Differentiation Macrophages Human Marrow in Liquid...

Differentiation of Macrophages from Normal HumanBone

Marrow in Liquid Culture

ELECTRONMICROSCOPYANDCYTOCHEMISTRY

DOROTHYF. BAINTON, Department of Pathology, University of California atSant Francisco, California 94143

DAVID W. GOLDE, Department of Medicine, University of California at LosAngeles School of Medicine, Los Angeles, California 90024

A B S T RA C T To study the various stages of humanmononuclear phagocyte maturation, we cultivatedbone marrow in an in vitro diffusion chamber withthe cells growing in suspension and upon a dialysismembrane. At 2, 7, and 14 days, the cultured cellswere examined by electron microscopy and cyto-chemical techni(lues for peroxidase and for morelimited analysis of acid phosphatase and arylsulfatase.Peroxidase was being synthesized in promonocytes of2- and 7-day cultures, as evidenced by reaction pro-duct in the rough-surfaced endoplasmic reticulum,Golgi complex, and storage granules. Peroxidase syn-thesis had ceased in monocytes and the enzyme ap-peared only in some granules. By 7 days, largemacrophages predominated, containing numerousperoxidase-positive storage granules, and heterophagyof dying cells was evident. By 14 days, the mostprevalent cell type was the large peroxidase-negativemacrophage. Thus, peroxidase is present in highconcentrations in immature cells but absent at laterstages, presumably a result of degranulation of peroxi-dase-positive storage granules. Clusters of peroxidase-negative macrophages with indistinct borders (epi-thelioid cells), as well as obvious multinucleatedgiant cells, were noted. Frequently, the interdigitatingplasma membranes of neighboring macrophagesshowed a modification resembling a septate junction-to our knowledge, representing the first documentationof this specialized cell contact between normal macro-phages. We suggest that such junctions may serve aszones of adhesion between epithelioid cells.

Address reprint requests to Dr. Bainton.Received for publication 31 October 1977 and in revised

form 10 February 1978.

INTRODUCTION

Recent studies have convincingly demonstrated thatmacrophages of inflammatory exudates originate fromprecursor cells in the bone marrow, are transportedas blood monocytes via the peripheral circulation,and differentiate further in tissues and cavities to be-come large macrophages (1-3), epithelioid cells, andmultinucleate giant cells (4-7). Seeking to identifythe precursor cells of mononuclear phagocytes inhuman bone marrow, Nichols et al. (8) and Nicholsand Bainton (9, 10) found that electron microscopyand peroxidase cytochemistry afford several advan-tages over light microscopy and routine electronmicroscopy in the definitive identification of thevarious developmental stages of immature leukocytes.In addition, the stages of monocyte maturation couldbe more explicitly recognized.

In normal human marrow, peroxidase is synthesizedearly, during the promonocyte stage, and is localizedin all granules and organelles of the secretory ap-paratus-i.e., cisternae of the rough endoplasmicreticulum (RER)l and Golgi complex. It has beendetermined that the peroxidase-positive granules alsocontain the enzymes arylsulfatase and acid phos-phatase and hence are modified primary lysosomes. Inthe monocyte stage, peroxidase production ceases andis no longer visible in the RER or Golgi complex,and a second population of granules is produced.Formation of these peroxidase-negative granules con-tinues while monocytes are being transported in theblood, but their content is unknown (9, 10).

I Abbreviation used in this paper: RER, rough endoplasmicreticulum.

J. Clin. Invest. C The American Society for Clinical Investigation, Inc., 0021-9738/78/0601-1555 $1.00 1555

Relatively little research has been dlevote(l to thecytochemiiical clcaracteristics of the huimiiani m11on1o-nuiiclear phagoeyte systemii in vivo afteir the l)looclmonocvte stage. However, the ecultivationi of normiialbone mnarrow cells in a li(qLli(l me(diumii in a Marlbrook invitro difutsioni chamber hlas recenitly leen achieved,and provi(les a convenienit meanis for stu(lving thedifferentiationi of this cell line in short-terim cutiltuires(11, 12). This system is advantageous l)ecause cellscan be eciltivated in the absence of exogenouis stimiula-tory substances anid are easily retrieved for analysisof' finetioni, cytogenetics, anidl cytochemi stry. Usinig thismetho(d of' clultture, Golde and Cline (11) were albleto idenitif'y the various maitiuratioiail stages ofi noriiialhumiainla milonloniutclear phagocytes by morphological anidcytochemilical procedures tusing light microscopy, anidalso denionistrated that miiacrophages miatuirinig invitro possess suirf'ace receptors for IgG anidl are capal)leof actively phagocytizinig microorganisms (11, 12).

The puIrp1ose of' the presenit studly was to characterizethese imimiatuire and milatuire hulitian milonioniulclearphagocytes by electron microscopy combined withperoxiclase and lysosomal einzymile localizationi. Thepreviousi investigationis by Nielols and Bainitoni (9, 10),using noriiail human bone marrow and 1)100(1, afforda basis for comparing in vivo cellular (liff'erenitiationwith that o)bserved in vitro after, 2, 7, or 14 aclysof' culture. In this report, we describe the fine struc-tiural and enzymatic alterations which ocecur (luiring thein vitro maturation of promoniocytes to large mnacro-phages andl giant cells.

METHODS

Materials. Bone marrow was obtained from tbe p)osterioriliac crests of five healthy, adlult volunteers.

The histochetnical reageints, Grade I 8-glycerophosphateand p-nitro-catechol sulfate, were obtained from Sigma Chemi-cal Co., St. Louiis, Mo., an(l 3,3'-diaminobenzidline tetra-hydrochloride vas stipplied by either Sigma Chemiiical Co.or ICN Nuitritional Biochemicals Div., Internationial C(hemical& Nuclear Corp., Cleveland, Ohio. Gluicose oxidase (A-grade)was purchase(d from Calbiochemil, San Diego, Calif.

Collec tion of tissues. Techini(qtues for obtaining stuspen-sionls of Ibone marrow cells and(l eiltuiring them in a li(quiidmedium have been describecl (1 1). Specimens were harvestedat 2, 7, andl 14 days.

Fixatiou). Cells processecl for enzyme localization wcerefixed in 1.5% (listilled gltutaraldlehydle in 0.1 M sodlitinucacodylate-HCI buffer (pH 7.4) containing 1% stucrose at4°C for 10-30 min.

Enzyme procedures. After fixation, cells were vashedthree times in sodiuim cacodylate-HCl buffer (pH 7.4) with7% sucrose. They wvere theni incubated at 22°C inl each ofthe followinig enizyme media containiing 5% suicrose: (a)Graham and Karnovsky's medium for peroxidase (13), pH 7.4,for 1 h (Figs. 1-3); (b) Tice's medium for peroxidase (14), usingthe glucose oxidase metho(d for the generation of H202, for2 h at 22°C (Figs. 4-8); (c) modified Gomori's imiedium foracid phosphatase (15), pH 5.0, for 90 min (Figs. 9 and 1OA);anid (d) Goldfischer's mediuimil for arylstilfatase (16), pH 5.5,

for 9() mi (Fig. lOB), followed by treatmlen(t with 2% (Nf14)2S(17).

Subsequent processing. Enzvymc' p)re'parations were( p)ost-fixed fOr 1 h at 40C( in 1%/(. OSo4 in acc'tate-\eronal (Win-throp Laboratories, Evanston, Ill.) bufter, pH 7.6, withl Orwithout staining en) bloc in uranvl acetate for 60 min, ad(then proccsse(l and( (cxaminedl as dlescril)d'l (8, 9).

RESULTS

The number and( (lifferential couinits of' norimial l)onemarrow cells grown in a liquid me(diu have Al-rea(dy b)een reporte(d (11). In brief,' at 2 clays (andup to 4 or .5 clays), the predlominiianlt cel linie is theneuitroplhilie granuhlovte; thereafter, the p)rop)ortion ofmaerophages progressively increases uintil 1b 7-14(lays, tlhexy are prepond(lerant. Tle.se macro)phages staii iintensely for a-naphthvl butrwate esterase, but peroxi-dcase activity hlias been nioted to (lecrcasc with miaitiura-tion (12).

Localization of l)eroxida.sc int 2-, 7-, anid 14-daiycunltu rcs

Promonociltes. The promonocytes (Fig. 1) caii beidlentifie(d without (liffictilty at 2 and( 7 clays, but wvererarie 1b 14 dclays (Table I). These imimiaiitiure cells ap-peare(l similar to thiose observed in freshly ol)btai nedniorimial hIiiiumani bone marrow (8-10), except thiait thepitsi)lia memiibranie finiger-like extenisionis were niot seenin vivo. Peroxidclase \as l)eing synthesized dutiring theiniitial stages of milatuiration, sinice reactivity wasvisible as a floccuilent density in alli cisternae of'the RER, all Golgi c'isterinae, and(i in both imimiaiitiureacI(i miaitiure cytoplasm ic grantiles (m-300 nini).

louocltecs. The milonobcyte differed firom the piro-monocyte in several aspects: (a) Thc' ntcleuis l)ec'amemore indented aInd often horseshoe shaped, and(I thecl'romilatinv was mo(lerately conclen se(l. (b) Peroxi(lasereAactioni productt disappeared f'romil the RERand Golgicisterinae, inidicatinig that synithesis of the enizymiie hadcease(d (Table II). (c) Two populations of' graniuileswere usuially presenit, one peroxiclase-positive anid theother peroxiclase-negative. These cutltuired monocytes(Fig. 2) differed f'rom typical 1lood10 monocvytes in thatthey contained f`ewer peroxiclase-negative graniuiles. Inaddition, secondary lvsosomes fille(d with peroxidlasereactioni producit were ob)served (arrov, Fig. 2), reflect-ing partial clegrantilation of the peroxidase-positivestorage granuiles. In other monotcytes (Fig. 3), we

commonly encouinteredl heteroplaigy of' cells such asneuitrophils anid eosinophils.

Macrophage.s iitl p)eroxidase-p)ositive gi'-a nules.By 7 (lays in cuiltiure, the milore typical feattiures ofmcaerophages emlergedl as follows: (a) there was amairke(d inerease in cellilar size which was mainilyattril)utal)le to the auigmienited cytoplasm; (b) an in-crease in the number of' militochond(lria, expanision of'

1556 D. F. Bainton and D. W. Golde

* .s44

a 4

t.

\ ;W11 es sa L AN

/ t 'A~~~~~~~~~~~~

^ i '" ' ,.," '' w

,.4~~~~~~~~4

(3 ^x_ _ b' *s ~ ~ ~ ~ ~ ~ ~ ~ ,I' *

CW

it~~~~~~~~si

-:eSC.4SX4

FIGuRE 1 Promonocyte from a 2-day culture reacted for peroxidase, which can be seen as afluocculent density in the RER(rer), perinuclear cisterna (pn), and all granules (p+g). The largenucleus (N) contains a large nucleolus (nu) and heterochromatin is sparse. A few mitochondria(m) are visible, and several cytoplasmic extensions (e) extrude from the cell surface. (x 14,000.)

the RER, an enlarged Golgi region with numerous into enormous cells with a host of storage granules.vesicles, and microfilaments (100 A) were also Some contained relatively few digestive vacuoles,apparent (Fig. 4); (c) characteristically, the plasma whereas, in others, these secondary lysosomes weremembrane had many narrow extensions. The only plentiful (Fig. 4, inset).peroxidase-positive organielles were the granules Macrophages lacking peroxidase -single macro-formed earlier, during the promonocyte stage, and phages. By 14 days, the dominant cell type was thesecondary lysosomes. Althouigh they were less promi- large macrophage without endogenous peroxidasenent at 14 days, macrophages with peroxidase-con- (Fig. 5 [Figs. 5-8 illustrate macrophages from 7- andtaining granules persisted and somiietimes developed 14-day cultures lacking peroxidase.]). Whereas some

Human Macrophage Differentiation In Vitro 1557

-A t.\*. . _'\.. .e

TABLE IStages of Monontclear Phagocqlte Differenitiation

in Liquid Cuiltture at Vhariou.s l'imeSs

Davs in cudltre

2 7 14

Promonocyte ++ + rareMonocyte ++ + rareMacrophage

(a) Peroxidase-positive rare ++ +(b) Peroxidase-negative rare + + +

Epithelioid cells 0 + + +Giant cells () 0 +

macrophages were free ini the liquiiid culture (Fig. 5),many formed cluimlps (Fig. lOB) or hald become ad-herent to the membrane (Fig. IOA).

Epithelioid cells. These tightly grouiped cells haddevelopecl elaborate cytoplasmic projections (Fig. 6A),similar to those described in immatuire epithelioidcells (7, 18). Occasionallv, the interdigitating pro-cesses of two adjacent macroplhages had made intimatecontact, an(l a modificationi of the two p)lasmna menm-

branes and cell sturfaces was evident (Figs. 6B and 7).This contact resembled the septate-like zone of ad-hesion described in other cell types (19), but it hasnot been observed before between normal macro-phages. In addition to cluisters of lipid-filled macro-phages with indistinct cell membranes (Fig. 8), weencouniitered conspictuouis multinuicleated giant cells(not illulstrate(l).

Localization of lylsosomal enzymes in 7- and14-dayj culttresMarrow f'rom two sulbjects was cultured for 7 and

14 days, and the cells in li(quiid suspension wereaspirated, fixed, and processed separately from thecells either adherenit to or loosely settled upon thesuirface of the membrane. Acid phosphatase distribu-tion followed two different patterns: in both suspendedand adherent cells, we located -25% that containedreaction produict within the entire RERand in diges-tive vacuioles (Fig. 9 [Figs. 9 and 10 depict 7- and 14-(lay ctulttured bone marrow cells which have been re-acted for lysosomal enzymes.]). These cells were notordinarilv in close association with other macrophages.

TABLE IIUltrastr,uctural Characteristics of Various Stages of Motlononelear Phagocyte Differentiation in Cultuire

Peroxidase localizationNuiclear/

Golgi Secondarv cytoplasmicCell type Cell size RER complex Grantiles lvsosomes ratio Ntcletis

Promonocyte (Fig. 1) 10-15 + + + - <1 (a) Round; indentations(b) Minimally condensed

chromatin; large nucleoli

Monocyte (Figs. 2 and 3) 8-10 - _ + + I to >1 (a) Variable shape(b) Heterochromatic, occasional

nucleoli

Macrophage(a) Peroxidase-positive 10-20 - - + + >1 (a) Usually eccentric; oval or

(Fig. 4) indented(b) Chromatin variable; may be

very condensed oreuichromatic

(b) Peroxidase-negative 10-20 - - - - >1 (a) Eccentric(Fig. 5) (b) Euichromatic, with large

nuicleoli

Epithelioid cell(a) Immature (Fig. 6) 15-20 - - - - >1 (a) Eccentric; indented(b) Matuire (Fig. 8) 15-20 - >1 (b) Euichromatic, with nucleoli

Giant cell >25 - - - - >1 (a) Mtultiple; randomly dispersed(b) Euichromatic, with large

nucleoli

1558 D. F. Bainton and D. Ws. Golde

b- .

T; 4,5zsk

:4:'

~ ~ ~ 'I.4

* ~ ~ ._G.<_St

'- ~ ~I1;)X-* \ 3

~) * \ve.r¢t.p_-ve

p-n

';.t6*E-24 -1-iA 4*~0atP +4 r

;loz-zz-'<'Nt 'es*-t ..'^i X i' tt,. 't,'g>~~~~~sx

wM h£s / *Xh ; 3XFb~~~~~~~~~~~~~~~~~~~~~~~~~~O

~~~.jk

Sr

..dl

4.I.

s

"',v

FIGURE 2 Monocyte from a 7-day cuiltuire. Peroxidase is present only in cytoplasmic granules(p+g) and in one large vacuole (arrow)-presumably a secondary lysosome-in which degranula-tion has occurred. The many other vacuoles (v) and vesicles (ve) contain no peroxidase. Twoperoxidase-negative granules can also he discerned (p-g). In the upper left-hand corner, note theportion of a promonocyte (P) with peroxidase in the RER(rer) and granules (p+g). (x 15,000.)

Human Macrophage Differentiation In Vitro

tbX^::v :;4 -X e

jjhjLj.

WIN T

!Wr-\% < x,li.rer.P8'w't._;s't.~~~~~~T

/'' 't s~~A

t*j<*tsS

..i. 141,

thm.

-,%W-

1559

4#

44~ ~

.:?.

.S.

FIGURE 3 Monocyte (or early macrophage) from a 7-day culture, demiionistrationi heterophagy ofother leukocytes. Note the many digestive vacuoles (dv) or secondary lysosomes filled withcellular debris. Obviouis eosinoplhil crystals (arrows) caII be idleitifie(l wvithin several of thesevacuoles. Some smaller peroxidase-positive granules (p*g) are still present. Cvtoplasmnic ex-

tensions (e) on this cell outnuimhber those of the more typical moinocyte (lep)icted in Fig. 2. (x 13,000.)

In the seconid pattern of distribution, weak reactioniprodctet wvas onily seen in some Golgi cisternae, anid

mainly in large digestive vacuioles. As is apparenit inFig. IOA, some of' these mlacrophages adhered to thedialysis membrane in comiipact nmasses. Cells found ineluLmps in the li(quid suspensioni medium also fol-lowevd this latter distribution. Reaction product for aryl-sulf;atase wlas less extensive than that for aci(d phospha-tase in that it was rarely seenin the RER or Golgicomlplex. Rather, it was observedi most ofteni in simiallvesicles anlle large seconidary lysosoimies (Fig. 1(B).

DISCUSSION

Monlocytes, uniilike nietitrophilic leukocytes, are niotfUlly, difterenitiated( when releasedcl f'romi the bone miiar-

ro\w. After cireculatinig in the bloo1 ( they eniter thetissuies, where they imiatture ilntO large, lonig-lived

macrophalges (1-3). The imiatturation of' rodent miiacro-

phages lias been extensively investigated both in vitroand in vivo (22-25). Cohni anid his co-workers (26-28)have shown that smiiall, i in miatture imiacrophages developinlto verv large phagocvtes with sizable nIticlei, plenlti-

1560 D. F. Bainton and D. W'. Golde

dv

AM. af .

,. ;',. .. ..1

fllfl

A

N

r-PI.

'4'.. ei 4t, i

.:.%....0.,. ,.-z;:49,4 :.:.

.,.j 'f.

t ':

-s'..w}

:~~~~~~~~~~~~~~~~~~~~~~~~~:#1 .. 47 <IA. >~~~~~~~~~~~~~~~~~~t -~~A

1I,

4/~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4

FIGURE 4 Large macrophage from a 14-day culture containing many peroxidase-positivegrannies (p+g). m, mitochondrion; mf, microfilament; rer, scattered RER. A large peroxidase-positive inclusion is designated by an arrow. Note the slender cytoplasmic extensions (6) of anadjacent macrophage which partially interdigitate with the extensions (e) of the other cell. Theinset shows two secondary lysosomes (arrows) from another macrophage, which contains whorlsof membrane (w), two eosinophil granules (E), and other debris. A lipid droplet is also present.( x 12,OOO; inset x14,OOO.)

Human Macronhauc Differentiation In Vitro 1561-- -- "V..-n- - --I./ - -.- --. -

p-g '(.7~~~~~~~~ ~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

XG !~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0

rer:

FIGURE 5 Peroxidase-negative iniacrophage f'rom at 14-clay cuiltuire. Althiough this type ofnmacro-phage is also observed at 7 days, it is the predominant type in thie 14-day cuilture. It hasan eccentric nutcleuts (N) with a distincet nuicleoluis (nu1) and( cytoplasm filledi with many of theorganelles described b)eiore-i.e. miitochiondria (nin); RER(rer); a large (;olgi comiplex (C") is niearthe plasma miemi-branie; andi ]Lnumerouis smiall vesicles (=-600) A). Note that f(o p)eroxidase reactionlproduct cani he detected at this late stage of matuiration. In adldition to occasional cle-ar xacuoles

(v, any incluisions wvith miodlerately dlense miatric-es are evident. Their content is uinknowni;we are presenitly designatintg thiemi as peroxidase-ntegative granutles (p-g). Also note the p)aucitNof, 1,000 A vlesicles, wvhich are uisually presenit in great num-bers in p)eritoneal mnacrophage'sfixed in vivo (8). (x 19,000.)

1562 D. F. Bainton and D. Wt'. Gwoldc

1-~~~~~~~~-

m~̂ >l fo-2 -Jyk\vX s wFe

M N

- ft~~~~~~~~~~~~~~~~~~~~~~4

PQM2 N c~~~~~~~~~~~~tQZQ ~~N'5;. A%N4

@;'' .'~'A"^,.'-rAtvN,~~~~~~~~~~~~~~~~~~~~~~~~>

4, -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

4N- (4' t9t''V3j,'t' ;'9' '; t\', * t < i' ' 'J'jQi-' %' *'v<'; -^'t' ' 5 *Y '/IGURE 6Aan)Hg agiiatosofajcetmarpagsreualyeiheiicl

'if4~~~~~~~~~~~~culturedJfor14days and reacted for peroxidase,illuttrating the numerous finger-like extensions

and ¾fcellrKontacts.(A) YkpictstYrtions4¾fytoplasmfrmthr

(m)and]00A mirflmns(f aepeet ()x1,0) B 2,0.

MI> .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~v

-4-'~~~~~~~

Nk >-iN~~~~tX~~

t M

FIGURE 6 (A and B). High magnifications of adjiacent macrophages, presumably epithelioid cells,cultuired for 14 days and reacted for peroxidasel, illustrating the numerous finger-like extensionsand cell contacts. (A) Depicts portions of cytoplasm fromn three macrophages (MI, M2, and M3);and (B) cytoplasm from two macrophages (MI and M2). At four sites (arrows), the plasma mem-branes of the two cells in (B) are in close conitact and mianifest a structuiral modification whichis better seen in Fig. 7-a higher magniification of two adjacent miembranes. These macrophagesrepresent early or immnature epithelioid cells, containing many lipid droplets (L) which non-specifically bind diaminobenzidine, thus accounting for their dark rims. Multiple mitochondria(nin) and 100 A microfilaments (mf) are present. ((A) x21,000; (B) x28,500.)

.-A. i"

,,;. ..... jp'. :... . W .1 , .......V i:.... -, , I. ...

FIGURE 7 A septate-like zone of adhesion with characteristicspacing between two macrophages. The space between thetwo plasma membranes (arrows) measures more than 200 Aand contains irregularly spaced extracellular particles (p). Theintracellular face exhibits fine filaments (f). (x 100,000.)

ful cytoplasm, numerous Golgi complexes, lysosomes,stacks of smooth and rough endoplasmic reticulum,and many mitochondria. These structural evolutionarychanges are associated with increases in the cellularcontent of protein, RNA, oxidative enzymes, andlysosomal hydrolases. More limited electron micro-scopic and cyto-chemical studies have been conductedwith human mononuclear phagocytes obtained frombone marrow (29, 30) and blood (31, 32), with softagar (29, 31, 32) or methylcellulose (30).

In this study, we used a liquid culture system andelectron microscopy and cytochemistry to facilitatethe detailed characterization of cellular organelles andenzyme contents of human bone marrow-derivedmononuclear phagocytes. The peroxidase technique isuseful for identifying promonocytes, because they arereactive for this enzyme and thus, can be distinguishedfrom peroxidase-negative cells such as erythroblasts,

lymphoblasts, and plasma cells. Although the granulo-cytic promyelocytes (neutrophilic, eosinophilic, andbasophilic) are also peroxidase positive, their storagegranules differ sufficiently in fine structure to makethem distinguishable from promonocytes (9, 33, 34).Hence, with van Furth and Fedorko (22), we supportthe conclusion that mononuclear phagoeytes do notdevelop from immature granulocytes but represent aseparate line of differentiation. In this in vitro system,we demonstrated that peroxidase is synthesizedearly, in the promonocyte stage, and packaged intostorage granules by the same pathway (RER -* Golgicomplex via vesicles-- granules) which has beendefined for other secretory proteins (35). Monocytesdid not synthesize the enzyme but contained it in theirstorage granules.2 With the endocytosis of dyingcells, these granules degranulated; and, within 7 to14 days, the cells became peroxidase-negativemacrophages.

Phagocytosis seems to be a major cause of the lossof storage granules in the liquid culture system. How-ever, some cells retain such peroxidase-positivegranules in great quantities, even after 14 days of invitro culture. Nevertheless, the vast number of macro-phages are peroxidase-negative by 14 days, and epithe-lioid and giant cells, also all negative for peroxidase,are frequently observed. The lipid-filled clusters ofepithelioid cells seen in Fig. 8 may correspondto the "giant fat" cells recently described by Dexter etal. (37) in the adherent cells cultivated from mousebone marrow.

Macrophages tended to form clusters, with extensiveinterdigitation between cells. Such groups have pre-viously been noted in animals (5, 7, 18, 38) and man(39-41) and referred to as "immature epithelioid cells"(7, 18). We found in this study that some of the cyto-plasmic extensions had modified membrane contactsresembling septate junctions. This membrane modifica-tion of the mononuclear phagocyte was similar to thatdescribed earlier by Friend and Gilula (19) in the ratadrenal cortex and in other steroid hormone-secretingcells. Along this distinctive cell contact (19), the mem-branes of apposing cells are separated by 210-300 Aand bisected by irregularly spaced 100-150-A extra-cellular particles. In freeze-fracture replicas, the cellmembranes in the area of the septate contact is no dif-ferent from non-junctional areas of membrane (19). Some-what similar membrane modifications have been de-scribed by Sanel and Serpick (42) in leukemic monocytes,by Daniel and Flandrin (43) in hairy cells undergoing

2 During studies on blood monocytes in vitro, Bodel et al.(36) observed that peroxidase activity rapidly appeared in theRERand perinuclear cisterna within 2 h after monocyte ad-herence to a fibrin-coated surface but disappeared after thecells had been cultured for 1-2 days.


.I. -.

..N.s 4 .~ i

N57' '7

01.Of~~~ ~~~~~~~~., < ; Ns ) B;R t\A66 , S ' .-.1

no ="4i

41V; X St*Wg i-;sssx -4t4

?-'4s'N4vC.( tt,<stett<,4,,t4

VC i

- " -- o'. < -. \ e' 2

FIGURE 8 A group of large peroxidase-negative macrophages from a 14-day culture, with a vievof the nuclei of five cells (N1 -N5). It is difficult to distinguish the cell borders (arrows)-a situationwhich has been noted previouisly in regard to mature "epithelioid" macrophages (18). Numerou1slipid droplets (L), some with very dense borders (L'), as well as mitochondria, many vesicles,clefts (C) (20), and several Golgi conmplexes (G) are present. Here, the inset reveals intracelluilarmembranes, including a Golgi region (G), at a higher inagnification. The vorm-like struetures(wo) have been demonstrated in guinea pig macrophages (21) and are thought to represent plasmamembranie modifications. (x 10,000; inset x 11,000.)

HImanI Macroplizage Differentiation In Vitro 1565

I%

t.

.c

w.,t,

VP~~~~~~~~6

Is ~ ~ A .. t. ...

;' * jt@rntst

/ ,, .~̂*e

it ,

t w t-. -s ~~~~I, ^ . t

F "~~~~~~re

4,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'0

A.

#h ~~~- ,di 4 -

Ap:r .:':.V P.;k 4 *" w

s t^. Ass;' . f t .. ''~T

FIGURE 9 A macrophage growing in suspension in a 7-day culture anld reacted f{r acid Phos-phatase. The entire RER(rer), excluding the perinuclear cisterna (pn), conitains moderate amountsof lead phosphate. In addition, this lysosomal enizymiie appears in imost Golgi cisterniae (Ge)as well as in some digestive vacuoles (dv). (x23,000.)

phagocytosis, anid by Bretoni-Gorius et al. (44) in ab-niormal erythroblasts. To our knowledge, this is thefirst report of septate-like zones of attachmiient in normnalmacrophages, and we believe that they may serve asadhesive elements in these cells. "Close junctions"hiave been observed in cultured imiurine imiacrophages(45) anid as so-called "junctional formiiations in im(ouse-spleeni cultures (46), but the imorphology of these imiemil-brane contacts is (1uite dissimiiilar from that illustrated

here. Moreover, this iimemiibranie miiodlification differredconsiderably from the cell coats of adjacent imiatrophagesdepicted by Brederoo and Daems (21), who (describeda spaciing of'630 A, muivch thicker thani ouir 300 A contact.

Another findinig of this research wats the observationthat adherent mlacrophages can pierce the dialysis miemii-brane, as demllonlstrated in F'ig. IOA. This imiay be an im-portant discovery, because som11e investigators arc de-signinlg experimenits anid interpretinig datat based oni the


.I

41 .;.,

.f'..f*. /.1,-4.

lw~~~mJ~

p.. ". )B:?4L'

-.4"R~~~~~~~~~~~~~~~~~~I

'', 4X>

>G.,INfi, CkAsi

FIGURE 10 (A) Adherent macrophages (M'-M2) growing on the membrane in a 7-day cultureand reacted for acid phosphatase. Only scant amounts of lead phosphate are demonstrable insome Golgi cisternae (Gc) and a few digestive vacuoles (dv). Note the close cell-to-cell contact.One process has actually pierced the dialysis membrane (DM) (right corner). N, nucleus; rer, RER;m, mitochondrion. B: Three closely adjoining macrophages (M'-M3) from the liquid mediumof a 14-day culture which were reacted for another lysosomal enzyme-arylsulfatase. Only thecytoplasmic vacuoles (arrows) contain reaction product. Note the many plasma membrane ex-tensions (e); such cells have been called immature epithelioid cells because of their densepacking and numerous extensions. N, nucleus. ([A] x13,000; [B] x9,500.)

NN;

assumption that only small dialyzable molecules canpenetrate such membranes.

A further pertinent point is the unresolved questionof why some macrophages, usually single cells, areactively synthesizing acid phosphatase, whereas macro-phages in clusters seem to contain low levels of theenzyme. Perhaps the individual cells are better phago-cytes, and, having ingested a "meal," are induced tosynthesize lysosomal enzymes, as documented by Axlineand Cohn (47). Alternatively, the phagocytic capacityof cellular aggregates may be less than that of relativelyfree macrophages. In this connection, Spector andMariano (48) found that epithelioid cells phagocytizefewer latex beads and bacteria than do free macrophages.

In conclusion, it should be re-emphasized that theseobservations on the localization of peroxidase were de-rived from macrophages in culture. As far as we know,similar electron microscopic and cytochemical studiesof human macrophages in various tissues have not beenattempted. In animal experiments, certain tissue macro-phages have been found to contain peroxidase, cyto-chemically identifiable in the RERand perinuclear cis-tema. This enzyme reactivity in the RERis characteristicof resident peritoneal macrophages of the rat (49), guineapig (50, 51), rabbit (36), and mouse (52), as well asKupffer cells (53) and medullary lymph node macro-phages (54) of the rat. Such reactivity has not yet beenreported in the tissue macrophages of man (55), butthis point should be fuirther explored.

ACKNOWLEDGMENTS

We thank Ivy Hsieh and Yvonne Jac(ques for their excellenttechnical assistance and Rosamond Michael for her helpfulediting.

This study was supported by U. S. Public Health Servicegrants CA 14264 (Dr. Bainton) and CA 15688 and HL 20675(Dr. Golde).

REFERENCES

1. Volkman, A. 1976. Monocyte kinetics and their changesin infection. In Immunobiology of the Macrophage. D. S.Nelson, editor. Academic Press, Inc., NewYork. 291-322.

2. Cohn, Z. A. 1975. Recent studies on the physiology ofcultivated macrophages. In The Phagoeytic Cell in HostResistance. J. A. Bellanti, and D. H. Dayton, editors. RavenPress, New York. 15-24.

3. Van Furth, R., H. L. Langevoort, and A. Schaberg. 1975.Mononuclear phagocytes in human pathology-Proposalfor an approach to improved classification. In Mononu-clear Phagocytes in Immunity, Infection and Pathology.R. van Furth, editor. Blackwell Scientific PublicationsLtd., Oxford. 1-15.

4. Sutton, J. S., and L. Weiss. 1966. Transformation of mono-cytes in tissue culture into macrophages, epithelioid cells,and multinucleated giant cells. J. Cell Biol. 28: 303-332.

5. Papadimitriou, J. M., and W. G. Spector. 1971. The origin,properties and fate of epithelioid cells.J. Patiol. (Lond.).105: 187-203.

6. Mariano, M., and WN. G. Spector. 1974. The formation and

properties of macrophage polykaryons (inflammatory giantcells). J. Pathol. (Louid.). 113: 1 -19.

7. Adams, D. 0. 1974. The structure of mononuclear phago-cytes differentiating in vivo. I. Seqiuential fine and his-tologic studies of' the effect of' bacilluts Calmette-Guerin(BCG). Am. j. Pathol. 76: 17-40.

8. Nichols, B. A., 1). F. Bainton, and M. G. Farquhar. 1971.Differentiation of monocytes. Origin, nature, and fate of'their azurophil granules. J. Cell. Biol. 50: 498-515.

9. Nichols, B. A., (and D. F. Bainton. 1973. Differentiation ofhumllan moniocytes in bone miarrow and blood. Sequentialformation of'tw\o grainutle poptulations. Lab. lItvest. 29: 27-40.

10. Nichols, B. A., an(d 1. F. Baintoni. 1975. Ultrastructure andcytocheinistrv of monionuiclear phagocytes. In Mononu-clear Phagocytes in Immunnity, Inf'ection and Pathology.R. van Furth, editor. Blackwell Scientific PublicationsLtd., Oxford. 17-55.

11. Golde, D. WV., anid M. J. C'line. 1973. Growth of humlanbone miarrowv in liquiiid cutltuire. Blood. 41: 45-57.

12. Willcox, NM. B., D. XW. Golde, and NI. J. Cline. 1976. Cyto-chemical reactions ofhutian hematopoietic cells in li(quidculture. J. Ilistochieit. Cytochoem. 24: 979-983.

13. Graham, R. C., and NI. J. Karnovsky. 1966. The early stagesof' absorption of injected horseradish peroxidase in theproximal tubules of nmouse kidtney: Ultrastructural cyto-chemistry by a new technique.j. Histochem. Cy,tochem.14: 291-302.

14. Tice, L. W. 1974. Ef'fects of hypophysectomy and TSHreplacemnent on the ultrastructural localization of thyro-peroxidase. F ndocrinology. 95: 421 -433.

15. Barka, T., and P. J. Andersoni. 1962. Hlistochemical meth-ods for acid phosphatase using hexazonium pararosanilinas coupler. J. Ifistocheml. Cytochem. 10: 741-753.

16. Goldfischel-, S. 1965. The cytochemical demonstration oflysosomal aryl sulftatase activity by light and electronmicroscopy. j. Ilistoclcm. Cytochemn. 13: 520-523.

17. Holtzman, E., and( R. Dominitz. 1968. Cytochemical studiesof lysosonmes, Golgi apparatus and endoplasmic reticulumin secretion and proteini uptake by adrenal medulla cellsof the rat. j. Histochlenm. Ci tochem. 16: 320-336.

18. Adams, D. 0. 1976. The granulomatous inflammatory re-sponise. A review%. Amn. J. Patwol. 84: 164-191.

19. Friend, D. S., and( N. B. Gilula. 1972. A distinctive cellcontact in the rat adrenal cortex.J. Cell Biol. 53: 148-163.

20. Sanel, F. T., J. Aisnier, C. J. Tillman, C. A. Schiffer, andP. H. Wiernik. 1975. Evaluation of granulocytes harvestedby filtration leucapheresis: Functional, histochemical andultrastructural studies. In Leucocytes: Separation, Collec-tionl, andl Transf'usioni (Proc. Int. Symposium on LeucocyteSeparation, Royal Postgrad. Med. Sch., London, 9-11 Sept. 1974). J. NI. GColdman and R. M. Lowenthal,editors. Academnic Press, Inc. Ltd., London. 236-250.

21. Brederoo, P., anld XX'. Tb. Daems. 1972. Cell coat, worm-like structures, and labyrintlhs in guinea pig resident andexudate peritonical in-acrophages, as demonstrated by anabbreviated fixatior procedure for electron microscopy.Z. Zellforsch. Mikrosk. Aniat. 126: 135-156.

22. vanl Furth, R., and NM. E. Fedorko. 1976. Ultrastructure of-ouise mononoiuclear phagocytes in bone marrow colonies

growvn in vitro. Lab. Intcest. 34: 440-450.23. Goud, Th. J. L. NI.. C. Schotte, and R. van Furth. 1975.

Identificatioin of'the nionol)last in monionuclear phagocytecolonies grown in vitro. Inl Mononuclear Phagocytes inImmunt111ity, Infection ansd Pathology. R. van Furth, editor.Blackwell Scientific Publications Ltd., Oxford. 189-203.

24. Ben-Ishav. Z. 1975. tUltr.astrooctuiral stuidy of bone marrow-


derived granulocytic colonies in diffusioni chambers inrats. Scand.J. Ilactinatol. 15: 241-250.

25. Alleni, T. D., anid T. M. Dexter. 1976. Surf'ace morphologyanid ultrastructure of mulrine graniulocytes and monocytesin long-termii li(quid cultuire. Blood Cells. 2: 591-606.

26. Cohn, Z. A., and B. Benson. 1965. The differentiationof mononuclear phagocy tes. Morphology, cytochemistry,and biochemistry. J. Exp. Med. 121: 153-169.

27. Steinman, R. NI., and Z. A. Cohn. 1974. The metabolismand physiology of the mononuclear phagocytes. In TheInflammatorv Process. B. XV. Zweifach, L. Grant, and R. T.McCluskey, editors. Academic Press, Inc., NewYork. 2ndedition. 1: 449-510.

28. Gordon, S., and Z. A. Cohn. 1973. The macrophage. Int.Rev. Ci1tol. 36: 171-214.

29. Shoham, D., E. B. David, and L. A. Rozenszajn. 1974.Cytochemical and morphologic identification of macro-phages and eosinophlils in tissue cultuires of normal humanbone marrow. Blood. 44: 221-233.

30. Parmley, R. T., NI. Ogawa, S S. Spicer, and N. J. Wright.1976. Ultrastrtucture and cytochemistry of bone marrowgraniulocytes in culture. Exl). Hematol. (Oak Ridge). 4:75-89.

31. Zucker-Franklin, D., and G. Gruskv. 1974. Ultrastructuralanalysis of hemiatopoietic colonies derived from humanperipheral blood. A newly developed method.J. Cell Biol.63: 855-863.

32. Zucker-Franklin, D., G. Grusky, and P. L'Esperance.1974. Granulocyte colonies derived from lymphocyte frac-tions of normal human peripheral blood. Proc. Natl.Acad. Sci. U. S. A. 71: 2711-2714.

33. Bainton, D. F., J. L. Ullyot, and M. G. Far(quhar. 1971.The development of neutrophilic polymorphonuclear leu-kocytes in human bone marrow. Origin and content ofazurophil and specific granules. J. Exp. Med. 134: 907-934.

34. Bainton, D. F. 1977. Differentiation of human neutrophilicgranulocytes: normal and abnormiial. In The Granulocyte:Function and Clinical Utilization. T. J. Greenwalt andG. A. Jamiiiesoni, editors. Alan R. Liss, Inc., NewYork. 1-27.

35. Palade, G. E. 1975. Intracellular aspects of the processof protein synthesis. Science (W17asht. D. C.). 189: 347-358.

36. Bodel. P. T., B. A. Nichols, and D. F. Bainton. 1977. Ap-pearance of peroxidase reactivity within the rough endo-plasmic reticulumii of blood monocytes after surface ad-herence.J. Ex). Med. 145: 264-274.

37. Dexter, T. M., T. D. Allen, and L. G. Lajtha. 1977. Condi-tions controlling the proliferation of haemopoietic stemcells in vitro.J. Cell. Physiol. 91: 335-344.

38. Smyth, A. C., and L. W'eiss. 1970. Electron microscopicstudy of inhibition of macrophage migration in delayedhypersensitivity.J. Immunol. 105: 1360-1374.

39. Epsteini, XV. L. 1967. Granulomatous hypersensitivity. Prog.Allergy. 11: 36-88.

40. Elias. P. M., and W. L. Epstein. 1968. Ultrastructural ob-servations on experimentally induced foreign-body andorgaanized epithelioid-cell granulomas in man. Am. J. Pat hol.52: 1207-1223.

41. XVilliams. XV. J., D. A. Erasmus, E. M. V. James, and T.Davies. 1970. The finie structure of sarcoid and tubercu-lous graniulomiias. Postgrad. Med. J. 46: 496-500.

42. Sanel, F. T., and A. A. Serpick. 1970. Plasmalemmal andsubsurface complexes in human leukemic cells: Mem-brane bonding by zipper junctions. Science (Wash. D. C.).168: 1458-1460.

43. Daniel, M. Th., and G. Flandrin. 1974. Fine structure ofabnormal cells in hairy cell (tricholeukocytic) leukemia,with special reference to their in vitro phagocytic capacity.Lab. Invest. 30; 1-8.

44. Breton-Gorius, J., G. Flandrin, M. T. Daniel, J. Chevalier,M. Lebeau, and F. T. Sanel. 1975. Septate-like junctionsin abnormal erythroblasts. Cytochemical, ultrastructuraland freeze-etch studies. Virchows Arch. B Cell Pathol.18: 165-180.

45. Levy, J. A., R. M. Weiss, E. R. Dirksen, and M. R. Rosen.1976. Possible communication between murine macro-phages oriented in linear chains in tissue culture. Exp.Cell Res. 103: 375-385.

46. McIntyre, J. A., M. F. La Via, T. F. K. Prater, and G. D.Niblack. 1973. Studies of the immune response in vitro.I. Ultrastructural examiinationi of cell types and clusterformation and functional evaluation of clusters. Lab.Invest. 29: 703-713.

47. Axline, S. G., and Z. A. Cohn. 1970. In vitro inductionof lysosomal enzymes by phagocytosis.J. Exp. Med. 131:1239-1260.

48. Spector, WV. G., and M. Mariano. 1975. Macrophage be-haviour in experimental granulomas. In MononuclearPhagocytes in Immunity, Infection and Pathology. R. vanFurth, editor. Blackwell Scientific Publications Ltd., Ox-ford. 927-942.

49. Robbins, D., H. D. Fahimi, and R. S. Cotran. 1971. Finestructural cytochemical localization of peroxidase activityin rat peritoneal cells: Mononuclear cells, eosinophils andmast cells.J. Histochem. Cytochem. 19: 571-575.

50. Cotran. R. S., and M. Litt. 1970. Ultrastructural localiza-tion of horseradish peroxidase and endogenous peroxidaseactivity in guinea pig peritoneal macrophages. J.Imnmuntol. 105: 1536-1546.

51. Daems, W. Th., and P. Brederoo. 1973. Electron micro-scopical studies on the structure, phagocytic properties,and peroxidatic activity of resident and exudate peritonealmacrophages in the guinlea pig. Z. Zellforsch. Mikrosk.Anat. 144: 247-297.

52. Lepper, A. W. D., and P. D'Arcy Hart. 1976. Peroxidasestaining in elicited and nonelicited mononuclear peritonealcells from BCG-sensitized and nonsensitized mice. Infect.Immun. 14: 522-526.

53. Fahimi, H. D. 1970. The fine structural localization ofendogenous and exogenous peroxidase activity in Kupffercells of rat liver.J. Cell Biol. 47: 247-262.

54. Hoefsmit, E. Ch. M. 1975. Mononuclear phagocytes, retic-ulum cells, and dendritic cells in lymphoid tissues. InMononuclear Phagocytes in Immunity, Infection and Pa-thology. R. van Furth, editor. Blackwell Scientific Publica-tions Ltd., Oxford. 129-146.

55. Watanabe, K., S. Masubuchi, and K. Kageyama. 1973.Ultrastructural localization of peroxidase in mononuclearphagocytes from the peritoneal cavity, blood, bone marrowand omentum, with a special reference to the origin ofperitoneal macrophages. Recent Adv. RES(Reticuloendo-thel. Syst.) Res. 11: 120-147.

Human Macrophage Differentiation In Vitro 1569

Differentiation Macrophages Human Marrow in Liquid...

Documents

Transcript of Differentiation Macrophages Human Marrow in Liquid...