sorter profiles represented by F9 (Fig. 4), a high percentage of these cells were OKMl’, Ia+, and sIgG+. While 84-9 1% of these cells were esterase positive, 76-83% were OKMl’, indicating that not all of the monocytes were identified by this monoclonal antibody. The majority of the cells stained with anti-IgG (y) most likely represent cytophilic IgG and associated with Fey receptors on the monocytes, since most of the contaminating cells in these fractions are T cells (5-10%) (Table 2) with only l-2% sIgD+ cells, which may represent a few large activated B cells. The binding of cytophilic IgG to Fcr receptors on monocytes was demonstrated by removing the IgG from the monocytes in F7-9 by acid treatment and by the ability to restore binding by adding rabbit IgG dimers to these cells (Table 3). No difference in the percentage of Fey receptors between F7 and 9 was detected. The IgM+ cells OKMl

I

-.

I.

7

bG

OKT3

I

%.AL

am

.. 400

en0

alo

1m

al0 RELATIVE

400 FLUORESCENCE

PO la0 INTENSITY

FIG. 4. Fluorescence profilesof cells from F9. Cells were stained with antibodies against OKMl , Ia, IgG, OKT3, IgD, and IgM (- - -) or with the appropriate control antibody (-).

SEPARATION

OF HUMAN

MONONUCLEAR

CELLS BY CCE

381

TABLE 3 Fc, Receptor Determination in Monocyte-Enriched

Fractions0

sIgG-positive cells Fraction 1

8 9

Pre-acid

Post-acid

Addition of IgG dimers

826 78 83

3 4 4

85 83 86

’ Cells from F7 to 9 obtained by elutriation were assayed for Fc, receptors as described under Materials and Methods. b The data are expressed as percentages and are from a representative experiment.

in F7-9 may represent cytophilic IgM on the monocytes. The median fluorescence intensity of the IgM staining cells in M-9 was severalfold lower than that of the IgM+ cells in F2-3. DISCUSSION The data presented here demonstrate that large numbers of peripheral blood cells can be rapidly separated into B-lymphocyte-, T-lymphocyte-, and monocyte-enriched cell populations by the CCE technique. An additional advantage of this procedure is the ability to isolate >98% of the platelets in the first fraction, thus significantly decreasing platelet contamination in subsequent mononuclear fractions. Depletion of platelets is particularly important when determining the production of biologically active mediators by lymphocytes and monocytes since related factors may also be produced by platelets (35-37). The absence of detectable monocytes, in the B-cell-enriched population in F2 is of interest since most purification techniques for B cells either fail to eliminate monocytes (nylon-wool columns and negative rosetting) or risk possible activation of B cells (EAC rosetting and immunoglobulin binding). The absence of monocytes in F2 permits further purification of B cells by additional negative selection techniques such as T-cell rosetting (E-positive rosettes). The second fraction contained the smallest (148 pm3) lymphocytes which is consistent with the observation that resting B cells are small (38). The inability of Contreras et al. (27) to specifically isolate B cells using CCE at a centrifuge speed (2030 rpm) approximately the same as ours (1960 rpm) may be related to their initial flow rate (10 ml/min), which was faster than the 6 ml/min utilized to deplete platelets and the subsequent 7 ml/min required to obtain the enriched B-cell fractions. Griffith (39) reported isolation of an enriched B-cell fraction on CCE by utilizing varying BSA gradients. Fraction 4 represented tbe most T cell enriched population. Less than 1% monocytes (Table l), a few B cells, and OKMl’ cells (Table 2), which most likely represent null cells (3 I), were also present. It was evident that not all the lymphocytes in F2-4 were accounted for by the specific monoclonal antibodies which suggests that there may be subpopulations of lymphocytes which lack these antigens. Alternatively, a portion of the specific fluorescence may have overlapped with the fluorescence of the control antibody and as a result was subtracted from the total. The lymphocytes in F4 were

382

WAHL

ET

AL.

found to be functionally monocyte depleted by their inability to respond to soluble antigens (3 1). Fractions 7-9 were highly enriched for monocytes (84-9 1%). A significant percentage (-20) of the total monocytes was also found in F6 and could be separated from the lymphocytes in F6 by collecting the first 55-65 ml (lymphocyte rich) of F6 separately from the subsequent monocyte enriched 100 ml (data not shown). The volumes (414 pm3) of the monocytes in F6-11 were very similar. This is in contrast to a bimodal distribution of small and large monocyte populations which has been observed by Sanderson (24) and Yasaka (30) utilizing CCE. Some of these differences in monocyte size may be related to osmolarity differences in elutriation media or cell identification techniques. Monocyte subpopulations in the fractions were not apparent based on monoclonal antibodies staining or Fey receptor expression. While these findings demonstrate that human peripheral blood cells were separated into enriched B cell, T cell, and monocyte populations by countertlow centrifugal elutriation, they also show that small resting B cells were isolated free of monocytes. Since this procedure does not rely on adherence or binding of cells to specific immunoadsorbant surfaces, the risk of cell activation during isolation is minimized. Rapid separation of mononuclear cells into their subpopulations makes it possible to perform multiple functional assays in a single experiment (31). ACKNOWLEDGMENTS The authors thank Dr. Ann Sandberg and Dr. Dov Pluznik for their critical review of the manuscript, Dr. Fred D. Finkelman for his helpful comments, Mr. Neil Hardegan for assistance with the FACS analysis, and Mrs. Elizabeth Walter and Ms. Janice Forrest for their preparation of the manuscript.

REFERENCES 1. 2. 3. 4. 5. 6.

Wofsy, L., Kimura, J., and Truffa-Bachi, P., J. Immunol. 107, 725, 1971. Schlossman, S. F., and Hudson, L., .I. Immunol. 110, 313, 1973. Chess, L., MacDermott, R. P., and Schlossman, S. F., J. Immunol. 113, 1113, 1974. Mage, M. G., McHugh, L. L., and Rothstein, T. L., J. Immunol. Methods 15, 47, 1977. Eisen, S. A., Wedner, H. J., and Parker, C. W., Immunol. Commun. 1, 571, 1972. Mendes, N. F., Tolnai, M. E. A., Silveira, N. P. A., Gilbertsen, R. B., and Metzgar, R. S., J. Immunol.

7. 8. 9. 10. 11. 12. 13. 14.

Bentwich, Z., and Kunkel, H. G., Transplant. Rev. 16, 29, 1973. David, J. R., Fed. Proc. 34, 1730, 1975. Arvilommi, H., and Riisanen, L., Nature (London) 257, 144, 1975. Wahl, S. M., Iverson, G. M., and Oppenheim, J. J., J. Exp. Med. 140, 1631, 1974. Lindahl, P. E., Nature (London) 161, 648, 1948. Lindahl, P. E., and Lindahl, K. M., Experientia 11, 310, 1955. McEwen, C. R., Stallard, R. W., and Juhos, E. Th., Anal. Biochem. 23, 369, 1968. McEwen, C. R., Juhos, E. Th., Stallard, R. W., Schnell, J. V., Siddiqui, W. A., and Geiman, Q. M., J. Parasitol. 57, 887, 1971. Grabske, R. T. Fractions, Beckman Technical Bulletin 1, 1, 1978. Sanderson, R. J., Bird, K. E., Palmer, N. F., and Brenman, J., Anal. Biochem. 71, 615, 1976. Sanderson, R. J., and Bird, K. E., In “Methods in Cell Biology” (D. M. Prescott, Ed.), p. 1, Academic Press, New York, 1977. Persidsky, M. D., and Olson, L. S., Proc. Sot. Exp. Biol. Med. 157, 599, 1978. Lionetti, F. J., Hunt, S. M., Lin, P. S., Kurtz, S. R., and Valeri, C. R., Transfusion 17, 465, 1977. Jemionek, J. F., Contreras, T. J., French, J. E., and Hartwig, V., Exp. Hematol. 6, 558, 1978. Nunn, A. D., and Gagne, G., Transfusion 18, 599, 1978. Contreras, T. J., Jemionek, J. F., French, J. E., and Shields, L. J., Transfusion 19, 695, 1979. Jemionek, J. F., Contreras, T. C., French, J. E., and Shields, L. J., Exp. Hematoi. 6, 801, 1978.

111, 860,

15. 16. 17. 18. 19. 20. 21. 22. 23.

1973.

SEPARATION

OF HUMAN

MONONUCLEAR

CELLS BY CCE

383

24. Sanderson, R. J., Shepperdson, F. T., Vatter, A. E., and Talmage, D. W., J. Immunol. 118, 1409, 1977. 25. Fogelman, A. M., Seager, J., Edwards, P. A., Hokom, M., and Popjak, G., Biochem. Biophys. Rex Commun. 76, 167, 1977. 26. Norris, D. A., Morris, R. M., Sanderson, R. J., and Kohler, P. F., J. Immunol. 123, 166, 1979. 27. Contreras, T. J., Jemionek, J. F., Stevenson, H. C., Hartwig, V. M., and Fauci, A. S., Cell. Zmmunol. 54, 215, 1980. 28. Weiner, R. S., and Shah, V. O., J. Immunol. Methods 36, 89, 1980. 29. DeMulder, P. H. M., Wessels, J. M. C., Rosenbrand, D. A. J., Smeulders, J. B. J. M., Wagener, D. J. Th., and Haanen, C., J. Immunol. Methods 47, 31, 198 1. 30. Yasaka, T., Mantich, N. M., Boxer, L. A., and Baehner, R. L., J. Immunol. 127, 15 15, 1981. 31. Wahl, S. M., Katona, I. M., Stadler, B. M., Wilder, R. L., Helsel, W. E., and Wahl, L. M., Cell. Immunol. 85, 384, 1984. 32. Koski, I. R., Poplack, D. G., and Blaese, R. M., In “In Vitro Methods in Cell-Mediated and Tumor Immunity” (B. R. Bloom and J. R. David, Eds.), p. 359. Academic Press, New York, 1976. 33. Finkelman, F. D., and Scher, I., J. Immunol. 122, 1757, 1979. 34. Breard, J., Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman, S. F., J. Immunol. 124, 1943, 1980. 35. Ross, R., Glomset, J., Kariya, B., and Harker, L., Proc. Natl. Acad. Sci. USA 71, 1207, 1974. 36. Heldin, C.-H., Wasteson, A., and Westermark, B., Exp. Cell Rex 109, 429, 1977. 37. Heldin, C.-H., Westermark, B., and Wasteson, A., Proc. Natl. Acad. Sci. USA 76, 3722, 1979. 38. Howard, J. C., Hund, S. V., and Gowans, G. L., J. Exp. Med. 135, 200, 1972. 39. Griffith, 0. M., Anal. Biochem. 87, 97, 1978.