Volumetric and Functional Heterogeneity

of Human Monocytes

EDWARDB. ARENSON,JR., MARTINB. EPSTEIN, and ROBERTC. SEEGER,Department ofPediatrics and Department of Radiological Sciences, University of California,Los Angeles School of Mfedicine and Physics Department, California StateUniiversity, Los Angeles, Los Angeles, California 90024

A B S T R A C T Volume analysis of purified humanblood monocytes revealed distinct populations of largeand small cells. Computer curve fitting suggested athird, intermediate-sized population. These monocyteswere designated N1, M2, and M3 in order of increasingsize, and their approximate volumes were 150, 250,and 480 ,um3, respectively. The three subpopulationswere present in all 30 normal individuals tested. Twonew techniques were developed that separate mono-cytes into M1 + M2 and M3 fractions; one used prefer-ential incorporation of carbonyl iron particles by M3cells and the other used the selective aggregationof N13 cells by thrombin in the presence of platelets.The chemotactic response to zymosan-activated humanserum by total monocytes, M, + MI2 monocytes, and M3monocytes was determined by the agarose platemethod. In all experiments M3 monocytes were 10-fold more responsive than MI + M2 monocytes andwere significantly more so than total monocytes. Thesefindings suggest that M3 cells are the major subpop-ulation capable of directional migration. This investi-gation establishes the existence of volumetricallydefinable subpopulations of human monocytes thatare functionally distinct.

This work was presented in part at the 63rd Annual Meetingof the Federation of American Societies of ExperimentalBiology, Dallas Texas, April 1979 and published as an abstractin 1979. Fed. Proc. 38: 1170.

Dr. Arenson was initially supported by National Institutesof Health training grant CaO9120 and is currently supportedby CA23175 and CA16042 from the Natinal Cancer Institute,Department of Health, Education, and Welfare. Dr. Epsteinwas partially supported by Associated Western Universitiesand Dr. Seeger is the recipient of Research Career Develop-nent award CA00069 from the National Cancer Institute,Department of Health, Education, and Welfare. Addressrequests for reprints to Dr. Arenson.

Received for publicationt 26 June 1979 and in revisedform 19 November 1979.

INTRODUCTION

The mononuclear phagocyte system includes bonemarrow promonocytes, circulating blood monocytes,and tissue macrophages. The spectrum of functionsattributable to mononuclear phagocytes has broadenedmarkedly, and there is increasing evidence that thesecells are metabolically, physically, and functionallyheterogeneous (1). In this context, it is reasonable tohypothesize that the mononuclear phagocyte systemcontains unrecognized subpopulations of cells that arefunctionally distinct. Experiments reported in thiscommunication demonstrate that human peripheralblood contains volumetrically distinct subpopulationsof monocytes with markedly different chemotacticcapabilities.

METHODSMonocyte purification. Washed mononuclear leukocytes

(MNL)l that contained lymphocytes and monocytes obtainedby centrifugation of heparinized blood over Ficoll-Hypaque(Pharmacia Fine Chemicals, Inc., Piscataway, N. J.) (2) weresuspended in RPMI-1640 medium (Grand Island BiologicalCo., Grand Island, N. Y.) that contained 20% human ABserum at a concentration of 3-5 x 106 cells/ml. 5 ml of thissuspension was placed in flasks (No. 3014 Falcon Labware,Div. of Becton, Dickinson & Co., Oxnard, Calif.), gassed with5%CO2 in air, and centrifuged (5 min at 60 g) onto the 25-Cm2plastic surface. After incubation for 30 min at 37°C, nonad-herent cells were removed by vigorously washing (four times)with warm Hanks' balanced salt solution (HBSS). Adherentcells then were removed by incubation for 10-20 min inRPMI-1640 medium that contained 20%fetal calf serum (FCS)and 0.4% lidocaine (Elkins-Sinn, Inc., Cherry Hill, N. J.)followed by gentle squirting. Cells were washed once,counted, and resuspended in HBSSfor analysis. Cell popula-tions obtained in this manner were >90% viable and con-

'Abbreviations used in this paper: FCS, fetal calf serum;HBSS, Hanks' balanced salt solution; MNL, mononuclearleukocytes.

J. CGn. Inivest. (D Thle American Society for Clinical Investigation, Inc. * 0021-9738/80/03/0613/06 $1.00Voltume 65 March 1980 613-618

613

tained 89±5% monocytes by nonspecific esterase staining(3). Phagocytosis of opsonized yeast or staphylococci wasused as an additional monocyte marker in initial experimentsand correlated closely with nonspecific esterase staining.Fewer than 5% granulocytes were present by morphologicexamination (Wright's stain), and B cells never exceeded 6%as determined by fluorescent staining for membrane immuno-globulin. This method isolated virtually all monocytes thatwere present in MNL, because <2% of nonadherent cellswere monocytes by nonspecific esterase staining, Wright'sstain morphology, and readherence to plastic.

Monocyte volume analysis. Volume analyses were per-formed on purified monocytes with a multiparameter cellsorter/analyzer built according to the design of Steincampet al. (4). Before analysis, monocytes were gently flushedthrough 25-gauge needles and observed microscopically toinsure that only single cells were analyzed. Dilute sus-pensions of monocytes (5 x 105/ml in 2 ml) were incubatedfor 10 min in fluorescein diacetate (12 ug/ml in HBSS) andfiltered through a 40-,um nylon mesh before analysis. Volumeanalyses were gated on fluorescence to insure that only viablenucleated cells were analyzed (5). Cell numbers were re-corded over 200 cell volume channels. Each volume analysiswas calibrated with 10-gm diameter latex beads (10-,um stand-ard bright fluorespheres; Coulter Electronics, Inc., Hialeah, Fla.

Separation of monocyte subpopulations. Two newmethods were developed for separating monocyte subpop-ulations. The first was based upon the hypothesis that vol-umetrically defined monocyte subpopulations would differin their ability to incorporate carbonyl iron particles. Totalmononuclear cells were suspended in 5 ml of RPMI-1640at a concentration of 10 x 106/ml and combined with 10 mlof RPMI-1640 that contained 25% FCS and 0.5% carbonyliron (iron powder SF special 11-63771, lot 64; GAF Corp.,NewYork) (6). After incubating on a rotor at 37°C for 15 min,iron-laden cells and excess iron were completely removedby passing the suspension over a magnet followed by cen-trifugation over Ficoll-Hypaque (specific gravity 1.077)at 200 g for 15 min. Iron-free MNL containing residualmonocytes were harvested, and iron-laden cells were dis-carded. Monocytes in this iron-free MNLpreparation wereeither assayed directly for chemotactic activity or wereanalyzed volumetrically after purification by adherence. Pur-ified adherent cells were 76±10% (10 donors) nonspecificesterase positive.

The second method was based upon the observation thataggregates that contained monocytes formed when MNLwere incubated with thrombin in the presence of platelets(7). MNL, prepared from heparinized blood by centrifugationover Ficoll-Hypaque, were suspended in HBSSthat contained0.5% FCS at a concentration of 10-20 x 106 cells/ml and 1 mlwas distributed into polystyrene tubes (No. 4-978-148;Fisher Scientific Co., Pittsburg, Pa.). 10 ,ul of thrombin(100 U/ml in normal saline and 50% glycerol; Thrombin-Topical, lot. 916923A; Parke, Davis & Co., Detroit, Mich.)was added to each tube. The tubes were capped and rotatedat room temperature for 1-2 min until large aggregatesof platelets and leukocytes formed. These aggregates formeda pellet after centrifugation (1,000g, 4 s, Fisher model 59centrifuge). Cells in the supernatant fluid were placed inclean Fisher tubes, cell pellets were disaggregated and re-suspended by gentle pipetting, and both fractions werewashed twice with HBSS. Monocytes in pellet and super-natant fractions were assayed directly for chemotactic activityor were analyzed volumetrically after purification by ad-herence. Purified adherent cells from the supernatant and pelletfractions were 69±19% and 93±2% nonspecific estereasepositive (five donors), respectively.

Monocyte chemotaxis. The chemotactic activity of wholeMNLwas compared with that of MNLthat contained selectedmonocyte subpopulations. MNL were tested instead ofpurified monocytes to minimize manipulations that mightdifferentially affect monocyte subpopulations. With theexception of the iron depletion method, in which wholeMNL were not exposed to iron, all cell populations weresubjected to exactly the same experimental manipulations.Volume analyses were performed on monocytes purifiedby adherence from each MNL preparation to corroboratesuccessful enrichment for monocyte subpopulations.

Monocyte chemotaxis was assayed by the agarose methodas previously reported (8). Briefly, 35-mm petri dishes thatcontained 0.75% electrophoresis-grade agarose (FisherScientific Co.) in medium 199 with 10% FCS were prepared,and three wells (4-mm Diam) were punched 2 mmapart.Unfractionated MNLor MNLenriched for monocyte sub-populations by iron depletion or thrombin aggregation weresuspended in medium 199 with 10% FCS and the proportionof monocytes in each preparation was assessed by morphology(Wright's stain) and nonspecific esterase staining. The per-centage of monocytes in each preparation was equalized byadding autologous iron-depleted nonadherent lymphocytes(>98% lymphocytes by Wright's and nonspecific esterasestaining), and 105 monocytes in 10 ,ul were added to centralwells of triplicate plates. 10 ,ul of zymosan-activated humanserum was added to the left well as the chemoattractant,and 10 ,ul of normal saline was added to the right well as thecontrol. After incubating plates for 16 h at 37°C in an humidi-fied environment that contained 5%CO2 in air, the numberof cells migrating toward the chemoattractant and the controlwas determined by using a 5 x 5-mm microscope eyepiecegrid to count cells between central and lateral wells. Themean± 1 SD of specifically migrating cells (number towardchemoattractant minus number toward control) was calculatedfor triplicate plates.

RESULTS

Identification of monocyte subpopulations. Repre-sentative cell volume distributions of purified mono-cytes from four healthy adults are shown in Fig. 1.Two major peaks of large and small cells were readilyapparent. However, computer analysis with optimalGaussian curve fitting required a third peak betweenthe two major peaks. Hence, three peaks, designatedMI, M2, and M3 in order of increasing size, were usedfor distribution analysis; their approximate meanvolumes were 150, 250, and 480 um3, respectively.The volume of Ml and to a lesser extent M2 monocytesoverlapped that of purified lymphocytes, whereas M3monocytes were significantly larger. A distinct peak ofcells of M3volume was readily identified in whole MNLpreparations. For the purpose of functional analysis,Ml and M2 monocytes were regarded as a single pop-ulation, MI + M2.

A summary of our analysis of purified monocytesfrom 30 normal individuals is provided in Table I.Ml, M2, and M3cells always were present with percentdistributions of 28±15, 23±14, and 49±12, respec-tively. The percentage of cells in each of these sub-populations exceeded that of nonmonocytes (11±+-5%)

614 E. B. Arenson, Jr., M. B. Epstein, and R. C. Seeger

'I

LLJ

Dz

LLJ-jC)

bJ

LLJ

II I-i

lL ...-IV|v* ,.|VIIV Cl:L: -IVI IvI: .v-_: -,IV v ,

. 2915 iii. 56 -i lH ......II. 8 71

50 100 150 50 100 150

CHANNELNUMBER(cell volume)

FIGURE 1 Cell volume distributions of purified monocytesfrom four normal adults. Channel numbers correlate directlywith cell volume. Each distribution was calibrate-d with 10 ,li.tDiam (524 ,m3) latex beads the position of which is indi-cated by an arrow (4). Approximate diameters and volumesfor M1, M2, and M3 cells are 7 ,um and 150 ,tm3, 8 gium and 250,um3, and 9 ,tm and 480 ,um3, respectively. Mean channelNo. + 1 SD (l - I) as well as relative percentage of cellsare shown for each peak. To calculate percentages of cellsin each peak, volume data were represented by three ratherthan two Gaussian distributions because three always gavethe best fit by chi-square analysis; distribution parameterswere optimized with a PDP 11/10 minicomputer using anonlinear least-squares algorithm; areas under volume peakswere computed using these parameters.

in the adherent cell preparations; therefore., monocyteswere abundant in each peak. For individuals of age20-50 yr, there were no sex-related differences inmonocyte volume distributions. Our data for a fewhealthy individuals of different age groups (Table I)suggested that their monocyte volume distributionswere similar to the 20-50-yr group.

To determine if these subpopulations of monocyteswere reproducibly present, we performed multipleanalyses on several normal individuals. Monocytesubpopulations were consistently present for periods

TABLE IPerCenttage of AlotiocIltes i All, A12, Mf1 + Al12, antd A13 Peaks*

M1onioc\tes in each peak

Age N1, \12 M, + M'2 \13

20-50 Male (i = 19) 25±16 25±14 50+16 50 +1520-50 Female (i = 11) 26±16 21+ 15 47±18 53+ 19<10 (ni = 3) 40±21 18±12 58±13 42±13>50 (ni = 5) 31± 14 16± 17 47+4 53±3All ages (i = 38) 28±15 23±14 51± 15 49±12

* 30 determinations o01 22 individuals betweeni 20 and 50 yrof age were analyzed accor(linlg to sex. Three dleterminationson children <10 yr an(d five dleterminiationis on adults over50 yr are shown separately for coimparisoni. Each value ismean±I SD.

up to 6 m11o; for example, four determuinations on oneindividual resulted in 20±8, 26+12, ancd 54+10% M1,M12, and \13 cells, respectively. These experimentsalso demiionstrated that monocyte subpopulations ofsome in(livi(luals could be consistently at one extremeof the normal range (e.g., M,, M\I2, and M3 of 17±3,4+5, and 79+8%).

Separation of nifonoct te sub popiulations. Mlethodswere developed for separating Ml + MN2 andl M3 sub-populations so that functionial capabilities could be(letermiinecl. Incubation of MNLwith carbonyl ironparticles (15 min, 37°C) and then complete separationof iron-laden cells removed 80% of the adherent mono-cytes. Remnaiining adherent cells were 76+10% non-specific esterase positive, and .74% were M, cellsby v\olumnetric analysis (Fig. 2). Variable numbers ofM2 cells were present, but M3 cells were virtuallyahsent. Incorporation of carbonyl iron by M3moniocytescould be abrogated by performing the incubation at4°C or in the presence of 1 mMN-ethylmaleimide,which prevents phagocytosis (9). This suggested thatthe differential ability of monocyte subpopulations toincorporate carbonyl iron involved phagocytic diver-sity rather than a simple difference in surface areaand binding properties.

A second method was developed with thrombin,which provided M1 + M2 and M3 subpopulations. In-cubation of MNLwith thrombin in the presence ofplatelets caused the formation of aggregates that con-tained monocytes that were readily pelleted by cen-trifugation (7). Volume analysis of adherent monocytesfrbm both the pellets and supernates demonstratedthat the supernatant cells (69+±19% nonspecific esterasepositive) contained predominantly Ml + M2 (.80%)monocytes, whereas those in the pellet (93±+2% non-specific esterase positive) were M3 monocytes (.90%)(Fig. 3).

Alonocyte Heterogeneity 615

LiimDz

-LJ

LUJ

'I

LUi

a b

.1026 64 ::74::-26

50 150 5.0 1S0CHANNELNUMBER(cell volume)

FIGURE 2 Differential incorporation of carbonyl ironparticles by monocyte subpopulations. Monocytes werepurified from total mononuclear leukocytes (a) and from ali-quots of these cells which were previously depleted of ironincorporating cells (b). Iron incorporating cells were removedafter incubating MNLat 37°C with 0.33% iron for 15 min(b). Volume analysis was performed as' detailed in the textand Fig. 1.

Chemnotactic responses of inonocyte subpopulations.Directional migration is an important function ofmononuclear phagocytes. Therefore, we asked whethervolumetrically defined subpopulations of monocytesdiffered in their chemotactic capabilities. To minimizethe likelihood that chemotactic differences betweenmonocyte subpopulations were due to purification pro-cedures, two independent methods were used to pre-pare monocyte subpopulations.

Preferential depletion of M3 monocytes with car-bonyl iron allowed the comparison of total monocytesto enriched M1 + M2monocytes (Table II). The chemo-tactic response of total monocytes was significantlygreater than M1 + M2 monocytes in all cases. By usingthe thrombin aggregation technique, total monocytes,enriched M3 monocytes, and enriched M1 + M2 mono-cytes could be compared (Table III). Brief exposureto thrombin did not significantly alter monocyte chemo-taxis because the chemotactic response of untreatedtotal monocytes was the same as thrombin-treatedtotal monocytes. Comparison of monocyte subpop-ulations demonstrated that the chemotactic response ofM3 cells was significantly (P < 0.05) greater than totalmonocytes and 10-fold greater than M, + M2 cells(P < 0.001). This observation corroborated the sug-gestion from the iron depletion experiments that M3monocytes were the most chemotactically active sub-population.

DISCUSSIONOur experiments clearly demonstrated the existenceof large and small human blood monocytes. Because

aTOTAL

MONOCYTES

LUJm

z

J

Lg

LUI

LUI

. MI+M2 M3ENRICHED ENRICHED i

:MI M2: M MI M349. 33 I..I 2 98t-

............ I.....

.....

CHANNELNUMBER(cell volume)

FIGURE 3 Differential aggregation of monocyte subpopula-tions by thrombin. Monocytes were purified from the pelletsand supernates of MNLsuspensions after thrombin aggrega-tion and centrifugation as described in the text. a shows thevolume distribution of total monocytes after thrombin treat-inent; although the M1 peak appears larger than the M2 peak,the M2 peak is actually larger owing to its greater width

mean channel No.-+1 SD). Volume distributionsof supernatant (b) and pellet (c) monocytes demonstrateenrichment of M, + M2 monocytes in the supernate andenrichment of M3 monocytes in the pellet.

TABLE IIChernotactic Activity of Total Monoct tes vs. EnrichedM1 + M2 J'otnoct tes Prepared by Iron Depletion Method*

Donor Cell preparation No cells imiigratinigt

1 Total monocytes 129+9M1 + M2 monocytes 23±+-5

2 Total monocytes 158±23Ml + M2 monocytes 38+7

3 Total monocytes 88+15M1 + M2 monocytes 27+9

* Mononuclear leukocytes containing either total monocytesor >72% MI + M2 monocytes were compared for theirchemotactic activity. M3 monocytes were removed by theiron depletion method.t Mean±+l SD of cells migrating specifically toward chemo-attractant for triplicate plates.

616 E. B. Arenson,Jr., M. B. Epstein, and R. C. Seeger

TABLE IIIChemotactic Activity of Total, Enriched M3, and Enriched

M1 + M2 Monocytes Prepared by ThrombinAggregation Method*

Donor Cell preparation Cells migratingt

1 Total monocytes 104+6Total monocytes (thrombin) 109+13M3 monocytes 247-+11M1 + M2 monocytes 22±5

2 Total monocytes 275+7M3 monocytes 306+17M1 + M2 monocytes 28±2

3 Total monocytes 134+9M3 monocytes 202±11M1 + M2 monocytes 17±3

* MNL that contained either total monocytes, -90% M3monocytes, or 280% M1 + M2 monocytes were comparedfor their chemotactic activity. Enriched M1 + M2 and M3monocytes were prepared by the thrombin aggregationmethod.4 Mean±1 SD of cells migrating specifically toward chemo-attractant for triplicate plates.

computer analysis of volume distributions suggestedthe presence of a third, intermediate-sized population,we termed these subpopulations MI, M2, and M3 inorder of increasing size. All three subpopulations werepresent in 30 normal individuals. In chemotaxis tests,large M3 monocytes were 10 times more responsivethan the smaller Ml + M2 monocytes.

Considerable evidence supports the contentionthat these volumetrically defined subpopulations ofadherent cells are in fact subpopulations of monocytes.Adherent cells containing these three subpopulationswere 89+5% monocytes by nonspecific esterasestaining and phagocytosis of yeast and staphylococci.Because the percentage of cells in the MI, M2, and M3peaks always exceeded the percentage of nonmono-cytes, it was clear that most of the cells in each sub-population were monocytes. Additional evidence sup-porting the identity of these cells as monocytes in-cluded the ability of carbonyl iron particles to eliminate95% of the adherent cells under optimal conditionsand the differentiation of enriched M, + M2 cells intomacrophages in cell culture (unpublished data).

Although it was clear that adherent Ml, M2, and M3cells were indeed monocytes, it was conceivable thatvolumetric differences could be an artifactual effect ofisolation and purification procedures rather than a truecharacteristic of monocyte subpopulations. However,extensive volume analysis of buffy-coat leukocytes,MNL, and cells obtained at each step of the purificationprocess indicated that Ml, M2, and M3 monocyteswere normally present in blood and that purification

procedures neither created nor appreciably alteredthese volume differences (data not shown). Thesefindings strongly supported the existence of volumet-rically definable monocyte subpopulations.

Having identified volumetric subpopulations ofmonocytes, we sought functional differences amongthem. Chemotaxis assays comparing iron-depletedmonocytes to total monocytes demonstrated thatMI + M2 monocytes were less active than total mono-cytes. Similar experiments using thrombin-purifiedfractions confirmed these results and localized es-sentially all of the chemotactic activity in the M3 sub-population. In addition, preliminary experiments thatused an adaptation ofthe agarose plate method in whichmigrating monocytes could be harvested and analyzedvolumetrically indicated that the chemotactically ac-tive cells among whole MNLwere >90% M3 mono-cytes (data not shown). These experiments providedclear evidence that volumetrically-defined monocytesubpopulations have markedly different functional ca-pabilities. Although the effect of cell size on chemotaxisis unknown, it is unlikely that the 10-fold differencebetween MI + M2 and M3 monocytes can be attributedto size difference because there is less than a three-fold difference in cell volume and less than a twofolddifference in surface area.

Others have recently observed volumetric subpop-ulations of human monocytes and murine mononuclearphagocytes. Norris et al. (10) fractionated human MNLby elutriation and found a population of small peroxi-dase-positive and phagocytic cells in predominantlylymphocyte fractions. These cells were found to berelatively deficient in antibody-dependent cellularcytotoxicity compared to purified large monocytes.Breard et al. (11) identified two populations of humanmonocytes with a monocyte specific heteroantiserumthat differed in size. Lee and Berry (12) used velocitysedimentation to fractionate Cortnlebacteriunin parvuiinu-induced mouse peritoneal mononuclear phagocytesinto small, medium, and large populations. Smalland medium cells functioned as accessory cells duringin vitro generation of antibody to sheep erythrocytesbut were not cytotoxic for tumor cells. Conversely,the large cells were cytotoxic and either suppressedor were unable to serve as accessory cells in the anti-body response.

The demonstration of volumetrically distinct andfunctionally different populations of human monocytesis consistent with the hypothesis that the mononuclearphagocyte system is composed of identifiable subsetswith discrete functions. However, neither our datanor that of others elucidates whether these functionallydiverse subpopulations represent cells following inde-pendent pathways of differentiation analogous to T andB lymphocytes, or whether they represent transientbut functionally discrete steps in a continuous dif-

Monioci te Heterogenteity 617

ferentiation sequence. The ontogeny and additionalfunctional characteristics of monocyte subpopulationsmust now be determined.

In summary, we have demonstrated that cell volumedifferences provide a means of identifying and quanti-tating subpopulations of human blood monocytes thatare functionally distinct. Additional studies of thesesubpopulations may stubstantially enhance our under-standing of the mononuclear phagocyte system and itscomplex role in the host defenses of man.

ACKNOWLEDGMENTSWe thank Ms. Sylvia A. Rayner, Mr. Calvin Myers, Ms.Dorothy Haskett, and Ms. Sherrie Gard for excellent technicalassistance. Wealso thank Dr. Robert I. Lehrer for his criticalreview of this manuscript.

This work was supported in part by grants CA12800 andCA22794 awarded by the National Cancer Institute, Depart-ment of Health, Education, and Welfare and contract EY-76-S003-0034 awarded by the U. S. Department of Energyto the University of California.

REFERENCES

1. Walker, W. S. 1976. Functional heterogeneity of macro-phages. In Immunobiology of the Macrophage. D. S.Nelson, editor. Academic Press, Inc., NewYork. 91-110.

2. Boyum, A. 1968. Isolation of leukocytes from humanblood. Scand. J. Clin. Lab. Invest. 21(Suppl. 97): 51-76.

3. Tucker, S. B., R. V. Pierre, and R. E. Jordan. 1977. Rapididentification of monocytes in a mixed mononuclearcell preparation. J. Immunol. Methods. 14: 267-269.

4. Steincamp, J. A., M. J. Fulwyler, J. R. Coulter, R. D.Hiebert, J. L. Horney, and P. F. Mullaney. 1973. A newmultiparameter separator for microscopic particlesand biological cells. Rev. Sci. Instrum. 44: 1301-1310.

5. Rotman, B., and B. Papermaster. 1966. Membraneproperties of living mammalian cells as studied byenzymatic hydrolysis of fluorogenic esters. Proc. Natl.Acad. Sci. U. S. A. 55: 134-141.

6. Holm, G., D. Pettersson, H. Mellstedt, E. Hedfors, andB. Bloth. 1975. Lymphocyte subpopulations in peripheralblood of healthy persons. Clin. Exp. Inmtyunol. 20:443-457.

7. Terasaki, P. I., D. Bernoco, M. S. Park, G. Ozturk, andY. Iwaki. 1978. Microdroplet Testing for HLA-A-B-Cand D Antigens. Am. J. Clin. Pathol. 69: 103-120.

8. Klein, R. B., T. J. Fischer, S. E. Gard, M. Biberstein, K. C.Rich, and E. R. Stiehm. 1977. Decreased mononuclearand polymorphonuclear chemotaxis in human newborns.Pediatrics 60: 467-472.

9. Cox, J. M., and T. P. Stossel. 1976. Measurement ofphagocytosis by macrophages. In In Vitro Methods inCell-Mediated and Tumor Immunity. B. R. Bloom andJ. R. David, editors. Academic Press, Inc., New York.363-368.

10. Norris, D. A., R. M. Morris, R. J. Sanderson, and P. F.Kohler, 1979. Isolation of functional subsets of humanperipheral blood monocytes.J. Immunol. 123: 166-172.

11. Breard, J., H. Lazarus, and S. F. Schlossman. 1979. Gen-eration and specificity of an heteroantiserum to humanperipheral blood monocytes. Fed. Proc. 38: 1418. (Abstr.)

12. Lee, K. C., and D. Berry. 1977. Functional heterogeneityin macrophages activated by Corynebacterium parvum:characterization of subpopulations with different ac-tivities in promoting immune responses and suppressingtumor cell growth. J. Inmumnol. 118: 1530-1540.

618 E. B. Arenson, Jr., M. B. Epstein, and R. C. Seeger