Murine leukocytes with ring-shaped nuclei include granulocytes, monocytes and their precursors

Murine leukocytes with ring-shaped nuclei include granulocytes,monocytes, and their precursors

Heike Biermann,* Barbara Pietz,* Rita Dreier,† Kurt W. Schmid,†

Clemens Sorg,* and Cord Sunderkotter*‡

*Institute of Experimental Dermatology, †Institute of Pathology,and ‡Department of Dermatology, University of Munster, Germany

Abstract: Leukocytes with ring-shaped nuclei (ringcells) are present in bone marrow (BM; ,50% ofBM cells), in peripheral blood (PB), and in inflam-matory infiltrates of mice, but also in humansduring myeloproliferative disorders. They are usu-ally referred to as polymorphonuclear cells (PMN),but we demonstrate that they additionally encom-pass different types of mononuclear (MNC)-likering cells. PMN ring cells had constricted ring-shaped nuclei with a wide cytoplasmic center andwere sorted among the GR-1high fraction. TheMNC-like ring cells belonged to the GR-1low frac-tion. Their nuclei were not segmented and thecytoplasmic center of their nuclei was small. Theywere heterogeneous with one subgroup containingmonocytes/macrophages according to ultrastruc-ture, immunophenotype (BM8, F4/80, CD13, ER-HR3), activity of unspecific esterase, and phago-cytosis of Leishmania major. A second subgroupcontained myeloic precursor cells as they prolifer-ated (Ki67), expressed ER-MP12, and showed onultrastructure distribution patterns of peroxidaseactivity compatible with myelocytes, promyelo-cytes, or promonocytes. A third subgroup of cellshad large, sometimes lobulated nuclei, waslineage markernegative/low (GR-1, Mac-1, B220 etc.),CD382, but c-kit1 and sca-11, and thus belonged toa close progeny of murine hematopoietic stemcells. In PB, ring cells encompassed mainly PMN,but also monocytes and cells with characteristics ofboth the granulocytic and monocytic lineage. Thus,ring cells comprise mature and precursor forms ofmyeloic cells. Their analysis revealed that in mice aclear distinction between the granulocytic andmonocytic lineage beyond the GM-CFU stage is notalways feasible. J. Leukoc. Biol. 65: 217–231;1999.

Key Words: ring cells · myeloic cells · precursor cells · macrophage

INTRODUCTION

A conspicuous finding in murine bone marrow (BM) areleukocytes with ring-shaped nuclei (ring cells). Their percent-age is marked but, in contrast to other well-defined BM cells,

they have not been characterized thoroughly and often are noteven mentioned in studies on murine blood cells. When theyare addressed they are usually referred to as polymorpho-nuclear neutrophils (PMN) or eosinophils and sometimes theyare also considered to be myelocytes [1–8]. In a few studies ithas been suggested that cells with a ring-shaped nucleus couldalso include mononuclear phagocytes [9–11]. This concept,however, has not been generally accepted, the main reasonbeing that a complete characterization of these cells is stilllacking. Thus, there is no agreement on the nature of a cell thatcomprises about 50% of freshly isolated murine BM cells andthat we observed also in peripheral blood (PB) and in tissueduring different forms of inflammation [12–14]. Cells withring-shaped nuclei apparently are not confined to the murinesystem but are also detected in other rodents such as rats andhamsters [1, 3, 8, 15] as well as in humans, e.g., in the PB ofpatients with infectious mononucleosis [16] or with myeloprolif-erative diseases [17–19]. In addition, granulocytic ring cellshave been described in patients with severe alcoholism [20].Yet, these human cells also have not been further classified.

The aim of our study was to characterize murine ring cellsboth functionally and morphologically and to provide criteriafor a distinction between possible subpopulations of these cells.On the basis of morphological (including ultrastructural),immuno-, and enzymohistochemical, as well as functionalcriteria we could show that several populations of cells withring-shaped nuclei exist in the murine system: polymorpho-nuclear-like (PMN-like) ring cells, which mostly include ma-ture cells of the granulocytic lineage, and a more heteroge-neous, mononuclear cell-like population (MNC-like ring cells)that encompass mature monocytes, but also precursor cells ofboth the granulocytic and monocytic lineage and even progeny

Abbreviations: ring cells, cells/leukocytes with ring-shaped nuclei; BM,bone marrow; PB, peripheral blood; PMN, polymorphonuclear cells/leukocytes;MNC, mononuclear cells; PBS, phosphate-buffered saline; DAB, diaminobenzi-dine; BrdU, bromodeoxyuridine; BSA, bovine serum albumin; HBSS, Hanks’balanced salt solution; PI, propidium iodide; FSC, forward light scattering;SSC, orthogonal light scattering; PO, peroxidase; GM-CFC, granulocyte-macrophage colony-forming cells; M-CFC, macrophage colony-forming cells;FITC, fluorescein isothiocyanate; DMEM, Dulbecco’s modified Eagle’s me-dium.

Correspondence: Cord Sunderkotter, M.D., Institute of Experimental Derma-tology and Department of Dermatology, Von-Esmarch-Strasse 56, D-48149Muenster, Germany. E-mail: [email protected]

Received April 23, 1998; revised September 30, 1998; accepted October 3,1998.

Journal of Leukocyte Biology Volume 65, February 1999 217

of hematopoietic stem cells. Thus it is not correct to considercells with ring-shaped nuclei solely as myelocytes or granulo-cytes, which is done explicitly in several studies on murine orrat leukocytes [1–3, 5–7] and apparently also implicitly in otherstudies.

METHODS

Mice

BALB/c mice from Charles River Wiga GmbH (Sulzfeld, Germany), 8–12weeks of age, were used in all experiments.

Preparation of BM cells

BM cells were obtained as described previously [13, 21]. Briefly, BM cells wereflushed from femurs of BALB/c mice. After osmoshock cells were grown inDulbecco’s modified Eagle’s medium (DMEM) supplemented with 15% heat-inactivated fetal calf serum (PAA Laboratories, Linz, Austria), 5% L cell-conditioned medium [as a source of macrophage colony-stimulating factor(M-CSF), 2% L-glutamine (Seromed, via Biochrom, Berlin, Germany], 1%Kanamycin (Seromed), and 1% nonessential amino acids. BM cells at a densityof 4 3 105 cells/mL culture medium were cultured in Teflon-coated bags at37°C with 7% CO2 for 1–6 days. Cytospin preparations of 2 3 105 cells/mL offreshly isolated cells and cells cultured for 1, 2, 3, 4, 5, and 6 days were fixedfor 10 min in acetone.

Preparation of peripheral blood cells

PB was obtained by axillary puncture of mice after anesthesia with intraperito-neal injection of 300 µL ketamine (Ketanest, Parke-Davis, Berlin, Germany)and collected in heparin-coated tubes. Erythrocytes were depleted by lysisusing distilled water. Leukocytes were resuspended in double-concentratedDMEM (Seromed) with NaHCO3 (3.7 g/500 mL) and washed. PB cells werecultured for 1 and 2 days with supplemented medium as described for BM cells.Cytospin preparations of freshly isolated and cultured PB cells were made asdescribed above.

Morphological analysis

Differential morphological analysis of BM and PB cells was performed onMay-Grunwald-Giemsa-stained cytospin preparations and blood smears. Percytospin or blood smear 400 cells were counted at 3400 magnification andassigned according to morphological criteria presented in Table 1 [1, 3–8, 11].

Detection of apoptosis

BM cells (day 2) were fixed in 4% neutral buffered formalin for 10 min beforepreparation of cytospins. Apoptotic cells in cytospins were detected by directimmunoperoxidase detection of digoxigenin-labeled genomic DNA using an insitu apoptosis detection kit (ApopTag Plus Peroxidase, Oncor, Gaithers-burg, MD).

Adherence of macrophages to cultureplates/plastic surfaces

For characterization of adherent macrophages BM cells in culture medium wereincubated at a density of 5 3 105 cells/mL in plastic Petri dishes (Greiner,Solingen, Germany) and in tissue culture chambers (Lab-Tek, chamber slide,Nunc, Wiesbaden-Biebrich, Germany). After 2 and 3 days of culture nonadher-ent cells in the supernatant were removed by thorough washing with phosphate-buffered saline (PBS) and the cells attached to the bottom of Petri dishes andchamber slides were stained by May-Grunwald-Giemsa or Hemalum.

Phagocytosis of latex and L. major

For determination of phagocytic activity a standardized latex suspensioncontaining 5.4 3 109 latex particles/mL with an average diameter of 0.81 µm(Difco Laboratories, Detroit, MI) was added at a concentration of 2.5 µL per 106

cells (ratio of cells to latex particles 1:14) to the BM cultures in Teflon bags or toadherent cells on chamber slides. After 24 h of incubation (37°C) cytospinpreparations were made as described above. Phagocytosis was also studiedusing L. major promastigotes (strain L38) grown in RPMI medium as describedpreviously [13]. Freshly cultured BM cells were exposed to promastigotes at aratio of 1:6 and harvested after 18 h of incubation. Half of the cells (5.5 3 106

cells) were processed for electron microscopy as described below, the otherswere used for cytospin preparations. Phagocytosis of L. major was evaluatedimmunohistochemically using the monoclonal rat IgG2b antibody 8E7 againstL. major promastigotes and amastigotes [22]. Cells were examined at 3400magnification for presence of intracellular parasites or ingested latex particles.Because one or two single particles are sometimes loosely associated with thecell surface, cells were scored as positive when they contained at least three ormore latex particles in the cytoplasm.

Electron microscopy

For ultrastructural analysis of BM cells, freshly isolated cells and cultured cellsfrom day 1 and 3 were processed according to the method described by Spurrusing Epon embedding [23]. On the basis of ultrastructural criteria [6, 7,24–32] (Table 2) cells were divided into five categories.

Peroxidase stainingElectron microscopy

For ultrastructural detection of endogenous peroxidase, freshly isolated andcultured BM cells from day 3 and PB cells were processed according to themethod described by Robbins et al. [33]. Briefly, cultured cells were washed inPBS and fixed in 2% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 30min at 4°C, then incubated for 30 min in 10 mL Tris buffer containing 10 mg3,38-diaminobenzidine (DAB) tetrahydrochloride (Merck, Darmstadt, Ger-many) in the absence of H2O2, followed by incubation for 60 min in thepresence of 0.01% H2O2. Cells were then postfixed for 60 min in 2% OsO4 inPBS, dehydrated in ethanol, and embedded in Epon. Thin sections wereexamined uncontrasted and also contrasted with uranyl acetate (1 minincubation) and lead citrate (2 min incubation). Controls consisted of cellsincubated in the DAB-Tris medium in the absence of H2O2.

DNA synthesis

Incorporation of bromodeoxyuridine (BrdU) into replicating cells was measuredto determine the proliferative ability of the different BM cell populations and ofPB cells. Ten micromolar BrdU (Sigma, Deisenhofen, Germany) was added tothe BM cultures and to the PB cultures. After 24 h of incubation, cytospinpreparations were made and BrdU-incorporation was detected immunohisto-chemically with the use of a monoclonal rat antibody against BrdU [34] (HarlanSera-Lab, Sussex, UK) 1:25 diluted in PBS containing 1% bovine serumalbumin (BSA). Controls were done with nonspecific rat IgG and by staining

TABLE 1. Morphological Classification of Murine Leukocytes withClosed Nuclei and of Leukocytes with Ring-Shaped Nuclei

Nucleus Cytoplasm

Monocytes round, ovoid orkidney-shaped,occupies up to 50% ofthe overall area

dark, greyish-blue,frequently withvacuoles

PMN/neutrophils ring-shaped, constrictedor lobular

faintly stained granules,sometimes not visible;no vacuoles

MNC-like ringcell type

broad ring of round orovoid shape andmostly regular contour

diameter of thecytoplasmic center issmaller than the widthof the ring

PMN-like ringcell type

slender ring, constrictedor lobular and ofirregular contour

diameter of thecytoplasmic center islarger than the widthof the ring

According to the studies of Schermer [1], Hardy [3], Beran and Tribukait [4],Gardner [5], Goud [6], Goud et al. [7], Schalm et al. [8], and Pels et al. [11]according to our criteria.

218 Journal of Leukocyte Biology Volume 65, February 1999 http://www.jleukbio.org

cells cultured in BrdU-free medium. The proportion of cells incorporating BrdUwas scored by counting at least 400 cells and the number of positive cellsbelonging to the respective cell types was recorded.

Proliferation assay

To investigate the proliferative ability of BM and PB cells, cytospin prepara-tions of freshly isolated and cultured BM and PB cells of day 1 were studiedimmunohistochemically with the use of a polyclonal rabbit antibody directedagainst the murine Ki67 antigen (Dianova, Hamburg, Germany). Normal rabbitserum was used as control.

EnzymohistochemistryLight microscopy

For analysis of chloroacetate esterase, cytospin preparations of freshly isolatedand cultured cells were fixed in a mixture of 37% buffered formalin (pH 6.6),

acetone, and distilled water (25:45:30) at 4°C for 30 s, and then stained by themethod of Yam et al. [35], using naphthol AS-D chloroacetate (Sigma) assubstrate. Nonspecific esterase activity was demonstrated according to themethod of Yam et al. using a-naphthyl acetate (Sigma) as substrate after fixationof cytospins in acetone for 10 min. Counterstaining was done with hemalum for 30 s.

Immunohistochemistry

Immunohistochemical staining of cytospin preparations was performed usingthe indirect immunoperoxidase assay as described previously [12] and usingthe antibodies listed in Table 3 [36–58] in appropriate concentrations. Doubleimmunostaining of cells was performed using the immunofluorescence method.After preincubation in 1% BSA/PBS for 1 h, cytospin preparations wereincubated in sequence with an unconjugated primary antibody (ER-MP12 orER-HR3) for 1 h, then with phycoerythrin (PE)-conjugated F(ab’)2 fragmentgoat-anti-rat IgG (H 1 L; Dianova) 1:50 diluted in BSA/PBS for 1 h, followed

TABLE 2. Ultrastructural Criteria of Murine Leukocytes

Criteria Neutrophil Monocyte

NucleusShape small with several nuclear lobes large, kidney-shaped, irregular, or indentedNucleolus absent one or more (small)Chromatin heterochromatic euchromatic, peripherally condensed

CytoplasmSurface smooth or with pseudopodia, free from larger granules many projections, pseudopodia filled with organellesPrimary granules

Number numerous less numerous than in PMNSize/shape large, round, ellipsoid, or elongated variable in shape (mostly round); smaller than in

PMN, but larger than in promonocytesSecondary (specific) granules even more numerous, smaller, and less dense than

primary granulesabsent

Golgi apparatus small, centrally situated, adjacent to the nucleus well developed, associated with centriolesMitochondria sparse numerous, long, and slim

Rough ER scanty moderate amount; much more cisternae than in PMNVesicles/vacuoles a few (tiny) vesicles numerous vesicles, even more than in promonocyte;

vacuolesLocalization of peroxidase-activity primary granules 1 to 111 positive primary granules 1 to 11 (1) positive

Criteria Promyelocyte Myelocyte Promonocyte

NucleusShape large, round, slightly indented,

or doughnut-shapedlarge, eccentric, more indented than

precursorlarge, round, reniform, irregular, or

indentedNucleolus one or more if present, smaller than in promyelocyte two or moreChromatin euchromatic less euchromatic euchromatic/moderately condensed

CytoplasmSurface smooth smooth finger-like projections, microextensions,

only one pseudopod (with organelles)Primary granules

Number frequently numerous variable number a fewSize/shape large, round, or ellipsoid large, round, or ellipsoid small, round, or elongated; mainly

perinuclearSecondary (specific)

granulesabsent variable number; less dense than

primary granulesabsent

Golgi apparatus large large largeMitochondria several large, numerous several, adjacent to the nucleus

Rough ER moderate amount, dilatedcisternae

dilated moderate amount (few strips), not dilated

Vesicles/vacuoles rarely rarely clear vesicles and a few vacuoles arefrequent

Filaments rarely bundled or scattered filament are typicalLocalization of peroxidase-

activityall granulesa 111rough ER 1nuclear envelope 1Golgi apparatus 1

only some primary granules: 1 to 111 many (not all) granules 11 (1)rough ER 1nuclear envelope 1Golgi apparatus 1

According to van Furth et al. [24], Florey [25], Bainton et al. [26], Nichols and Bainton [27], Goud [6], Goud et al. [7], Cohn [28], Bainton [29], Cross and Mercer[30], Bloom and Fawcett [31], Morioka et al. [32].

a Granules in the promyelocyte may contain crystalline inclusions that are never observed in promonocytes [27].

Biermann et al. Murine leukocytes with ring-shaped nuclei 219

by a 1-h incubation with a fluorescein isothiocyanate (FITC)-conjugatedprimary antibody [ER-MP20 FITC or FITC Ly-6G (GR-1)]. For negativecontrols, specific antibodies were replaced by an isotype of rat IgG. All thesecontrols were negative, indicating specificity of our antibodies.

Cell sorting

For analysis of GR-1-positive cells, single-cell suspensions in 96-well microti-ter plates were incubated with 1% BSA in Hanks’ balanced salt solution (HBSS;Seromed) to minimize unspecific staining and incubated with GR-1 FITC for 30min on ice.

For double labeling with GR-1 and BM8, cells were first incubated withunconjugated BM8 (Table 3), then with goat-anti-rat IgG PE (Dianova), andfinally with GR-1 FITC. Corresponding controls encompassed those forautofluorescence (cells in HBSS only), goat-anti-rat IgG PE, rat IgG (isotype)with goat-anti-rat IgG PE, rat-IgG FITC, rat IgG (isotype) with goat-anti-rat IgGPE, and rat-IgG FITC (isotype). Just before FACS analysis and sorting, cellswere stained with propidium iodide (PI) at 1 µg/mL in HBSS. Flow cytometryand cell sorting were performed on a Moflo High Speed Sorter and Analyzer(Cytomation) using an argon ion laser (Coherent, Palo Alto, CA) with anexcitation wavelength of 488 nm. Data acquisition was done with the CyCLOPS3.4 software (Cytomation). Forward light scattering (FSC), orthogonal lightscattering (SSC), and fluorescence signals (FL1, FL2, FL3) were acquired andstored in listmode data files. Each measurement contained a defined numberof 2 3 104 vital, PI-negative cells. Gates of the bivariate and univariatehistograms were logically connected in the following order: vital cells from gate1 of PI/FSC to FSC/SSC, gating leukocytes to FITC/PE. The separately gatedFITC signals at the mode values of GR-1low and GR-1high with the defined FSC,SSC, FL1, and FL3 characteristics were simultaneously isolated with a sort rateof 2 3 104 cells per second.

Cells negative for leukocytic lineage markers were selected by gating outcells stained with a cocktail of unconjugated primary antibodies containingMac-1, GR-1, B220, CD3, Ter 119 [57, 58], followed by incubation withgoat-anti-rat IgG PE. After exclusion of the PE bright cells the remainingnegative cells were stained with anti-Sca-1, anti-c-kit, and anti-CD38 (Table 3).

For subsequent morphological characterization of sorted cells at least 5 3

104 cells from each population were sorted into plastic tubes filled with 150 µLHBSS. Cell suspensions were concentrated on slides by cytocentrifugation. Foranalysis of further development in vitro, sorted cells were collected intoeight-well glass tissue culture chambers (Nunc) at 5 3 104 cells in 200 µL andcultivated as described for BM cells.

Animal model

The left ears of mice were injected subcutaneously with 20 µg of LPS(Escherichia coli serotype 055:B5, Sigma) or solely with PBS as a control. After3, 6, 8, 12, 16, 24, 36, and 48 h, mice were killed and ears were fixed in 10%buffered formalin and embedded in paraffin. Paraffin sections were cut on amicrotome and stained with hematoxylin-eosin and the PAS method. Tissuesections were examined with a Zeiss microscope. To compare areas of equal size10 graticule fields were evaluated in each section.

RESULTS

Distinction of ring cells by morphology andadherence to plastic surfaces

On smears or cytospin preparations of murine BM cultures wewere able to assign several leukocytes to the granulocytic ormonocytic or lymphocytic lineage (Fig. 1) using the criteriaelaborated by several detailed studies (e.g., nuclear shape,qualities of cytoplasm) and summarized in Table 1 [1, 3–8, 11].However, these criteria, especially those based on nuclearfeatures, did not allow to unambiguously classify the cells thathad ring-shaped nuclei (ring cells). The percentage of thesecells was marked in the first days of culture, comprising at least50% of freshly isolated BM cells and 35% of cultured BM cellsat day 2 (Fig. 1a, Table 4). We provisionally establishedmorphological criteria to distinguish between two types of ringcells (Table 1). The nuclei of some of these cells clearly showedconstrictions into lobes as also seen in PMN, and the width oftheir karyoplasmic ring was smaller than the diameter of itscytoplasmic center. We referred to them as PMN-like ring cells.Most cells had a granular cytoplasm and a few had eosinophilicgranules (eosinophils). The percentage of PMN-like ring cellssteadily declined from 37.4% on day 0 to 0.5% on day 6 (Table4). In another population the nuclei had a round or ovoid,

TABLE 3. Monoclonal Antibodies Against Markers of Murine Leukocytes and Erythroid Cells

Antibody/clone (cl) Rat-Ig class Specificity Reference/source

F4/80 cl: A3-1 IgG2b differentiation marker for cells of the monocytic lineage after the M-CFCstage; eosinophils

[36–40] Serotec

ER-HR 3 IgG2c mature cells of monocytic lineage [40–42] BMABM8 IgG2a monocytes and myeloid precursors [40] BMACD13 cl: R3-242 IgG1, k mouse aminopeptidase N in macrophages, monocytes, dendritic cells, B cells [43] PharMingenER-MP 12 IgG2a myeloid precursor cells/early macrophage precursors; GM-CFC, lymphocytes;

identical to CD31 (Pecam-1)[44, 45, 40, 46–49] BMA

ER-TR 3 IgG2b murine MHC class II antigen [50] BMAGR1 (Ly-6G) RB6-8C5GR-1 FITC

IgG2b differentiation antigen on granulocytes. Myelocytes 1 granulocytes; strongexpression; myeloic precursors; low expression (transiently also expressedon the monocytic lineage)

[51, 52] PharMingen

MRP 14 IgG2b granulocytes and transiently also expressed on cells of the monocytic lineage [14] own productionMAC-1 (Ly-40) cl: M1/70 IgG2b CD11b/CD18; myeloic cells beyond the M-CFC stage; natural killer cells [40, 53–56] SerotecER-MP 20 (Ly-6C)ER-MP 20-FITC

IgG2a mid-stage macrophage precursors from MCF-C until monocytes, somegranulocytes; lymphocytes (own observation)

[40, 45–47] BMA

Sca-1 Ly-6A/E IgG2a more committed progeny of murine hematopoietic stem cells in BM [57, 58] PharMingenc-Kit (2B8) IgG2a CD117; hematopoietic progenitor cells [58] PharMingenCD 38 (cl. 90) IgG2a T and B cells [58] PharMingenTER 119 (Ly-76) IgG2b erythroid cells [58] PharMingenCD3 cl: 17A2 IgG2b T lymphocytes [58] PharMingenB220 IgG2a CD45R; B cells [57, 58] PharMingen

Serotec, Wiesbaden, Germany; BMA Biomedicals AG, Augst, Switzerland; PharMingen, Hamburg, Germany.


smooth contour, compatible with, e.g., mononuclear cells(MNC), and their karyoplasmic ring was wider than thediameter of its cytoplasmic center. This group of MNC-like ringcells comprised 15.1% of all cells and decreased to 0.6% onday 6 of culture (Table 4). Thus it did not contain the fullymature forms of macrophages, but still appeared to be morpho-logically heterogeneous with regard to size and cytoplasmicqualities. According to hue of the cytoplasm, the quality, andnumber of granules or vacuoles, they were likely to encompassmyeloic precursor cells such as promyelocytes or promonocytesin addtion to monocytes (Fig. 1b). There was one form of large

ring cells that sometimes contained one narrow segment in thering-shaped nucleus similar to PMN-like ring cells, but whichwe assigned to MNC-like ring cells because of their size andbecause the remaining segments of their large nuclei werebroader than in PMN-like ring cells and broader than thediameter of the cytoplasmic center. As outlined below, this cellpresented a precursor cell (see Fig. 10).

We first tested the possibility that ring cells encompassmonocytes or macrophages, which would challenge the thesisthat they present only PMN and myelocytes [1–3, 5–7].

Adherence of ring cells

One criterion to differentiate monocytes or macrophages fromPMN and most other leukocytes is adherence to glass or plasticsubstrates after a longer period of culture [7, 24, 59, 60, 61].Therefore we cultured BM cells in plastic dishes and tissueculture chambers. After 2 days and 3 days 1–2% of adherentcells revealed ring-shaped nuclei fulfilling the criteria forso-called MNC-like ring cells (Table 1).

Apoptosis

Another criterion for differentiation between PMN and mono-cytes is rapid apoptosis of PMN compared to longevity ofmonocytes [62–64]. We therefore labeled 38-OH ends offragmented DNA. We found that the majority of labeled(apoptotic) cells on day 2 were PMN and small PMN-like ringcells (53 and 25%, respectively), whereas MNC and MNC-likering cells comprised only a small percentage (19 and 3%,respectively) among the total number of apoptotic cells. Thus,both MNC and MNC-like ring cells do not undergo apoptosis asrapidly as PMN and PMN-like ring cells.

Phagocytosis of latex particles and L. major

After addition of latex particles or L. major, 43 or 44%,respectively, of MNC-like ring cells had phagocytosed latexparticles or L. major, many of them internalizing more than 20particles or more than 3 to 4 parasites (Figs. 2 and 3a). Incontrast, only a few small PMN-like ring cells had phagocy-tosed latex particles or L. major, and none of them hadinternalized more than six particles or more than two L. majorparasites. Thus, the MNC-like ring cells are phagocytes whosephagocytic activity was markedly higher than that of PMN-likering cells. It did not completely reach the phagocytic potentialof differentiated macrophages with closed nuclei.

Fig. 1. (a) Murine BM cells after 2 days of culture. The following cell types arepictured: PMN with segmented nuclei (thin arrows), PMN-like ring cells withlobular or constricted ring-shaped nuclei and a cytoplasmic center that is largerthan the width of the ring (thin double arrows), MNC with closed nuclei (thickarrows), MNC-like ring cell with a ring-shaped nucleus of smooth contour and acytoplasmic center that is smaller than the width of the ring (thick doublearrow). Cytospin preparations, May-Grunwald-Giemsa, original magnification3360. (b) Different types of ring cells among freshly isolated BM cells.PMN-like ring cell (A) with a nuclear ring showing several constrictions and alarge ring center; precursor type of MNC-like ring cell (B) showing a broadnuclear ring of round shape and regular contour, a small ring center, and abasophilic cytoplasm filled with small whitish granules; myelo/monocytic typeof MNC-like ring cell (C) showing a small cytoplasmic center, hardly visiblecytoplasm, and a nucleus of slightly irregular contour. Cytospin preparation,May-Grunwald-Giemsa, original magnification 31440.

TABLE 4. Percentages of Murine Bone Marrow Cellsat Different Time Points in Culture

Day ofculture

Monocytes/promonocytes

MNC-likering cells PMN

PMN-likering cells

Othercellsa

0 0.3 6 0.2 15.1 6 4.1 8.1 6 3.4 37.4 6 7.7 38.7 6 9.81 7.8 6 2.4 11.4 6 6.1 20.2 6 7.7 32.6 6 9.8 25.7 6 3.12 34.5 6 5.7 10.1 6 2.7 14.4 6 2.3 25.3 6 4.9 15.8 6 2.54 92.1 6 2.3 1.6 6 0.5 2.25 6 0.5 4.1 6 2.2 06 98.5 6 0.3 0.6 6 0.3 0.5 6 0.4 0.5 6 0.1 0

a Other cells include megakaryocytes, plasma cells, blast cells, and precur-sors of the myelocytic, monocytic, megakaryocytic, and erythroid lineage.


Ultrastructure

For a detailed characterization of ring cells with regard to theclassic cytomorphological criteria (Table 2), we performedultrastructural studies of BM cells. Our microscopic criteriareferring to the shape of nuclei, however, could not beuniversally applied on ultrathin sections, because it was notknown at which level the individual cells had been cut. Thus,the criterion ‘‘width of the ring’’ could become misleadingsimilarly to the criterion ‘‘contour of the ring’’: classic mono-cytes as well as monocytic ring cells sometimes showedmarked, but narrow indentations of their nuclei, not recogniz-able under the light microscope. However, we were able todistinguish between MNC-like ring cells and PMN-like ringcells by using defined ultrastructural criteria: one majorpopulation of ring cells, detected predominantly among freshlyisolated BM cells and among cells cultured for 1 day, fulfilledultrastructural criteria compatible with PMN such as cytoplasmwith lower density of organelles, but with numerous primaryand secondary granules (Table 2, Fig. 3b). Some ring cells

contained birefringent granules and were thus eosinophils.Another population of ring cells, which was found predomi-nantly among cells cultured for 3 days, showed qualitiescompatible with the ultrastructure of monocytes (Table 2; Fig.3c). In contrast to PMN-like ring cells these cells contained areduced number of granules of one single type and some cellshad a visible nucleolus in a mainly euchromatic nucleus. Inmany ring cells it was notable that some organelles, especiallythe Golgi apparatus, granules, and sometimes centrioles, wereconcentrated in the inner cytoplasmic center of the nuclear ring(Fig. 3, d–f ).

Thus, with regard to morphology, adherence, apoptosis, andphagocytosis we have revealed that one group of ring cellsconsists of PMN, including eosinophils, whereas another groupencompasses monocytes without the fully mature form ofmacrophages.

However, besides these two groups there was still anothersubpopulation of ring cells that on ultrastructure showed asimilar arrangement of organelles as the monocytic MNC-likering cells, except that their nucleus was almost completelyeuchromatic and that their cytoplasm contained fewer vesicles,no vacuoles, but variable numbers of granules (Fig. 3d). Thesecells could therefore correspond to precursors of myeloic cells.

Ultrastructural localization of peroxidase activity

To analyze which types of precursor cells occur among ringcells we analyzed BM cells for their intracellular localization ofendogenous peroxidase (PO) activity [24, 26, 27, 29] (Table 2).We found that ring cells showed different patterns of POdistribution and that they included cells fulfilling the criteriafor promonocytes (Fig. 3e) and monocytes as well as forpromyelocytes (Fig. 3f ), myelocytes, and PMN (Fig. 3g).Although the total number of precursor ring cells was low, wecould clearly identify ring cells whose PO activity was not onlyfound in granules, but also in nuclear envelope, rough endoplas-mic reticulum (ER), and Golgi apparatus (promyelocytes andpromonocytes). Closer analysis revealed that some of thesecells (4%) contained fewer granules that, however, were not allpositive and that had no crystalline inclusions. They thusfulfilled the criteria for promonocytes (Fig. 3e). Other, thoughfewer, cells (2% of all ring cells) were promyelocytes as theyhad a higher number of granules of different, ellipsoid, and

Fig. 3. (A) Electron micrograph of a monocytic MNC-type ring cell, day 1 of culture, 18 h after incubation with L. major. The cell shows the typical characteristics ofa mononuclear phagocyte: vesicles, mitochondria (m), cisternae of rough ER (er), and fingerlike surface projectons (p). The cytoplasm including the ring centercontains several L. major parasites (short arrows). Original magnification 36600. (B) Electron micrograph of a PMN-like ring cell, day 1 of BM culture. Like PMN thecytoplasm of this cell contains numerous granules of different size and shape: large, dense primary granules (P) and smaller, less dense secondary granules (S) Otherorganelles are sparse. The nucleus reveals a heterochromatic chromatin pattern. Original magnification 311,500. (C) Electron micrograph of a monocytic MNC-typering cell , day 3 of BM culture. Like a monocyte with a closed nucleus this cells shows a euchromatic nucleus with nucleolus (nu) and a cytoplasm filled with severalcisternae of rough ER (er), mitochondria (m), many clear vesicles, and a very few granules (g) of one single type. Original magnification 38900. (D) Electronmicrograph of a precursor MNC-type ring cell, day 3 of BM culture. This cell shows an almost completely euchromatic nucleus with nucleolus (nu), a moderateamount of cytoplasmic granules (g), numerous cisternae of rough ER (arrow), some mitochondria, a few vesicles, and a centriole (c) in the cytoplasmic ring center.Original magnification 36600. (E) Electron micrograph of a promonocytic ring cell. Day 0 of BM culture, staining for intracellular peroxidase, slightly contrasted for3 min. Strong activity of peroxidase is detected in granules (g); weak to moderate activity in rough ER (er) and nuclear envelope (ne). In addition to positive granules,peroxidase-negative granules (g-) are also present (see inset, original magnification 321,000), which identify this cell as promonocyte and differentiate it frompromyelocytes. Original magnification 36600. (F) Electron micrograph of a promyelocytic ring cell. Day 0 of BM culture, staining for intracellular peroxidase,slightly contrasted for 3 min. Strong activity of peroxidase is detected in all granules (g), weak to moderate activity in cisternae of rough ER (er) and nuclear envelope(ne). Note the accumulation of granules in the ring center of the cell. The absence of negative granules identifies this cell as promyelocyte. Note the dilated cisternaeof rough ER in the promyelocyte compared to the smaller cisternae in the promonocyte (panel e). Original magnification 36600. (G) Electron micrograph of a PMN(cell on the right) and a PMN-like ring cell (cell on the left), day 3 of BM culture, staining for intracellular peroxidase, slightly contrasted for 3 min. Peroxidaseactivity is exclusively found in granules (short arrows). Peroxidase-negative granules are also seen (long arrows). Original magnification 36600. ≤≤≤≤≤c

Fig. 2. Monocytic MNC-type ring cell (double arrow) and a PMN-like ring cell(double arrowhead) after 2 days of BM culture, 24 h after addition of latexparticles. The MNC-like ring cell contains eight phagocytosed latex particles inthe cytoplasm, whereas this PMN-like ring cell has not phagocytosed latexparticles. Cytospin preparation, May-Grunwald-Giemsa, original magnification31440.



round shape, sometimes revealing crystalline inclusions andalso a dilated ER (Fig. 3f ). It is easier to identify cells aspromonocytes than as promyelocytes because the presence ofone negative granule (which would exclude the promyelocyte)cannot always be ruled out with certainty. Similarly, nucleoli,for example, are sometimes, but not always, present in myelo-cytes, thus impairing differentiation between myelocytes andPMN. After 3 days of culture we also observed monocytic ringcells that were completely negative for PO, thus fulfilling thecriteria of resident macrophages or their immediate precursors[65, 66].

Thus, when considering ultrastructural criteria and distribu-tion of PO activity we found that ring cells also encompassprecursors of PMN (promyelocytes, myelocytes) and precursorsof monocytes (promonocytes). We experienced that the ultrastruc-tural criteria for differentiation between promonocytes andpromyelocytes [24, 26, 27, 29], however, are rather arbitraryand do not always allow a clear distinction.

Proliferative activity (DNA synthesis anddetection of Ki67)

To find out whether precursor forms of ring cells are capable ofDNA synthesis we investigated whether cultured ring cells inBM showed incorporation of BrdU. We found that on day 1 ofBM culture 33% of all MNC-like ring cells had incorporatedBrdU (Fig. 4), whereas less than 2% of PMN-like ring cellswere BrdU-positive. We also analyzed the expression of Ki67, aproliferation-associated nuclear protein, in freshly isolated andcultured BM cells (day 1 of culture). In contrast to PMN andPMN-like ring cells, which were only rarely positive for Ki67(,2% of positive cells), up to 9% of MNC-like ring cellsshowed presence of Ki67.

Enzymohistochemistry

Activity of different enzymes is used in hematology to assignleukocytes to the granulocytic or monocytic lineage. Strongactivity of unspecific esterase is used as a marker for monocytes[35]. In BM the percentage of MNC-like ring cells positive forunspecific esterase was continuously increasing during culturefrom 0.1% on day 0 to 100% on day 6 (Table 5, Fig. 5) [35].Most PMN and PMN ring cells showed no or very weak activityfor unspecific esterase (Table 5).

Moderate to strong activity of chloroacetate esterase, anenzyme primarily but not exclusively found in cells of thegranulocytic lineage, including promyelocytes and myelocytes[35], was consistently detected in almost all PMN-like ring cellsand PMN. In MNC with closed nuclei there was weak activity. Itis interesting to note that, among most MNC-like ring cells thepercentage of intensely positive cells was as high as in PMNand in PMN-like ring cells (Table 5). The intensity of stainingwas most marked in the inner cytoplasmic center of the nuclearring, consistent with our ultrastructural finding of accumulatinggranules in the center of nuclear rings. Statistically, there mustbe one population of MNC-like ring cells after 3 days of culturethat was positive for both unspecific and chloroacetate esteraseand that could not be clearly assigned to the monocytic orgranulocytic lineage.

Immunohistochemistry

Because macrophages, their subtypes, and their precursors, aswell as other leukocyte subtypes, can partially be identified bytheir staining pattern with different antibodies (Table 3), wefurther analyzed the immunophenotype of ring cells. They didnot include T or B lymphocytes because they were negative forCD3 and CD45R (B220). A panel of antibodies claimed topreferentially recognize macrophages (F4/80, BM8 CD13,ER-HR3) stained several MNC-like ring cells in BM (Fig. 6a),especially during later stages of culture (Table 6) and stainedcorrespondingly almost all ring cells that adhered to glasssurfaces or that had phagocytosed latex or L. major. Markers forboth monocytes and PMN (MRP14, MAC-1, and ER-MP20)were detected on both MNC-like and PMN-like ring cells(Table 6, Fig. 6b). The marker GR-1, usually claimed tospecifically stain PMN, stained 95% of the PMN-like ring cells(BM day 0). In contrast to MRP14, it did not stain large cellswith lobulated nuclei which on immunohistochemistry couldnot always be unambiguously assigned to PMN- or MNC-likering cells. It was, however, expressed by some typical MNC-likering cells, especially early during culture (Table 6). These cellscould be myelocytes but GR-1 is also transiently expressed onprecursor cells of the monocytic lineage [51, 67]. ER-MP12, amarker for myeloid precursors [44, 45], was positive on 26% ofMNC-like ring cells on day 0, but not on MNC-like ring cells oflater stages. It was not positive on adherent MNC-like ring cellsor on adherent monocytes, so it appears to be down-regulatedbefore monocytic cells become adherent. Thus, some MNC-likering cells also immunohistochemically present parameters ofprecursors of monocytes.

Because, especially during day 2 of culture, a high percent-age of MNC-like ring cells was positive for GR-1 and for F4/80or ER-HR3 (Table 6), and because 32% of chloroacetateesterase-positive MNC-like ring cells also showed high activityof unspecific esterase (Table 5), we wondered whether someMNC-like ring cells could simultaneously present markers ofboth lineages. On double immunofluorescence staining withGR-1 and ER-HR3 or BM8, we found that indeed there wereGR-11BM81 as well as GR-11ER2HR31 MNC-like ring cellspresenting 8% (7/83 cells) on day 1 of culture (Fig. 7, a and b).We then analyzed whether these cells could be further differen-tiated according to the density of the GR-1 surface antigen [67].

Cell sorter analysis

Cell sorting offers an elegant way both to distinguish cellsaccording to antigen density or according to presence of acombination of surface markers and to subsequently analyzetheir morphology or their further development. Sorting forGR-1-positive cells we found that in BM (and in PB) there arecells with a low mean and a high mean fluorescence for GR-1(GR-1low, GR-1high), as described by Lagasse and Weissman[67]. BM cells day 2 contained 1.5-fold more GR-1low cells thanGR-1high cells, whereas PB contained 3.8-fold more GR-1high

than GR-1low cells (data not shown). Among the macrophagemarkers, BM8 was best suited for intense and distinct stainingpatterns. Sorter analysis of BM cells day 0 and day 2 revealed asmall percentage of GR-1high and BM81 cells (BM day 0, 3.8%;BM day 2, 0.78% Fig. 8), and a larger percentage of GR-1low


BM81 cells (BM day 0, 5.8%; day 2, 3.7%). A high percentagewas GR-1high or GR-1low, but BM8-negative (Fig. 8). Subsequentmorphological analysis of the sorted populations revealed thatGR-1high BM82 cells contained mostly PMN or small PMN-likering cells. In the small fraction of GR-1high BM81 cells, mostcells turned out to contain mainly monocytic cells that hadphagocytosed (apoptotic) GR-1high particles of PMN. The GR-1low

population was very heterogeneous and contained severalMNC-type ring cells as well as various cells with closed nuclei.It is known to comprise promyelocytes, promonocytes, mono-cytes, and B cells [67]. When this fraction, however, wasrestricted to the GR-1low BM81 population it was enriched formonocytic cells (Fig. 9), i.e. for monocytes (45%), promono-cytes and promyelocytes (49%), and MNC-like ring cells (6%;not further differentiated). The GR-1negativeBM81 populationwas negligible at day 2 (0.55%), but later increased andcontained monocytes or macrophages (data not shown). In afurther sorter analysis we analyzed the double-negative popula-tion. It was found to contain small and large cells with closed

nuclei, but also large cells with ring-shaped nuclei whoseslightly lobulated karyoplasm sometimes contained one narrowkaryoplasmic segment (Fig. 10). Thus, double labeling andcell sorting revealed that PMN-like ring cells can be separated

Fig. 4. DNA-synthesizing monocytic ring cell. Day 3 of BM culture, 24 h afterincubation with BrdU, cytospin preparation immunostained for BrdU. TheMNC-like ring cell (double arrow) but not the monocyte shows intranuclearincorporation of BrdU. Original magnification 31440.

Fig. 5. Activity for unspecific esterase in MNC-like ring cells. Day 3 of BMculture, cytospin preparation stained for unspecific esterase. Two monocyteswith closed nuclei and a MNC-like ring cell (double arrow) reveal activity forunspecific esterase. Original magnification 31440.

Fig. 6. (A) F4/80-positive monocyte and monocytic MNC-like ring cell. Day 4 ofBM culture, cytospin preparation stained for F4/80. A monocyte (arrow) and amonocytic MNC-like ring cell (double arrow) are positive for F4/80. Originalmagnification 3360. (B) MRP14-positive cells. Day 3 of BM culture, cytospinpreparation stained for MRP14. PMN with segmented nuclei (short arrow), PMN-likering-cells (short double arrow), a monocytic cell (long arrow) and a MNC-like ring cell(long double arrow) are positive for MRP14. Original magnification 3360.

TABLE 5. Distribution of Unspecific Esterase/ChloroacetateEsterase in Murine Bone Marrow Cells

Intensityof staining(in caseof positivereaction)day of culture

Cells of theMNC lineage

moderate to strong11 to 111/

1 to 1 (1)

MNC-likering cells

moderate to strong11 to 111/

11 to 1111

PMNvery weakto weak

(1) to 1/11 to 1111

PMN-likering cellsvery weakto weak

(1) to 1/11 to 111

0 33a/0%b 0.1/100% 0/94% 0/94%1 21/33% 3.2/96% 1.8/96% 0/95%2 54/63% 32/100% 3.2/100% 13.5/99%4 95/80%d 67/100% 13/86% 30/100%6 95/92%d 100/100% 33/100% /c

a Percentage of unspecific esterase-positive cells of all cells of the MNClineage in bone marrow.

b Percentage of chloroacetate esterase-positive cells of all cells of the MNClineage in bone marrow.

c Cells of this type were not detectable.d From day 3 on 7–10% of cells of the MNC lineage also revealed a strong

111 staining intensity.


from MNC-like ring cells on the basis of high expression ofGR-1. It also added further evidence that MNC-like ring cellscomprise a heterogeneous group of monocytic cells in whichmonocytes and their late precursors (e.g., promonocytes) areenriched among the GR-1low BM81 population. However, thereremains a small group of MNC-like ring cells that do not stainmarkedly with the lineage markers GR-1 and BM8 and thatcould therefore present myeloic progenitors.

For further analysis of the precursor status we stained day 0BM cells immunohistochemically with a cocktail of lineagemarkers for lymphocytes (B220, CD3) for erythroid cells(Ter119) as well as for myeloic cells (Mac-1, GR-1) [57, 58] andanalyzed negative cells. About 20% of the negative cellspresented with a large lobulated ring-shaped nucleus (Fig. 10).This karyoplasm contained broad and narrow karyoplasmicsegments (Fig. 10) and left a cytoplasmic center of variable size.Some of these nuclei at first glance appeared similar to nuclei ofPMN-like ring cells, however, their broad sections were widerthan those seen in PMN-like ring cells and wider than thecytoplasmic center, and in addition the whole cell was largerthan PMN. We then sorted for cells that were negative or onlyfaintly positive for lineage markers. Cells that were negative orbarely positive for the cocktail of lineage markers (lin-eagenegative/low) amounted to 10.2%. These lineagenegative/low cellsstill contain cells that are not committed yet to the myeloic orlymphocytic lineage [57, 58]. Morphological analysis of thesecells revealed that 20.2–25.1% of these lineagenegative/low cellswere those ring cells with large nuclei that sometimes containeda segment (Fig. 10). These cells comprise about 2–2.5% of BMcells day 0. Sorting for lineagenegative/low cells was able to at least

enrich this particular population of large ring cells, whereasother analyses were not suitable to partially separate cells withring-shaped nuclei from those with closed nuclei. In a firstattempt to further characterize these lineagenegative/low ring cellswe stained them for the markers CD38, c-kit, and sca-1. Over95% of these cells were CD382, but c-kit1 and sca-11.According to Randall et al. [58] this staining pattern wouldassign these ring cells to the more committed progeny of murinestem cells. Although this progenitor status needs to be furthercharacterized our study has revealed that murine BM ring cellsappear very early during leukocyte differentiation.

Cells in peripheral blood

Because ring cells encompass PMN and monocytes it was notsurprising to find these cells also in PB. On differential bloodcount MNC-like ring cells comprised 0.95% and PMN-like ringcells 12% of all leukocytes, whereas monocytes and PMN

Fig. 7. (A and B) Ring cells positive for both the macrophage marker ER-HR3and the granulocyte marker GR-1. Day 2 of BM culture, cytospin preparation,immunostained for both GR-1 (A) (green fluorescence) and ER-HR3 (B) (redfluorescence). A MNC-like ring cell with a small cytoplasmic center showsstaining for GR-1 especially in the ring center (long arrow). The red reactionproduct of ER-HR3 staining (visualized by a second antibody conjugated withPE) appears as distinct, brightly yellow staining within the green fluorescenceof FITC-GR-1-staining (short arrow) (original magnification 31500). (b) Samesection as in panel a. The ER-HR3 staining appears as red fluorescence at thecell periphery (short arrow). Note the other cells which are FITC-GR-1-positivebut ER-HR3-negative (5PMN) (arrowheads) and the GR-1-positive andER-HR3-negative (PMN-like ring cell; double arrowhead). Original magnifica-tion 31500.

TABLE 6. Immunophenotyping of Murine Bone Marrow Cells

Cells of theMNC lineage

MNC-likering cells PMN

PMN-likering cells

Macrophage-markerF4/80

2 52%a 33% 3.4% 2.5%4 83% 60% 0% 0%6 90% 100% 0% 0%

BM82 63% 32% 20% 25%4 71% 51% 8% 7%6 90% 75% 0% /b

Marker for early macro-phage precursors

ER-MP 122 10% 27% 0% 0%4 0.3% 25% 0% 0%6 0% 0% 0% 0%

PMN markerGR 1

2 3.4% 48% 96% 95%4 0.3% 17% 100% 89%6 0.26% 33% 100% 100%

Marker for bothMRP 14

2 13.6% 88.6% 100% 100%4 0.3% 40% 100% 75%6 0% 0% 100% 100%

a Percentage of positive cells of all cells of the MNC lineage in bone marrow.b Cells of this type were not detectable.


comprised 1 and 32%, respectively. Strong activity of chloroac-etate esterase was consistently detected in almost all PMN-likering cells, PMN, and in several MNC-like ring cells. Around17% of monocytes were weakly positive. Immunohistochemi-cally all MNC-like ring cells expressed F4/80 or BM8, severalexpressed GR-1, and some expressed CD13 (25%) or ER-HR3(33%). Thus, at least one-third of MNC-type ring cells in PBhad the phenotype of monocytes. It is interesting to note thatsome of the cells released into the PB have retained theircapacity to proliferate in certain situations as some cellswith closed nuclei (2%) and also some ring cells (1.5%)stained positive with Ki67 and 0.5% of ring cells werepositive in the BrdU assay. Sorter analysis of PB cells forGR-1high and GR-1low as mentioned above, showed that PBcontained 3.8-fold more GR-1high than GR-1low cells [67]. Mostof the GR-1high are known to be PMN, whereas GR-1low areknown to contain lymphocytes in addition to monocytes andmyelocytes [67]. Our results imply that in murine blood there

are some ring cells that bear characteristics of both thegranulocytic and monocytic lineage, not allowing a cleardistinction. Sorting for PB cells that were negative or faintlypositive for leukocytic lineage markers yielded some, but veryfew cells (, 0.01%).

Cells in inflammatory infiltrates

Ring cells were also found in inflammatory infiltrates after localinjection of LPS, but here an accurate quantitative analysis isdifficult. PMN-like ring cells could be identified more easily

Fig. 8. FACS analysis and sorting of cellsdouble-labeled for GR-1 and BM8. MurineBM cells (day 2) were stained with unconju-gated BM8, PE-conjugated secondary anti-body, and anti-GR-1 FITC. Cells were addi-tionally differentiated according to low andhigh expression of GR-1. The different frac-tions are outlined and described in Table 7.

Fig. 9. MNC-type ring cell among a sorted fraction of GR-1low- andBM8-positive cells. Cells belonged to sorted cell fractions described in Figure8. Cytospin preparation, May-Grunwald-Giemsa, original magnification 3540.

Fig. 10. Lineagenegative/low ring cell (BM day 0), which was negative afterimmunohistochemical staining with a cocktail of lineage markers. It shows aslighty lobulated nucleus and one narrow karyoplasmic segment in addition tothe remaining broad nuclear segments. These cells were enriched by sorting forlineagenegative/low cells and consisted of large ring cells whose nuclei wereslightly lobulated and sometimes contained one narrow segment similar to thesmaller PMN-like ring cells, but whose remaining karyoplasmic sections werebroader than the cytoplasmic center and broader than in PMN-like ring cells.Cytospin preparation, original magnification 31440.


than MNC-like ring cells due to the larger centers of theirring-like nuclei. They comprised between 20 and 25% of allgranulocytes at early time points after injection of LPS.MNC-like ring cells formed a small percentage (,1%), which,however, rose within 48 h together with the percentage ofmonocytic cells with closed nuclei (from 3 to about 40%). Atthis time there was no monocytosis on differential blood count(0.8 6 0.4% monocytes vs. 0.4 6 0.2% before injection ofLPS). Immunohistochemically the rising percentage of mono-cytes was confirmed, but the morphology of cells in cryopre-served tissue was not well enough maintained for accurateanalysis of ring cells.

DISCUSSION

The aim of our study was to classify murine cells withring-shaped nuclei. We found that these cells encompass about50% of all cells from freshly isolated murine bone marrow andare also observed in peripheral blood and in inflammatoryinfiltrates. For a detailed evaluation we performed morphologi-cal (including ultrastructural), enzymohistochemical, immuno-histochemical, as well as functional studies, thus combining theclassic morphological methods with recently developed proce-dures. By widely accepted criteria [7, 21, 24, 32, 37, 40, 42, 44,68] we were able to clearly assign ring cells to differentsubtypes of leukocytes. One morphologically distinct group ofring cells showed the characteristics of granulocytes, such assecondary and primary granules, high expression of GR-1, highrate of apoptosis, and only low uptake of L. major. Theremaining, smaller group was fairly homogeneous with regard toshape of the nuclear ring, but was heterogenous with regard toother criteria. Closer analysis showed that this group included(1) monocytes, (2) precursors of both the granulocytic and themonocytic lineages, i.e. promyelocytes, myelocytes as well aspromonocytes, and (3) even early progeny of hematopoieticstem cells. This latter population from BM day 0 was small andconsisted of large ring cells in which the wide karyoplasmicring was sometimes interrupted by a narrower segment. Some ofthem could therefore be mistaken for PMN-like ring cells, andthis may be the reason why a similar percentage of cells wascounted as GR-12 PMN-like ring cells on immunohistochemis-try where fixation does not always sufficiently preserve a goodmorphology. Yet they can be assigned morphologically toMNC-like ring cells because of their size and their broadkaryoplasmic sections. This morphological assignment was

supported by the fact that most of these cells presented no oronly faint expression of lineage markers and were thus notPMN.

In most previous studies ring cells were classified as beingexclusively PMN, eosinophils, or, if at all, precursors of PMN[1–8, 15]. In contrast, it was not generally accepted that ringcells also encompass monocytes. In some studies on murine BMcells, ring cells have been mentioned among cells that wereable to phagocytose plastic microspheres [9] and to bind CSF[10]. The only study that has shown that certain ring cells sharecharacteristics of monocytes, such as adherence, longer sur-vival time in culture, and pattern of enzymatic activity, waspublished by Pels et al. [11]. In addition to Pels et al. weelaborated further criteria that proved that ring cells encompassmonocytes and that were also able to separate PMN-like ringcells from MNC-like ring cells (high density of several organ-elles, expression of surface markers for monocytes/macro-phages, reduced apoptotic rate, high phagocytosis of L. major).MNC-like ring cells do not appear to contain mature BMmacrophages because they have almost disappeared after 1week of BM culture. However, by applying the spectrum ofcriteria we detected that the so-called MNC-like ring cellsadditionally contain different precursor forms. Although mu-rine ring cells had been suggested by some authors to begranulocyte precursors [1–4, 6–8, 69, 70], they were notconsidered so far to contain precursors of monocytes. We foundthat some MNC-like ring cells stained with ER-MP12, a markerexpressed on GM-CFC, M-CFC, and on early macrophageprecursors until the promonocytic stage [32, 44, 45]. When weanalyzed distribution of peroxidase activity, we found cellsexhibiting PO activity as described for promyelocytes andpromonocytes [7, 24, 26, 27, 29, 32, 66]. These ultrastructuralcriteria established by Bainton and collaborators [26, 27, 29]have been derived from human leukocytes. Our study showedthat they are also compatible with the murine system. We areaware, however, that there are certain restrictions to thesecriteria when distinguishing between promonocytes and pro-myelocytes as well as between myelocytes and PMN [24, 26,27, 29] because the presence of one negative granule (whichwould exclude the promyelocyte) cannot always be ruled outwith certainty and because, e.g., nucleoli are sometimes, butnot always, present in myelocytes. This fact demonstrates thatthe distinction between cells of the granulocytic and monocyticlineage is not always rigid. It encourages the combination ofdifferent methods for characterization of leukocytes, even if thecombined use of ultrastructural as well as immuno- andenzymohistochemical procedures still remains unable to unam-biguously assign all myeloic cells to either monocytes orgranulocytes.

The precursor types among ring cells also comprised formspreceding the promonocyte or promyelocyte because theyincluded cells with low or absent expression of (leukocytic)lineage markers. When we performed sorter analysis in anattempt to separate PMN-like ring cells from MNC-like ringcells by distinct expression of surface markers (e.g., low or highdensity of GR-1), we found some ring cells among those cellsthat were negative or only faintly positive for several lineagemarkers. Further analysis revealed that they encompass

TABLE 7. Percentages of FACS-Analyzed and Sorted CellFractions of Murine BM Cells Day 2

Region Number of cells (%) Mean Y Combination of antibodies

Tot. Stats. 100.00 8.91 3.09 8.4 GR-1neg/BM82

2 31.97 6.5 GR-1low/BM82

3 59.97 4.6 GR-1high/BM82

4 0.55 81.2 GR-1neg/BM81

5 3.7 79.7 GR-1low/BM81

6 0.78 56.1 GR-1high/BM81

See Figure 8.


CD382, sca-11 and s-kit1 cells, which, according to a recentstudy, would present a more committed progeny of hematopoi-etic stem cells [58]. This is remarkable because the fraction oflineagenegative/low ring cells was enriched for the particularMNC-like ring cell with slender nucleus (25% of sorted cells)and did not contain MNC-like ring cells with smooth nuclearcontour and PMN-like ring cells. We obtained a slightly higherpercentage of lineagenegative/low BM cells than Randall et al. [58],but this may be due to differences between inbred mousestrains (Randall et al. used C57Bl/6 mice) and to variations inthe selected gates because the appearance of lineage markerson the cell surface is a very gradual process. A closer analysisof the development of these progenitor cells was beyond thescope of this study, but with L cell-conditioned medium theyyielded monocytic cells.

Although the majority of ring cells in BM belong to thegranulocytic lineage, their identification as macrophages, mono-cytes, and various precursors is an important finding. Quantita-tive or even qualitative studies on murine leukocytes could beflawed as long as they do not leave the concept that ring cellsexclusively belong to the granulocytic lineage.

Because we found that one type of ring cells even presentsprogeny of hematopoietic stem cells we suggest that cells withring shaped nuclei present a myeloic cell lineage parallel tomyeloic cell lines with closed nuclei. At this point we cannotaccurately state at which point in the development from theomnipotent stem cell ring cells first occur nor if they are able todevelop into cells with closed nuclei.

One reason for the lack of detailed studies on the monocyticnature of some murine ring cells may originate from thelong-standing discussion about the origin and kinetics ofmacrophages. According to the concept of the mononuclearphagocyte system (MPS) by van Furth et al. [71] all macro-phages from BM, circulation, and different tissues belong to onecell line originating in the BM and differentiating there intomonocytes. In contrast, other authors have suggested thatmacrophages (1) may derive from more than one progenitor cellor lineage and (2) may additionally derive from cells other thanthe typical kidney-shaped blood monocyte, e.g., from proliferat-ing local macrophages [72–74]. Leder found cells in BM andPB with strong activity of both unspecific esterase and chloroac-etate esterase and concluded that they may present an interme-diate form between the granulocytic (promyelocyte) and mono-cytic (monocyte) lineage [74, 75]. Metcalf also had evidencethat murine macrophages in vitro could arise by transformationof granulocyte precursors [69, 70]. We have found someMNC-like ring cells (and mononuclear cells with closed nuclei)with the enzymo- and immunohistochemical characteristics ofboth granulocytes and monocytes. Although sorter analysisrevealed that some of the double-positive cells were GR-1low

and not GR-1high as granulocytes, it remains a fact that somecells with closed and especially with ring-shaped nuclei cannotbe clearly assigned to the granulocytic or monocytic lineage byconventional methods. These cells could present so-calledbipotential cells with common characteristics of monocytes andgranulocytes beyond the stage of the GM-CFU myelomonocyticstem cell [69, 70, 74, 75].

On kinetic studies on murine models of inflammation one

striking finding is a rather sharp rise in the percentage andabsolute numbers of macrophages in the infiltrate [12]. It hasbeen explained by an increased rate of immigration viamacrophage-specific endothelial adhesion molecules and che-mokines [76] and by rising numbers of monocytes in PB(monocytosis). However, the rise of monocytic cells in theinfiltrate precedes monocytosis in murine blood. An additionalexplanation could therefore be that the early infiltrate containsmacrophages that proliferate locally or that at this point havenot been recognized as macrophages, but as, for example,granulocytes, such as MNC-like ring cells. We have shown thatMNC-like ring cells in PB have retained their capacity toproliferate.

At this point we can only speculate whether the ring-shapednucleus of murine myeloic cells is of biological relevance. Thering-shaped nucleus disposes of a larger surface and could thusallow a higher rate of interactions between nucleus andcytoplasm, e.g., with regard to exchange of metabolites orsignals. The cytoplasmic center also could offer a certaindegree of compartmentalization because we have observed thatseveral organelles are concentrated within it. Because theincidence of ring cells in the human system is increased inpatients with different myeloproliferative disorders [19, 77] orwith growth factor therapy for neutropenia [78], it has beensuggested that their occurrence in humans may be consideredas a marker for increased abnormal granulopoiesis associatedwith cytoplasmic immaturity [19, 78].

Our extensive analysis of murine leukocytes with ring-shaped nuclei has yielded a characterization of the differentforms of these cells and has also revived questions about theprovenance of monocytes/macrophages in blood and in inflam-matory infiltrates.

ACKNOWLEDGMENTS

This work was supported by the Deutsche Forschungsgemein-schaft, SFB 293, project A8 (Su). We thank Ms. K. Fischer, Ms.S. Merfeld, and Ms. B. Hillmann for excellent technicalassistance as well as Dr. M. Klempt, Dr. F. Schonlau, and Ms. L.Gassen for valuable discussions.

REFERENCES

1. Schermer, S. (1958) Die weisse Maus. Die Blutmorphologie der Laboratori-umstiere, (S. S. Leipzig, J. A. Barth, eds.) Berlin: Verlag, 61–74.

2. Bradley, T. R., Metcalf, D. (1966) The growth of mouse bone marrow cellsin vitro. Austr. J. Exp. Biol. Med. Sci. 44, 287–300.

3. Hardy, J. (1967) Haematology of Rats and Mice. In Pathology ofLaboratory Rats and Mice (E. Cotchin and F. J. C. Roe, eds.) Oxford, UK:Blackwell Scientific 501–536.

4. Beran, M., Tribukait, B. (1971) The quantitative bone marrow and spleencomposition in male NMRI and CBA mice. Acta Haematol. 45, 55–64.

5. Gardner, M. V. (1974) The blood picture of normal laboratory animals. Areview of the literature 1936–1946. J. Franklin Inst. 243, 172–176.

6. Goud, T. J. L. M. (1975) Identification and Characterization of theMonoblast. Leiden, Thesis.

7. Goud, T. J. L. M., Schotte, C., van Furth, R. (1975) Identification andcharacterization of the monoblast in mononuclear phagocyte coloniesgrown in vitro. J. Exp. Med. 142, 1180–1199.

8. Schalm, O. W., Jain, N. C., Carroll, E. J. (1975) Veterinary Hematology(O. W. Schalm, N. C. Jain, E. J. Carroll, eds.) Philadelphia: Lea andFebiger, 235–258.


9. Buhles, W. C. (1979) Studies on mononuclear phagocyte progenitor cells:morphology of cells and colonies in liquid culture of mouse bone marrow. J.Reticuloendothel. Soc. 25, 363–378.

10. Guilbert, L. J., Stanley, E. R. (1980) Specific interaction of murinecolony-stimulating factor with mononuclear phagocytic cells. J. Cell Biol.85, 153–159.

11. Pels, E., de Groot, J. W., Mullink, R., van Unnik, J. A. M., den Otter, W.(1980) Identification of two different types of mouse peritoneal exudatecells with ring-shaped nuclei. J. Reticuloendothel. Soc. 27, 367–375.

12. Sunderkotter, C., Beil, W., Roth, J., Sorg, C. (1991) Cellular eventsassociated with inflammatory angiogenesis in the mouse cornea. Am. J.Pathol. 138, 931–939.

13. Sunderkotter, C., Kunz, M., Steinbrink, K., Meinardus-Hager, G., Goebeler,M., Bildau, H., Sorg, C. (1993) Resistance of mice to experimentalleishmaniasis is associated with more rapid appearance of mature macro-phages in vitro and in vivo. J. Immunol. 151, 4891–4901.

14. Roth, J., Sunderkotter, C., Goebeler, M., Gutwald, J., Sorg, C. (1992)Expression of the calcium binding proteins MRP8 and MRP14 inexperimental contact dermatitis. Int. Arch. Allergy Immunol. 98, 140–145.

15. Vogel, M. (1974) The femoral bone marrow cells of the albino rat. Am. J.Med. Sci. 213, 456–462.

16. Peichev, M. (1986) Ring cells in infectious mononucleosis. Br. J. Haema-tol. 62, 397–398.

17. Stavem, P., Hjort, P. F., Vogt, E., van der Hagen, C. B. (1969) Ring-shapednuclei of granulocytes in a patient with acute erythroleukaemia. Scand. J.Haematol. 6, 31–32.

18. Langenhuijsen, M. M. A. C. (1984) Neutrophils with ring-shaped nuclei inmyeloproliferative disease. Br. J. Haematol. 58, 227–230.

19. Kanoh, T., Saigo, K., Yamagishi, M. (1986) Neutrophils with ring-shapednuclei in chronic neutrophilic leukemia. Am. J. Clin. Pathol. 860,748–751.

20. Knecht, H., Eichhorn, P., Streuli, R. A. (1985) Granulocytes withring-shaped nuclei in severe alcoholism. Acta Haematol. 73, 184.

21. Neumann, C., Sorg, C. (1980) Sequential expression of functions duringmacrophage differentiation in murine bone marrow liquid cultures. Eur. J.Immunol. 10, 834–840.

22. Kunz, M., Meinardus-Hager, G., Goebeler, M., Mahnke, K., Sunderkotter,C., Sorg, C. (1993) 8E7: a new monoclonal antibody against Leishmaniamajor parasites. Arch. Dermatol. Res. 285, 74 (Abstract).

23. Spurr, A. R. (1969) A low-viscosity epoxy resin embedding medium forelectron microscopy. J. Ultrastruct. Res. 26, 31–43.

24. van Furth, R., Hirsch, J. G., Fedorko, M. E. (1970) Morphology andperoxidase cytochemistry of mouse promonocytes, monocytes, and macro-phages. J. Exp. Med. 132, 794–812.

25. Florey, L. (1970) General Pathology (L. Florey, ed.) Philadelphia: Saun-ders.

26. Bainton, D. F., Ullyot, J. L., Farquhar, M. G. (1971) The development ofneutrophilic polymorphonuclear leukocytes in human bone marrow. J. Exp.Med. 134, 907–934.

27. Nichols, B. A., Bainton, D. F. (1975) Ultrastructure and cytochemistry ofmononuclear phagocytes. In Mononuclear Phagocytes in Immunity, Infec-tion, and Pathology (R. van Furth, ed.) Oxford, UK: Blackwell Scientific,17–55.

28. Cohn, Z. A. (1968) The structure and function of monocytes andmacrophages. Adv. Immunol. 9, 163–214.

29. Bainton, D. F. (1993) Neutrophilic leukocyte granules: From structure tofunction. Adv. Exp. Med. Biol. 336, 17–33.

30. Cross, P. C., Mercer, K. L. (1993) Blood In Cell and Tissue Ultrastructure. AFunctional Perspective (P. C. Cross, K. L. Mercer, eds.) New York: Freeman,163–187.

31. Bloom, W., Fawcett, D. W. (1994) A Textbook of Histology (W. Bloom, D. W.Fawcett, eds.) New York: Chapman and Hall, 108–132, 234–259.

32. Morioka, Y., Naito, M., Sato, T., Takahashi, K. (1994) Immunophenotypicand ultrastructural heterogeneity of macrophage differentiation in bonemarrow and fetal hematopoiesis of mouse in vitro and in vivo. J. Leukoc.Biol. 55, 642–651.

33. Robbins, D. (1971) Fine structural cytochemical localization of peroxidaseactivity in rat peritoneal cells, eosinophils and mast cells. J. Histochem.Cytochem. 19, 571–575.

34. Aten, J. A., Bakker, P. J. M., Stap, J., Boschman, G. A., Veenhof, C. H. N.(1992) DNA double labeling with IdUrd and CldUrd for spatial andtemporal analysis of cell proliferation and DNA replication. Histochem. J.24, 251–259.

35. Yam, L. T., Li, C. Y., Crosby, W. H. (1971) Cytochemical identification ofmonocytes and granulocytes. Am. J. Clin. Pathol. 55, 283–290.

36. Austyn, J. M., Gordon, S. (1981) F4/80, a monoclonal antibody directedspecifically against the mouse macrophage. Eur. J. Immunol. 11, 805–815.

37. Hirsch, S., Austyn, J. M., Gordon, S. (1981) Expression of the macrophage-specific antigen F4/80 during differentiation of mouse bone marrow cells inculture. J. Exp. Med. 154, 713–725.

38. Hume, D. A., Robinson, A. P., Macpherson, G. G., Gordon, S. (1983) Themononuclear phagocyte system of the mouse defined by immunohistochemi-cal localization of antigen F4/80. J. Exp. Med. 158, 1522–1536.

39. McGarry, M. P., Stewart, C. C. (1991) Murine eosinophil granulocytes bindthe murine macrophage-monocyte specific monoclonal antibody F4/80. J.Leukoc. Biol. 50, 471–478.

40. Leenen, P. J. M., de Bruijn, M. F. T. R., Voerman, J. S. A., Campbell, P. A.,van Ewijk, W. (1994) Markers of mouse macrophage development detectedby monoclonal antibodies. J. Immunol. Meth. 174, 5–19.

41. de Jong, J. P., Voerman, J. S. A., van der Sluijs-Gelling, A. J., Willemsen,R., Ploemacher, R. E. (1994) A monoclonal antibody (ER-HR3) againstmurine macrophages. I. Ontogeny, distribution and enzyme histochemicalcharacterization of ER-HR3-positive cells. Cell Tissue Res. 275, 567–576.

42. de Jong, J. P., Leenen, P. J. M., Voerman. J. S. A., van der Sluijs-Gelling,A. J., Ploemacher, R. E. (1994) A monoclonal antibody (ER-HR3) againstmurine macrophages. II. Biochemical and functional aspects of theER-HR3 antigen. Cell Tissue Res. 275, 577–585.

43. Hansen, A. S., Noren, O., Sjostrom, H., Werdelin, O. (1993) A mouseaminopeptidase N is a marker for antigen-presenting cells and appears tobe co-expressed with major histocompatibility complex class II molecules.Eur. J. Immunol. 23, 2358–2364.

44. Leenen, P. J. M., Slieker, W. A. T., Melis, M., van Ewijk, W. (1990) Murinemacrophage precursor characterization. I. Production, phenotype anddifferentiation of macrophage precursor hybrids. Eur. J. Immunol. 20,15–25.

45. Leenen, P. J. M., Melis, M., Slieker, W. A. T., van Ewijk, W. (1990) Murinemacrophage precursor characterization. II. Monoclonal antibodies againstmacrophage precursor antigens. Eur. J. Immunol. 20, 27–34.

46. Slieker, W. A. T., van der Loo, J. C. M., de Rijk-de Bruijn, M. F. T. R.,Godfrey, D. I., Leenen, P. J. M., van Ewijk, W. (1993) ER-MP 12 antigen, anew cell surface marker on mouse bone marrow cells with thymus-repopulating ability: II. Thymus-homing ability and phenotypic character-ization of ER-MP 12-positive bone marrow cells. Int. Immunol. 5,1099–1107.

47. de Bruijn, M. F. T. R., Slieker, W. A. T., van der Loo, J. C. M., Voerman,J. S. A., van Ewijk, W., Leenen, P. J. M. (1994) Distinct mouse bone marrowmacrophage precursors identified by differential expression of ER-MP 12and ER-MP 20 antigens. Eur. J. Immunol. 24, 2279–2284.

48. van der Loo, J. C. M., Slieker, W. A. T., Kieboom, D., Ploemacher, R. E.(1995) Identification of hematopoietic stem cell subsets on the basis oftheir primitiveness using antibody ER-MP 12. Blood 85, 952–962.

49. Ling, V., Luxenberg, D., Wang, J., Nickbarg, E., Leenen, P. J. M., Neben,S., Kobayashi, M. (1997) Structural identification of the hematopoieticprogenitor antigen ER-MP12 as the vascular endothelial adhesion mol-ecule PECAM-1 (CD31). Eur. J. Immunol. 27, 509–514.

50. van Vliet, E., Melis, M., van Ewijk, W. (1984) Monoclonal antibodies tostromal cell types of the mouse thymus. Eur. J. Immunol. 14, 524–529.

51. Hestdal, K., Ruscetti, F. W., Ihle, J. N., Jacobsen, S. E. W., Dubois, C. M.,Kopp, W. C., Longo, D. L., Keller, J. R. (1991) Characterization andregulation of RB6-8C5 antigen expression on murine bone marrow cells. J.Immunol. 147, 22–28.

52. Fleming, T. J., Fleming, M. L., Malek, T. R. (1993) Selective expression ofLy-6G on myeloid lineage cells in mouse bone marrow. J. Immunol. 151,2399–2408.

53. Springer, T., Galfre, G., Secher, D. S., Milstein, C. (1979) Mac-1: amacrophage differentiation antigen identified by monoclonal antibody. Eur.J. Immunol. 9, 301–306.

54. Ho, M-K., Springer, T. A. (1982) MAC-1 antigen: Quantitative expressionin macrophage populations and tissues, and immunofluorescent localiza-tion in spleen. J. Immunol. 128, 2281–2286.

55. Beller, D. I., Springer, T. A., Schreiber, R. D. (1982) Anti-Mac-1selectively inhibits the mouse and human type three complement receptor.J. Exp. Med. 156, 1000–1009.

56. Rosen, H., Gordon, S. (1987) Monoclonal antibody to the murine type 3complement receptor inhibits adhesion of myelomonocytic cells in vitroand inflammatory cell recruitment in vivo. J. Exp. Med. 166, 1685–1701.

57. Spangrude, G. J., Heimfeld, S., Weissman, I. L. (1988) Purification andcharacterisation of mouse hematopoietic stem cells. Science 241, 58–62.

58. Randall, T. R., Lund, F. E., Howard, M. C., Weissman, I. L. (1996)Expression of murine CD38 defines a population of long-term reconstitut-ing hematopoietic stem cells. Blood 87, 4057–4067.

59. Koller, C. A., King, G. W., Hurtubise, P. E., Sagone, A. L., LoBuglio, A. F.(1973) Characterization of glass adherent human mononuclear cells. J.Immunol. 111, 1610–1612.


60. Johnson, W. D., Mei, B., Cohn, Z. A. (1977) The separation, long-termcultivation, and maturation of the human monocyte. J. Exp. Med. 146,1613–1626.

61. Mosier, D. E. (1981) Separation of murine macrophages by adherence tosolid substrates. In Methods for Studying Mononuclear Phagocytes (D. O.Adams, P. J. Edelson, H. S. Koren, eds.) New York: Academic, 179–186.

62. Savill, J. S., Wyllie, A. H., Henson, J. E., Walport, M. J., Henson, P. M.,Haslett, C. Macrophage phagocytosis of aging neutrophils in inflammation.Programmed cell death in the neutrophil leads to recognition by macro-phages. J. Clin. Invest. 83, 865–875.

63. Whyte, M. K. B., Meagher, L. C., MacDermot, J., Haslett, C. (1993)Impairment of function in aging neutrophils is associated with apoptosis. J.Immunol. 150, 5124–5134.

64. van der Vliet, H. J. J., Wever, P. C., van Diepen, F. N. J., Yong, S-L., TenBerge, I. J. M. (1997) Quantification of Bax/Bcl-2 ratios in peripheral bloodlymphocytes, monocytes and granulocytes and their relation to susceptibil-ity to anti-Fas (anti-CD95)-induced apoptosis. Clin. Exp. Immunol. 110,324–328.

65. van Furth, R., Diesselhoff-den Dulk, M. M. C., Raeburn, J. A., van Zwet,T. L., Crofton, R., Bluss van Oud Alblas, A. (1980) Characteristics, origin,and kinetics of human and murine mononuclear phagocytes. MononuclearPhagocytes. Functional Aspects (R. van Furth, ed.) The Hague: Martinus-Nijhoff, 279–298.

66. van der Meer, J. W. M. (1988) In vitro cultures of proliferating mononuclearphagocytes. In The Alveolar Macrophage. Local Immunity, Vol. 4. (C. Sorg,ed.) Munster: Regensberg und Biermann, 14–22.

67. Lagasse, E., Weissman, I. L. (1996) Flow cytometric identification ofmurine neutrophils and monocytes. J. Immunol. Meth. 197, 139–150.

68. Leenen, P. J. M., Melis, M., Kraal, G., Hoogeveen, A. T., van Ewijk, W.

(1992) The monoclonal antibody ER-BMDM1 recognizes a macrophageand dentritic cell differentiation antigen with aminopeptidase activity. Eur.J. Immunol. 22, 1567–1572.

69. Metcalf, D. (1969) Studies on colony formation in vitro by mouse bonemarrow cells. I. Continuous cluster formation and relation of clusters tocolonies. J. Cell Physiol. 74, 323–332.

70. Metcalf, D. (1971) Transformation of granulocytes to macrophages in bonemarrow colonies in vitro. J. Cell Physiol. 77, 277–280.

71. van Furth, R., Cohn, Z. A., Hirsch, J. G., Humphry, J. H., Spector, W. G.,Langevoort, H. L. (1972) The mononuclear phagocyte system: a newclassification of macrophages, monocytes and their precursor. Bull. WorldHealth Org. 46, 845–852.

72. Daems, W. T., De Bakker, J. M. (1982) Do resident macrophagesproliferate? Immunobiol. 161, 204–211.

73. Volkman, A., Chang, N. C., Strausbauch, P. H., Morahan, P. S. (1983)Differential effects of chronic monocyte depletion on macrophage popula-tions. Lab. Invest. 49, 291–298.

74. Leder, L-D. (1967) The origin of blood monocytes and macrophages. Blut16, 86–98.

75. Leder, L-D. (1967) Der Blutmonocyt (Leder, L-D., ed.) Berlin: Springer-Verlag.

76. Springer, T. A. (1994) Traffic signals for lymphocyte recirculation andleukocyte emigration: The multistep paradigm. Cell 76, 301–314.

77. Kanoh, T. (1991) Ring neutrophils in plasma cell dyscrasia. Arch. Pathol.Lab. Med. 115, 178–180.

78. Schmitz, L. L., McClure, J. S., Litz, C. E., Dayton, V., Weisdorf, D. J.,Parkin, J. L., Brunning, R. D. (1994) Morphologic and quantitativechanges in blood and marrow cells following growth factor therapy. Am. J.Clin. Pathol. 101, 67–75.


Murine leukocytes with ring-shaped nuclei include granulocytes, monocytes and their precursors

Documents