journal of immunological methods -...

Journal of Immunological Methods 385 (2012) 1–14

Contents lists available at SciVerse ScienceDirect

Journal of Immunological Methods

j ourna l homepage: www.e lsev ie r .com/ locate / j im

Research paper

Pure populations of murine macrophages from cultured embryonic stemcells. Application to studies of chemotaxis and apoptotic cell clearance

Lihui Zhuang a, John D. Pound a, Jorine J.L.P. Willems a, A. Helen Taylor a,b,Lesley M. Forrester a,b, Christopher D. Gregory a,⁎a University of Edinburgh/MRC Centre for Inflammation Research, Queen's Medical Research Institute, 47 Little France Crescent, Edinburgh, EH16 2TJ, UKb MRC Centre for Regenerative Medicine, Scottish Centre for Regenerative Medicine, University of Edinburgh, 5 Little France Road, Edinburgh, EH16 4UU, UK

a r t i c l e i n f o

Abbreviations: BMDM, bone marrow-derived macrtioned medium;DC, dendritic cell;EB, embryoid bodievirus;ESC, embryonic stem cell;ESDM, embryonicmacrophages;LIF, leukaemia inhibitory factor⁎ Corresponding author.

E-mail address: [email protected] (C.D. Grego

0022-1759/$ – see front matter © 2012 Elsevier B.V. Adoi:10.1016/j.jim.2012.06.008

a b s t r a c t

Article history:Received 5 April 2012Received in revised form 10 June 2012Accepted 11 June 2012Available online 18 June 2012

Embryonic stem cells provide a potentially convenient source ofmacrophages in the laboratory.Given the propensity of macrophages for plasticity in phenotype and function, standardisedculture and differentiation protocols are required to ensure consistency in population outputand activity in functional assays. Here we detail the development of an optimised cultureprotocol for the production of murine embryonic stem cell-derived macrophages (ESDM). Thisprotocol provides improved yields of ESDM and we demonstrate that the cells are suitable forapplication to the study of macrophage responses to apoptotic cells. ESDM so produced were ofhigher purity than commonly used primary macrophage preparations and were functional inchemotaxis assays and in phagocytosis of apoptotic cells. Maturation of ESDM was found to beassociated with reduced capacity for directed migration and increased capacity for phagocyticclearance of apoptotic cells. These results show ESDM to be functionally active in sequentialphases of interaction with apoptotic cells and establish these macrophage populations as usefulmodels for further study of molecular mechanisms underlying the recognition and removal ofapoptotic cells.

© 2012 Elsevier B.V. All rights reserved.

Keywords:Embryonic stem cellsMacrophagesEmbryonic stem cell-derived macrophagesChemotaxisApoptosisApoptotic cells

1. Introduction

Macrophages play diverse roles in tissue homeostasis,development, immunity, inflammation, tissue repair and regen-eration. Such functional diversity is reflected in the broadlydiverse activation states that macrophages display in responseto their environment. Current understanding of macrophagebiology has been generated to a significant extent fromreductionist investigations ofmacrophages and theirmonocyteprecursors in isolation in vitro. Two main approaches havebeen used: (a) isolation of primary macrophages or theirprecursors followed by culture tomature and/or activate them,

ophages;CM, condi-s;EBV, Epstein–Barrstem cell-derived

ry).

ll rights reserved.

and (b) culture of continuous macrophage-lineage cell lines,such as the widely used J774 and RAW 264 lines, derived frommalignant tumours (Ralph and Nakoinz, 1975; Raschke et al.,1978). Each of these approaches has its obvious advantages anddisadvantages. Acquisition of primary cells requires access toliving organisms, often disruptive isolation procedures, andproduces issues over purity of cell populations and limitationsin cell numbers. Furthermore, primary cells are inherentlydifficult to manipulate genetically which limits their use forstudying the molecular mechanisms associated with theirfunction. By contrast, macrophage cell lines provide suppliesof limitless numbers of pure cell populations that can be readilymanipulated but as they are transformed cells they are farremoved from the normal state. They do, however retaincertain features of normal macrophages and their precursorsand consequently are widely used in macrophage research.

Rapid progress in embryonic stem cell (ESC) culture andthe development of efficient differentiation strategies over

http://dx.doi.org/10.1016/j.jim.2012.06.008

mailto:[email protected]

http://dx.doi.org/10.1016/j.jim.2012.06.008

http://www.sciencedirect.com/science/journal/00221759

2 L. Zhuang et al. / Journal of Immunological Methods 385 (2012) 1–14

recent years has provided access to another in vitro source ofmacrophages. Derivation of macrophages and their precursorsfrom ESCs, at least in part, combines the advantages of tumourcell lines – including avoidance of in vivo access and isolationproblems – with the preferred option to study essentiallynormal cells, rather than their neoplastic, transformed coun-terparts. ESCs can be expanded in large numbers in vitro andare highly amenable to sophisticated genetic manipulationtechnology so they have the potential to provide highly purepopulations of wild-type or genetically manipulated macro-phages and their precursors representative of tissue-specificsettings. For example, murine ESC-derived macrophages(ESDM) have been applied to the investigation of macrophageactivities in atherosclerotic lesions (Moore et al., 1998) andmicroglial cells (Napoli et al., 2009) as well as generalmacrophage physiology (Chawla et al., 2001; Odegaard et al.,2007). Similarly, human ES cells and haemopoietic stem cellshave been used as productive sources of human macrophagesformechanistic studies (Karlsson et al., 2008; Subramanian et al.,2009; Way et al., 2009). Furthermore, cells apparently compa-rable with bone marrow-derived macrophages may be derivedin vitro from induced pluripotent stem cells (Way et al., 2009).

Given the plasticity of macrophages and the scope forminor changes in culture conditions to change profoundlytheir activation status, each methodological approach re-quires optimisation in order to achieve reproducibility inphenotype and function of the output cells in appropriateapplications. Here we describe an optimised and convenientmethod for the production of supplies of murine macro-phages and their precursors from ESCs with particularconsideration of their application to in vitro studies ofmacrophage chemotaxis and phagocytic clearance of apopto-tic cells — a fundamentally important macrophage activity.We compare the purity of these ESDM with bone marrow-derived and peritoneal macrophage populations, the primarycells most often used in murine macrophage studies. Ourfindings indicate that ESDM can be produced in requirednumbers at high purity and can be applied effectively toin vitro studies of migration and apoptotic-cell clearance.Furthermore the results show that the migratory andphagocytic responses of these cells are reciprocally related,being dependent upon their maturation state: ‘younger’, lessmature cells responded more vigorously than more matureESDM to chemotactic stimuli whereas more mature ESDMwere more active in phagocytosis than their precursors.These results establish ESDM as a potentially valuable in vitromodel system for studies of detection and removal ofapoptotic cells by mononuclear phagocytes.

2. Materials and methods

2.1. Cell lines

The murine E14 ESC line (Handyside et al., 1989) wasused for most of the investigations. For some experiments theESC line, GFP#7a, which expresses enhanced green fluores-cent protein (eGFP) constitutively (Gilchrist et al., 2003), wasused. Both lines behaved similarly in the macrophagedifferentiation protocols described. ESC lines were main-tained as undifferentiated cells in 0.1% gelatin-coated tissueculture flasks in GMEM (Life Technologies Ltd, Paisley, UK)

supplemented with 10% (v/v) batch tested foetal bovineserum (FBS, Globepharm Ltd, Guildford, Surrey, UK), 1% (v/v)non-essential amino acids (Life Technologies), 2mM L-glutamine (Life Technologies), 1mM sodium pyruvate (LifeTechnologies), 0.1mM 2-mercaptoethanol (Sigma-Aldrich,Poole, Dorset, UK) and 100 U/mL leukaemia inhibitory factor(LIF) that had been produced from Lif-transfected COS-7 cellsas previously described (Jackson et al., 2010). We designatethe media used to maintain ESCs in their undifferentiated,self-renewing state as GMEMSR. The EBV-negative humanBurkitt lymphoma (BL) cell line, BL2 and its Bcl-2-expressingcounterpart transfectant line, BL2-bcl-2, were derived andcultured as described (Wang et al., 1996).

2.2. Macrophage differentiation

Macrophage differentiation of ESCs was performed in amedium we termed ESDMDiff. This medium is similar toGMEMSR except that (1) LIF was excluded from the mediumand (2) the FCS used for differentiation had been pre-screenedfor optimal haematopoietic differentiation (Jackson et al., 2010).Furthermore, ESDMDiff contained recombinant IL-3 (Stem CellTechnologies, Genoble, France) and L929-derived macrophagecolony-stimulating factor (M-CSF; CSF-1). L929 fibroblasts werecultured in D-MEM: F12 with Glutamax (Life Technologies)supplemented with 10% FCS (Biosera Ltd, Ringmer, East Sussex,UK), 100IU/mL penicillin (PAA Laboratories Ltd, Yeovil, Somer-set,UK) and 100μg/mL streptomycin (PAA Laboratories). Con-ditioned medium (CM) containing CSF-1 was harvested fromadherent L929 cells 3days following confluency, filteredthrough a 0.22μm membrane to remove cell debris and storedat −20°C as described (Burgess et al., 1985). L929 CM wasincluded in ESDMDiff at a final concentration of 15% (v/v).

The culture media used for ESC culture and macrophagedifferentiation are summarized in Table 1.

2.3. Preparation of primary macrophages

Bone marrow-derived macrophages (BMDM) were pre-pared from Balb/c mouse femurs as described (Truman et al.,2004). Briefly, bone marrow was flushed into a 50mL conicaltube (BD Falcon) with DMEMsupp (DMEM: F12 with Glutamax(Life Technologies), 100IU/mL penicillin with 100μg/mL strep-tomycin, 20% (v/v) FBS (Biosera) and 10% (v/v) L929 cellconditioned medium) using a 10mL syringe (BD Plastipak) anda 26G needle (BDMicrolance). Cells were then passed through a23Gneedle and cultured in 20mLDMEMsupp in 95mmdiameterbacteriological grade Petri dishes at 37°C in humidified 5% CO2

in air. Culture medium was replaced after 2h and again after4days. Differentiated macrophages were used for experimentsafter a further 3days in culture. Peritoneal macrophages wereprepared by peritoneal lavage of a normal Balb/c mouse withPBS. Recovered cells were centrifuged, resuspended in 20mlDMEMsupp and cultured at 37°C in humidified 5% CO2 in air in95mm diameter bacteriological grade Petri dishes for 2h afterwhich non-adherent cells were removed by washing.

2.4. Detachment of macrophages

Cell culture mediumwas removed and macrophages werewashed twice with Hanks' BSS (without Ca2+/Mg2+/phenol

Table 1Summary of media used for ESC culture and for differentiation to ESDM.

Medium Basal Additives Application

GMEMSR GMEM 10% (v/v) FBS, 1% (v/v) non-essential amino acids, 2mM L-glutamine,1mM sodium pyruvate, 0.1mM 2-mercaptoethanol, 100U/mL LIF

ESC maintenance

ESDMDiff GMEM 10% (v/v) FBS (pre-screened for optimal haematopoietic differentiation),1% (v/v) non-essential amino acids, 2mM L-glutamine, 1mM sodium pyruvate,0.1mM 2-mercaptoethanol, 1ng/mL IL-3, 15% (v/v) L929 CM (CSF-1)

Differentiation of ESCs to EB producing non-adherentmacrophage progenitor cells

ESDMCult GMEM 10% (v/v) FBS (pre-screened for optimal haematopoietic differentiation),1% (v/v) non-essential amino acids, 2mM L-glutamine, 1mM sodium pyruvate,0.1mM 2-mercaptoethanol, 15% (v/v) L929 CM (CSF-1), 100IU/mL penicillin,100μg/mL streptomycin

ESDM maturation and growth

3L. Zhuang et al. / Journal of Immunological Methods 385 (2012) 1–14

red, PAA) and incubated in the detachment buffer (Hanks' BSS(without Ca2+/Mg2+/phenol red)) supplemented with 5mMEDTA (Fisher Scientific Ltd, Loughborough, UK) and 0.2% (w/v)low-endotoxin bovine serum albumin (BSA, PAA Laboratories)at 4°C for 20min before being detached with gentle scrapingusing a Cell Lifter (Corning, Ewloe, Flintshire, UK). Cells werewashed twice with Hanks' BSS (without Ca2+/Mg2+/phenolred) before use.

2.5. Cytocentrifugation and staining

Macrophages were detached, washed and blocked inDulbecco's PBS (without Ca2+/Mg2+/phenol red, PAA) with10% (v/v) normal mouse serum (NMS, Biosera) at 4°C for30min. After cytocentrifuging (300rpm, 5min, Cytospin, Ther-mo Shandon, Basingstoke, Hampshire, UK), the slideswere driedand fixed in acetone (Fisher) at 4°C for 10min. The slides wereair-dried again before further staining. For general morpholog-ical analysis, cells were stained with Reastain Quick-Diff(Reagena Ltd, Toivala, Finland) according to the manufacturer'sinstructions. Immunoperoxidase staining was performed usingrat anti-mouse CD68 (MCA1957, purified, working concentra-tion 5μg/mL), F4/80 (MCA497GA, purified, 2μg/mL) and CD11b(MCA711, purified, 2μg/mL). Isotype controls were rat IgG2a(MCA1212, purified, 5μg/mL) for CD68 and rat IgG2b(MCA1125, purified, 2μg/mL) for CD11b and F4/80. All primaryantibodies and isotype controls were from AbD Serotec(Kidlington, Oxfordshire, UK). Secondary antibody was bioti-nylated anti-rat IgG (H+L) (mouse absorbed) (BA-4001,2.5μg/mL, Vector Laboratories, Peterborough, UK). After beingwashed twice in PBS, slides were blocked in Protein Block,Serum-Free (Dako UK Ltd, Ely, Cambridgeshire, UK) for 10minat room temperature, incubated with primary antibodiesovernight at 4°C, followed by two washes in PBS and incubationwith secondary antibody for 30min at room temperature. Aftertwo more washes, Vector ABC reagent was added for 30min atroom temperature, and DAB chromagen (Dako) was added for5min at room temperature. Finally, the slides were washedwithdistilled water and counterstained in haematoxylin for 2.5min,washed in distilled water again and air-dried before mountingwith Histomount (Fisher Scientific). To quantify the percentageof positively stained cells, 10 non-overlapping fields from oneslide at 200× magnification under an Axioskop 2 (Carl Zeiss UKLtd, Welwyn Garden City, Hertfordshire, UK) microscope werechosen and the numbers of both positively and negativelystained cells were quantified per field. Data were expressed asthe mean percentage of positive cells per field (±SEM).

2.6. Flow cytometric analysis of macrophage phenotype

Macrophages were transferred to ice-cold Dulbecco's PBS(without Ca2+/Mg2+, PAA) containing 5% (v/v) normal goatserum (NGS, Biosera) and 10% (v/v) NMS for 10min to blockFc receptors before staining with antibodies against F4/80(0.5μg/mL, PE-conjugated, CaltagMedsystems Ltd, Buckingham,Buckinghamshire, UK), CD11b (1μg/mL, PE-conjugated, BDBioscience, BD U.K. Limited, Oxford, Oxfordshire, UK) andCD11c (2.5μg/mL, FITC-conjugated, eBioscience, Hatfield,Hertfordshire, UK). Staining of samples of BMDMwas analysedusing a FACSCalibur cytometer (Becton-Dickinson) while stain-ing of ESDM and peritoneal cells were analysed using an EpicsXL flow cytometer (Beckman Coulter (UK) Ltd, HighWycombe,Buckinghamshire). Flow cytometry list mode data were ana-lysed using Flowjo (Treestar Inc, Ashland, OR, USA).

2.7. Clonal density assay

In order to determine whether undifferentiated ESCs werepresent in cell suspensions, clonal density assays wereperformed. ESDM or undifferentiated E14 ESCs were seeded atdensities of 103, 104, 105 and 106 cells per well in 5mL, induplicate, onto 0.1% gelatin-coated 6-well plates in GMEMSR

(which is supplemented with LIF, Table 1). Cells were culturedfor 1day then the medium was changed to GMEMSR with orwithout LIF. After a further 5-day culture, the medium wasremoved and colonies were fixed and stained for alkalinephosphatase activity using the AP leukocyte kit (Sigma-Aldrich)according to the manufacturer's instructions. Stained colonieswere examined microscopically and scored as stem cell, mixed,or differentiated colonies according to their alkaline phospha-tase positivity.

2.8. Flow cytometric assay of phagocytosis of apoptotic cells

ESDM populations were tested for their ability to phagocy-tose apoptotic BL2 cells using a flow cytometric phagocytosisassay based on that described by Jersmann et al., 2003. BL2 cellswere washed twice in PBS, resuspended in serum-free RPMI1640 (20×106cells/mL), mixed with the far-red fluorescentdye DDAO-SE (1μL of dye per 10×106 cells, Invitrogen, LifeTechnologies) and incubated at 37°C for 15min. The cells werecentrifuged, resuspended at 4×106cells/mL in serum-freeRPMI 1640 and UV-irradiated (100mJ/cm2 UV-B) to induceapoptosis as described (Truman et al., 2008). UV-treated cellswere returned to culture at 37°C for 3h for apoptosis toproceed. The cells were then centrifuged and resuspended in


ESDMCult (Table 1) at 2×106cells/ml. Alternatively, BL2 cellswere washed twice in serum free RPMI 1640, and 20×106 cellswere stained with 4μL PKH26 Red Fluorescent Cell Linker dye(Sigma-Aldrich) followed by a 4-min incubation at 25°C. Thereaction was stopped by addition of 2mL FCS. Cells werecentrifuged and then washed 3 times in 10mL complete BLmedium(RPMI 1640, 10% (v/v) FBS (Biosera), 2mML-glutamine,100IU/mL penicillin, 100μg/mL streptomycin). PKH-26 stainedBL2 cells were thenwashed oncemore in serum-free RPMI 1640and resuspended at 5×106cells/mL in HypoThermosol (BioLifeSolutions, Bothell,WA, USA) in a polypropylene conical tube (BDFalcon) and incubated at 4°C for 18h. They were then washedtwice in serum-free RPMI and resuspended at 2×106cells/mL inESDMCult medium at 2×106cells/mL and cultured for 90min at37°C, 5% CO2 to allow apoptosis to occur. Apoptosiswas assessedby flow cytometric analysis of fluoresceinated annexin V bindingand propidium iodide uptake as described (Truman et al., 2008).

ESDM constitutively expressing eGFP were derived bydifferentiation of the ES cell line, GFP#7a (Gilchrist et al.,2003) and demonstrated a comparable efficiency of macro-phage differentiation to E14 ESCs (data not shown). Alterna-tively, after detachment, ESDM not expressing GFP and atvarious stages of maturation were stained with the PKH67Green Fluorescent Cell Linker dye (Sigma-Aldrich) using asimilar protocol to that described for PKH26 above. Cellswere resuspended at 4×105cells/mL in ESDMCult and 2×105

of the eGFP-expressing or PKH67-dyed cells in 0.5mL wereadded to each well of a 48-well plate (Nunc A/S, Roskilde,Denmark) and incubated overnight. Before the assay, themedium was removed from the ESDM by aspiration and cellswere washed gently with PBS before 0.5mL of BL2 cells(2×106 per mL, stained and induced to undergo apoptosis asdecribed above) was added. Phagocytosis was allowed toproceed for 0.5 or 1h at 37°C. ESDM alone and BL2 cells alonewere used as controls. All samples were tested in triplicate.After incubation, the medium was removed and wells werewashed gently with PBS. 0.5mL of trypsin/EDTA (PAALaboratories) was added to each well, followed by a 15minincubation at 37°C and then a 15min incubation on ice todetach ESDM. Finally, the cells were harvested from each wellby vigorous pipetting for analysis by flow cytometry.

2.9. Chemotaxis assay

Recombinant C5a, CCL2 and VEGF were obtained from R&DSystems Europe Ltd (Abingdon, Oxfordshire, UK); CCL5 wasfrom Prospec Bio (Rehovot, Israel). Chemotaxis was assayedusing Boyden chamber-type transwells of either 5μm or 8μmpore size (Corning BV Life Sciences, Amsterdam, Netherlands)placed inwells of 24-well plates. RPMI 1640 (Life Technologies)supplemented with 2mM L-glutamine, 100IU/mL penicillin,and 100μg/mL streptomycin was used as the chemotaxis assaymedium. 100μL input migratory cells were seeded in the upperchamber and 600μL chemoattractant-containing medium wasloaded in the lower chamber. The initial ESDM chemotaxisprotocol on which optimisation was based, was performed asfollows. The chemotaxis assay was performed with transwellsof 8μm membrane pore size. Recombinant human C5a (10ng/mL) was used as positive control and basal assay medium asnegative control. C5a and negative control were loaded intolower chambers and incubated with transwells at 37°C for 2h

to allow equilibration before loading ESDM. ESDM weredetached from Petri dishes, washed twice with Hanks' BSSand once in the assaymediumbefore resuspending in the assaymedium at 2×106/mL. ESDM (2×105 in 100μL) were added tothe upper chambers and were incubated at 37°C for 4h. At theend of the assay, non-migrated cells were removed from theupper surface of the transwells with cotton buds and thetranswells were fixedwith 100%methanol for 10min, air-driedand subsequently stained with Quick-Diff Blue for 20min. Thenumber ofmigrated cells on the underside of the transwell wasdetermined using an inverted microscope (Axiovert 25, CarlZeiss) and was calculated as the mean of the number ofmigrated cells per high power field (400×magnification) fromten random areas. Two changes to this protocol were found toenhance rigour of the assay: (1) ESDM were detached fromPetri dishes 1day prior to assay and cultured in suspension in20mL fresh ESDMCult medium in Teflon pots (Roland VetterLaborbedarf OHG, Ammerbuch, Germany) for 24h beforeharvesting and preparation for experiments; and (2) theequilibration step was dropped and ESDM and chemoattrac-tants were loaded simultaneously. In some experiments,supernatants from stressed, pre-apoptotic cells were used inchemotaxis assays. These were generated from BL2 cellsharvested from exponentially growing cultures. Cells werewashed with Dulbecco's PBS (PAA) twice, and withserum-free RPMI 1640 once. They were then resuspendedin serum free RPMI 1640 at 100×106cells/mL and culturedat 37°C for 1h. Supernatants were harvested, centrifuged at400g, 4 °C for 5min to remove cells and at 870g, 4 °C for10min to remove cell debris. High-density culture underthese conditions triggers synchronous apoptosis in 40–50%of BL2 cells.

3. Results

3.1. Optimisation of the cell culture system

Differentiation of macrophages from ESCs was initiated bythe formation of three-dimensional aggregates known asembryoid bodies (EB). Pilot experiments were performed tocompare the production of macrophages from EB that hadbeen generated in suspension culture with EB that had beengenerated in hanging drop culture (Jackson et al., 2010) or ina semi-solid medium culture system. We noted that the leastlabour-intensive system of classical aggregation in suspen-sion culture provided the most efficient yields of macro-phages (data not shown). Optimisation of the suspensionculture system was undertaken using the control E14 ESCline. We used two growth factors that are known to beimportant in macrophage development in vivo: interleukin 3(IL-3) plays a key role in the lineage commitment frompluripotent stem cell to myeloid-restricted progenitor andcan act on myeloid progenitors to promote their proliferation(Dorshkind, 1990). As IL-3-responsive progenitor cellsdifferentiate, they become responsive to CSF-1, which isrequired for the maintenance of the monocyte phenotypeand capable of supporting the differentiation and the long-term survival of macrophage-like cells (Young et al., 1990).Therefore, ESDMDiff, the medium used for ESC differentiationto EB, was supplemented with both IL-3 and CSF-1 forat least the first 10days of the differentiation protocol; once


macrophage progenitors had formed, the medium wassupplemented with CSF-1 alone (referred to as ESDMCult).Table 1 summarizes the differentiation and growth mediaused for the sequential phases of the ESDM culture protocol.

The optimised protocol for production of ESDM is summa-rized in Fig. 1. At the start of the differentiation procedure (day0), 6×105 trypsin-dissociated E14 ES cells were seeded insuspension culture in 20mL ESDMDiff (containing IL-3 andCSF-1, Table 1) in 95mmbacteriological-grade Petri dishes. ESCsstarted to differentiate and formed various sized EB thatincreased in size over the next few days. To avoid adherence ofthe larger aggregates, EB were gently dislodged and transferredto a new Petri dish on day 4 and day 6 and then on day 8, theywere transferred to a 95mm gelatin-coated tissue culture dish.EB adhered to the tissue culture plastic and over the next fewdays, non-adherent macrophage precursor cells began to bereleased into the medium. At day 10 the non-adherent cells inthemedium (termed ESDM-0) were harvested, centrifuged andre-suspended in 20mL ESDMCult (containing CSF-1, penicillinand streptomycin, Table 1) and plated onto a 95mm diameterbacteriological Petri dish. Over the next 2–7days (i.e. 12–17days

Fig. 1. Schematic outline of optimised protocol for production of macrophages frsuspension for 8days with CSF-1 and IL-3 to form embryoid bodies (EB) and then psupernatant of adherent EB which contains macrophage progenitor cells (ESDM-0)progenitor cells are further cultured for up to 7days either continuously as adhere1day in suspension in Teflon pots (ESDM-3 and ESDM-7 respectively). Since supernof ESCs can yield 2–4×106 ESDM, 12–24×106 ESDM can be harvested from one dishnumber of ESC dishes. Culture media for the various stages are indicated on the lef

from the start of the differentiation protocol) macrophage-likecells adhered to the plastic Petri dish and proliferated to form aconfluent monolayer. These adherent ESDM were harvested byusing cold detachment buffer and gentle scraping using a CellLifter. 20mL ESDMDiff were added to the original plate ofadherent EB and non-adherent cells could then be harvestedagain 2days later. Harvesting in this way could be repeatedevery 2days for up to a total of 20days (Fig. 1). Approximately2–4million macrophages can be produced from each harvestand, by multiple harvesting, one starting culture of 6×105 EScells can produce a total of 12 to 24million macrophages.Significantly higher numbers of macrophages can be gener-ated by simply increasing the number of starting dishes ofESCs (Fig. 1). The viability of ESDM after detachmentwas ~90%according to trypan blue exclusion (data not shown). For thestudies described here, ESDM-0 were either (1) matured inadherent culture for 7days and used directly after lifting(termed ESDM-Ad) or (2) matured in adherent culture for 2or 6days and used following subsequent suspension culturein Teflon pots (cells referred to as ESDM-3 and ESDM-7respectively) as summarized in Fig. 1.

om murine ESCs. ESCs (6×105 cells per 95mm Petri dish) are cultured inlated onto gelatinized dishes. EB become adherent after 2days. From day 10,is collected every 2days and plated onto 95mm Petri dishes, with CSF-1. Thent cells (ESDM-Ad) or for 2 or 6days as adherent cells followed by a furtheratant can be collected every 2days from day 10 EB to day 20 EB, and one dishof 6×105 ESCs. Simple scale-up can be achieved by multiplying the starting

t of the figure and are defined in Table 1.


3.2. Purity of derived ESDM

To assess their purity, ESDM-Ad were first characterisedby morphology both in culture and in stained cytocentrifugedpreparations. In adherent culture, the cells were stronglyadhesive to plastic and extended long plasma membraneprojections with varied morphology, including bipolar andmultipolar forms (Fig. 2A). The morphology indicated thatthese cells might be macrophages or dendritic cells (DC). Thecells were strongly adherent and detachment directly with acell lifter yielded a large proportion of non-viable cells, (50–60% by trypan blue exclusion). Pre-treating the cells withcold detachment buffer before scraping with a cell lifter couldincrease the viability to ~90%. Previous work has shown thatmouse DC are semi-adherent, and despite initially adheringto plastic surfaces, they become non-adherent after overnightculture (Steinman and Cohn, 1973). The strong adherence ofthe ES-derived cells is therefore more consistent with amacrophage phenotype than that of a DC.

After detachment, the cells were prepared as single-cellsuspension for cytocentrifugation and staining. Although thecells were moderately heterogeneous in morphology (Fig. 2B),the majority demonstrated characteristics of mature macro-phages, including large size with low nucleocytoplasmic ratio,eccentrically placed, round or kidney-shaped nucleus and acytoplasm containing numerous vacuoles, reflecting activepinocytosis (Fig. 2C). Some cells displayed an immaturemacrophage morphology with a higher nucleocytoplasmicratio and more densely stained basophilic cytoplasm. Amongthe mature cells, some either bound to cell debris or displayedevidence of apoptotic cells in phagocytic vacuoles, demon-strating the cells' active engagement in apoptotic-cell clearancein the cultures (Fig. 2C). Some of the cells were multinucleatebut none had a polymorphonuclear leucocytemorphology. The

Fig. 2. Morphological comparison of ESCs and ESDM. A: phase contrasmicrograph of ESDM-Ad in culture showing bipolar (arrow, B) andmultipolar (arrow, M) morphologies. B: ESDM differentiated from macrophage progenitor cells by adherent culture for 7days (ESDM-Ad). Cytocentrifuge preparation of detached ESDM-Ad revealed by Quick-Diff staining. CESDM bind and engulf apoptotic cells and cell debris. Cytocentrifugepreparation of detached ESDM-Ad revealed by Quick-Diff staining. Insethigher magnification showing internalized material. D: morphology of ESCrevealed by Quick-Diff staining of cytocentrifuge preparation.

t

--:

:s

morphology of adherent cells in cytocentrifuged preparationswas distinct from that of the originating ESCs (Fig. 2D).

Together, the adherent nature, morphology, pinocytic andphagocytic activities strongly suggested that the ESDM soderived were indeed macrophages. In order to confirm macro-phage identity, antibodies against macrophage-specific markersF4/80 CD68, and CD11bwere used for immunostaining (Fig. 3A).F4/80 is a cell surface glycoprotein which is a member of theEGF-TM7 protein family and it has been found that theexpression of F4/80 is heterogeneous and can vary duringmacrophage maturation and activation. Unlike other macro-phage markers, F4/80 is not expressed on other types of bloodcells (Nussenzweig et al., 1981). It can be detected onnon-adherent macrophage progenitor cells from bone marrowand the level of expression increasedwithmaturation (Hirsch etal., 1981). CD68, also named macrosialin, is a heavily glycosy-lated transmembrane protein, expressed by tissuemacrophages,Langerhans cells and at low levels by DC. It is the murinehomologue of the humanmacrophage glycoprotein CD68, and amember of the lysosomal-associated membrane protein familylocated predominantly intracellularly. Minute but definiteamounts of CD68 have also been detected on macrophagesurfaces. Cytocentrifuge preparation and fixation with acetonerendered the intracellular antigen accessible to antibody. CD11b,known as the integrin alpha M chain, is implicated in differentadhesive interactions of monocytes, macrophages and granulo-cytes. As shown in Fig. 3A, most ESDM-Ad showed very strongstaining of CD68, CD11b and F4/80, suggesting a high levelexpression of these antigens, while a minority were weaklystained. More than 80% of cells were positively stained byF4/80 and CD11b antibodies, and more than 90% were CD68positive (Fig. 3B). This macrophage marker profile, whichwas recapitulated exactly by ESDM-Ad populations stainedin situ using chamber slides (results not shown), confirmedthat the vast majority of the ESDM were truly macrophages.Importantly, undifferentiated E14 ESCs were negative for allthree macrophage markers (data not shown).

To investigate the purity of the ESDM further, we testedwhether contaminating undifferentiated ESCs were presentusing clonal assays based on the principle that the undiffer-entiated state of ESCs is characterised by a high-level expressionof alkaline phosphatase (Pease et al., 1990). To this end,ESDM-Ad were seeded at varying densities (103, 104, 105, and106 cells perwell) and undifferentiated E14 ESCswere seeded at103 perwell as a positive control (seeMaterials andmethods forfull details). Given that LIF is a key factor for ESC self-renewal,clonal assays were performed in the presence and absence ofthis factor. As expected, in the presence of LIF the majority ofcolonies produced from control, undifferentiated ESCs weretightly-packed stem-cell colonies with high alkaline phospha-tase activity (Fig. 4A, G). By contrast, in the absence ofLIF the majority of colonies were of the loosely-packedalkaline phosphatase-negative differentiated type (Fig. 4C, G).Some mixed colonies formed in both the presence and absenceof LIF (Fig. 4B,G). No stem-cell or differentiated colonies weregenerated when ESDM were plated in this clonal assay at anyseeding density tested (Fig. 4D, E, G) indicating that contami-nation by undifferentiated ES cells was less than one permillionESDM. We confirmed that the cell populations tested wereviable by demonstrating that ESDM grewwell when cultured inmedium supplemented with CSF-1 (Fig. 4F, G).

image of Fig.�2

Fig. 3. Immunocytochemical demonstration of macrophage markers expressed by ESDM. A: cytocentrifuge preparations of ESDM-Ad cells stained withperoxidase-conjugated rat anti-mouse macrophage specific antibodies against F4/80, CD68 and CD11b (upper panels) or isotype control reagents (lower panels).Binding was detected using DAB substrate (brown) and haematoxylin was used as counterstain. B: quantitative immunocytochemistry. The percentage ofpositively stained cells for each marker was determined by scoring ten randomly selected fields. Data are means+SEM for one representative experiment ofthree.


3.3. Comparison of ESDM with bone marrow-derived andperitoneal murine macrophage preparations

Further investigations of macrophage purity were under-taken by flow cytometry using the macrophage markers F4/80and CD11b and comparisons were made with bone marrow-derived macrophages (BMDM) and resident peritoneal cellswhich are commonly used sources of murine macrophages forin vitro studies. In these analyses, CD11c was used principallyas a marker of DC.

Macrophages from each of the three sources displayed awide range of forward and side light scatter which required logscales to be used for analysis (Fig. 5A–D, top panels). For bothBMDM and peritoneal macrophages, three subpopulationscould be distinguished by light scatter (Fig. 5A, B). For BMDM,cells with higher forward and side scatter (gate R1, Fig. 5A)expressed F4/80, CD11b and amajority expressed CD11c,whilethose with intermediate scatter (gate R2, Fig. 5A) expressedlower levels of all three markers. For peritoneal macrophages,cells in scatter gates R1 (Fig. 5B) were positive for F4/80 andCD11b, but negative for CD11c. 30–40% of peritoneal macro-phages in gate R2 (Fig. 5B) expressed F4/80 and CD11b butnone expressed CD11c. For both BMDM and peritonealmacrophages, cells with lower side scatter (gates R3, Fig. 5A,B) were negative for all threemarkers. These data indicate that

(1) both BMDM and peritoneal cells were contaminated bysignificant proportions of F4/80-negative and CD11b-negativecells and therefore presumably notmacrophages, and (2) BMDMcontained a significant proportion of cells expressing CD11cwhich could be either DC or CD11c+ macrophages. Analysis ofESDM-Ad also revealed three subpopulations defined by lightscatter (Fig. 5C), two major (gates R1 and R2) and one minor(R3). Unlike BMDM and peritoneal cells, ESDM-Ad in all three ofthese subpopulations were positive for F4/80 and CD11b, butnegative for CD11c. Similar to resident peritoneal macrophages,therewas also aminor population in gate 2 thatwas negative forF4/80 and CD11b.

We compared ESDM-Ad derived from progenitor cellsharvested from the adherent EB at early (day 10) and late (day20) stages of our differentiation protocols (Fig. 5C and 5D,respectively). Progenitor cells were cultured for 7days prior toflow cytometric analysis. No significant differences were ob-served in macrophage and DC marker expression between cellsderived from early and late cultures; both were found to bepositive for F4/80 and CD11b, but negative for CD11c (Fig. 5C, D).A small minority subpopulation of low side scatter cells fromearly cultures (R3, Fig. 5C) was positive for F4/80 and CD11b butnegative for CD11c. This subpopulation was not evident in cellsfrom late cultures. Propidium iodide staining indicated that theminor population of cells derived from both cultures that did not

image of Fig.�3

Fig. 4. Clonal density assay to detect potentially contaminating ESCs in ESDM preparations. Alkaline phosphatase activity (pink reaction product) was used as amarker for undifferentiated E14 ESCs. A–C: examples of the three categories of colonies formed by ESCs in the presence or absence of LIF. A: stem-cell colony(tight, strongly-stained, colony formed in the presence of LIF). B: mixed colony (alkaline phosphatase-positive stem cells and differentiated, unstained cells)formed in the presence or absence of LIF. C: differentiated colony (translucent dispersed colony with little or no stain) formed in the absence of LIF. D–F: ESDM-Adwere cultured in GMEMSR in the presence (D) or absence (E) of LIF. Data shown are from seeding 103 cells in 5mL, but no colonies were observed when up to 106

cells were seeded. F; ESDM-Ad cultured in ESDMCult. G: quantification of colonies formed by 103 ESCs or ESDM-Ad plated in the presence or absence of LIF asindicated. Data are the means+SD for the numbers of colonies per well for duplicate samples from two experiments.


express F4/80 (scatter gate R2, Fig. 5C, D) was non-viable (datanot shown). Non-viable cells comprised 1–10% of the total fromboth cultures and were evenly distributed between gates R1andR2.

3.4. Phenotype of ESDM and their precursors at different stagesof maturation

For certain applications (including investigations of macro-phage chemotaxis — see below) a preferred alternative toharvesting adherent cells from plastic surfaces was to harvestESDM directly from short-term suspension culture followingvarious periods of maturation under adherent conditions. Thus,for some experiments, the ESDM precursors harvested fromadherent EB were analysed directly (these cells being termedESDM-0), or were cultured adherent to Petri dishes for(a) 2days or (b) 6days, then detached and further cultured for

Fig. 5. Flow cytometric comparison of macrophage and dendritic cell markers on ESplots show the forward (FSC-H) and side (SSC-H) light scatter of the cells. Histogramdefined by R1–R3 after staining with F4/80, CD11b and CD11c antibodies. A: Bmacrophages from Balb/c mice. R1: 54%, R2: 9%, R3: 34%. C: ESDM-Ad obtained fromadherent EB at day 20. R1: 86%, R2: 13%. Isotype control staining is shown byRepresentative data from three experiments.

1day in suspension in Teflon pots. The resultant populationswere termed (a) ESDM-3 and (b) ESDM-7 respectively, reflect-ing the 3 or 7day further culture of precursors (Fig. 1). Withrespect to F4/80, CD11b and CD11c, the phenotype of ESDM-7was identical to that described above for adherent ESDM liftedon day 7 (data not shown). Furthermore, ESDM-0 and ESDM-3presented similar phenotypeswith high expression of F4/80 andCD11b but no expression of CD11c (Fig. 6A, B).

Taken together, these results indicate that the ESDMproduction protocols yielded macrophages and their pre-cursors in high purity.

3.5. Phagocytic activity of ESDM

Observations of ESDMcultures (e.g. Fig. 2C) suggested thatESDMwere capable of engulfing dead and dying cells. In orderto determine whether ESDM could be utilized in modelling

DM-Ad, BMDM and resident peritoneal macrophages.For each panel A–D dots are the fluorescence intensity distributions for cells in the light scatter gatesMDM from Balb/c mice. R1: 22%, R2: 9%, R3: 65%. B: resident peritonealadherent EB at day 10. R1: 78%, R2: 16%, R3: 3.8%. D: ESDM-Ad obtained fromthe shaded areas. Bars indicate the percentage positive for each marker.

image of Fig.�4


image of Fig.�5

Fig. 6. Flow cytometric comparison of macrophage and dendritic cell markers on ESDM-0 and ESDM-3. A: ESDM-0. R1: 83%, R2: 12%, R3: 4%. B: ESDM-3. R1: 86%,R2: 11%, R3: 1%. Other details as in legend to Fig. 5.


this process quantitatively, we evaluated the ability of ESDMto phagocytose apoptotic cells in a flow cytometry-based assayadapted from that applied previously to a study of humanmonocyte-derived macrophages (Jersmann et al., 2003).

We used the GFP-expressing ES cell line, GFP#7a, whichconstitutively expresses eGFP (Gilchrist et al., 2003) as thesource of ESDM in some experiments while lymphoma cellslabelled with the far-red fluorescent dye DDAO-SE induced toundergo apoptosis by UV-irradiation were phagocytic targets;apoptosiswasmonitored by annexinV/PI staining (Fig. 7A) andtheir far-red fluorescencewas assessed as shown in Fig. 7B (leftpanel “UV-BL2”). The fluorescence of GFP#7a-derived, consti-tutive eGFP-expressing ESDM-Ad, which were phenotypicallyidentical to E14-derived ESDM-Ad (exemplified by F4/80 andCD11b expression, Fig. 7C), was assessed as shown in Fig. 7B(middle panel “GFP-ESDM”). Co-incubation of apoptotic lym-phoma cells with ESDM resulted in combined green/red eventsin the top right quadrant of the dot plot as shown in Fig. 7B

Fig. 7. Phagocytosis of apoptotic cells by ESDM.ESDM were assessed for their abcytometric phagocytosis assay. A–C: DDAO-SE-labelled BL2 cells were UV-irradiatedGFP-expressing ESDM-Ad derived from GFP#7a ESCs. After washing and trypsinizatiused in phagocytosis assay shown in B, as assessed by flow cytometric analysis of aDDAO-SE label which did not affect apoptosis induction in this system (data not shoof apoptotic, DDAO-SE-labelled BL2 cells alone (UV-BL2, left panel), GFP-expressingof these cells after co-culture (UV-BL2+GFP-ESDM, right panel). Values indicate pe(lower right quadrants) and macrophages phagocytosing apoptotic cells (upper righanalysis of F4/80 and CD11b expression. Isotype control staining is shown by thPKH26-labelled BL2 cells were induced to undergo apoptosis by cold-shock treatmenAfter washing and trypsinization, cells were analysed by flow cytometry. Data are mBL2 cells and untreated controls.

(right panel, “UV-BL2+GFP-ESDM”). These represent apopto-tic lymphoma cells interacting with ESDM. Because theseinteractions were resistant to trypsin/EDTA treatment whichhas been reported to release apoptotic cells from othermacrophages, it seems highly probable that the double-positive, red/green events mainly represent ESDM that havephagocytosed apoptotic lymphoma cells. No interactions wereobserved in these assays between viable lymphoma cells andESDM (data not shown).

When ESDM at different stages of maturation were com-pared, significantly enhanced phagocytosis of apoptotic cellswas strongly correlated with maturational stage. For theseexperiments, interactions between apoptotic lymphoma cellsand non-GFP-expressing E14-derived ESDM at different stagesof maturation (ESDM-0, -3 and -7 as described above) wereassessed using fluorescence green-labelled macrophages andfluorescence red-labelled apoptotic cells. As shown in Fig. 7D,ESDM-0 interacted only weakly with apoptotic lymphoma cells

ility to phagocytose apoptotic Burkitt lymphoma (BL2) cells using a flow, incubated at 37°C for 3h to induce apoptosis and co-cultured for 30 min withon, cells were analysed by flow cytometry. A: apoptosis of UV-treated BL cellsnnexin V (AxV) binding and PI uptake. This was measured in the absence own). B: flow cytometric analyses of red (FL4) versus green (FL1) fluorescenceESDM alone (GFP-ESDM, middle panel) and trypsinized single cell suspensionrcentages of red apoptotic cells in upper left quadrants, green macrophagest quadrants). C: purity of ESDM from GFP#7a as measured by flow cytometrice shaded areas. Bars indicate the percentage positive for each marker. Dt and co-cultured for 1h with PKH67-labelled ESDM-0, ESDM-3, and ESDM-7eans+SE for the percentage ESDM demonstrating phagocytosis of apoptotic

f

:.

image of Fig.�6


image of Fig.�7

Fig. 8. Chemotactic activities of ESDM. Modified Boyden chamber assayswere used to measure migratory responses of ESDM in vitro. A: ESDM-7from Teflon pots were assayed for their ability to migrate towards variousconcentrations of CCL5, VEGF and C5a. Data are the means±SEM for thenumbers of migrated cells per high power field (HPF) for 10 random areas.B: demonstration of chemotaxis rather than chemokinesis of ESDM-7 whichmigrate towards C5a (50ng/mL) placed into the lower but not upper well ofthe Boyden chamber. Data are mean number of migrated ESDM/HPF+SEM.Control was migration in the absence of C5a. C: migration of ESDM at variousstages of maturation toward C5a (50ng/mL) and to supernatants ofapoptosis-triggered BL2 lymphoma cells. ESDM-0, -3 and -7 are macrophagepopulations at increasing stages of maturation as defined in Fig. 1. Data aremean number of migrated ESDM/HPF+SEM.


(in this case cold-shock-induced apoptosis), whereas ESDM-3and subsequently ESDM-7 became progressively more capableof phagocytosing apoptotic cells.

3.6. Chemotactic ability of ESDM

In order to determine whether the defined ESDM prepara-tions could be useful for studies of macrophage chemotaxis,

ESDM-0, -3 and -7 preparationswere tested for their abilities tomigrate towards a range of mononuclear phagocyte chemoat-tractants, including supernatants of lymphoma cells triggeredto undergo apoptosis. Direct use of macrophages isolated fromadherent culture in classical Boyden-chamber assays generatedinitially inconsistent results (data not shown). However, asshown in Fig. 8A, a regime of harvesting of macrophages fromPetri dishes followed by 24h culture in suspension in Teflonpots prior to use in Boyden assays generated highly reproduc-ible, dose-dependent migratory responses of ESDM to CCL5,VEGF and C5a. Such migratory responses were obtained whenchemokines were present in the lower, but not upper, wells ofthe chambers – as exemplified with C5a in Fig. 8B – confirmingthat the responseswere chemotactic rather than chemokinetic.Perhaps most significantly, chemotaxis was found to beinversely related to macrophage maturity. Thus as shown inFig. 8C, migratory activity towards both C5a and pre-apoptoticcell-derived supernatants was strongest in themost immature,ESDM-0 populations, with ESDM-3 and ESDM-7 cells display-ing markedly reduced responses. In the case of pre-apoptoticcell supernatants, responsiveness steadily diminished withincreasing macrophage maturity.

4. Discussion

The limited availability of convenient and appropriatemodel cell populations coupled with their inherent plasticityhas hampered investigation of macrophage biology in vitro.Primary cells are ideal in principle, but require regular accessto animals or human subjects, involve isolation proceduresthat impinge on cell functions and there is often significantcontamination by other cell types or by dead or dying cells thataffect macrophage activities. Widely used thioglycollate-elicited peritoneal macrophages can be isolated from animalsin high numbers however they are functionally distinct cellsrecruited from a monocyte subset and unrepresentative of theresident population (Ghosn et al., 2010). Macrophage-liketransformed cell lines are commonly used because theyare pure and readily available in unlimited numbers inthe laboratory, but are far from ideal because of theirnon-physiological state. Differentiation of ESCs affords theopportunity to produce virtually limitless, convenient suppliesof non-transformed macrophages and their precursors in highpurity and viability in the laboratory. Here we describeoptimised procedures for the production of highly purepopulations of ESC-derived murine macrophages (ESDM)at different stages of maturation and in particular considertheir application to studies of chemotaxis and phagocytosis,notably in response to cells undergoing apoptosis.

Given their advantages, murine and human ESDM havebeen used increasingly in recent years in wide-ranging studiesof the molecular cell biology of macrophages. However, littleattention has yet been paid to their potential application to theresponses of macrophages to detection and clearance of cellsundergoing apoptosis. Naturally, cell purity is paramount in allfunctional studies, but of additional critical importance to suchinvestigations of phagocyte responses to apoptosis is theviability of the macrophage population itself, since inadvertentexposure of the macrophages to cell death during isolation orproduction will have profound effects on the results of theinvestigations proper. Thus, presence of dead or dying cells –

image of Fig.�8


including dead or dyingmacrophages – during preparationwillaffect the sensitivity and reproducibility of subsequent con-trolled assays in which the prepared macrophages or theirprecursors are deliberately exposed to apoptotic or dead cells.Here we present an optimised protocol for the convenientpreparation of highly viable murine ESDM in high purity atdifferent maturational stages suitable for studies of macro-phage chemotaxis and phagocytosis, including responses toapoptotic cells. For studies of chemotaxis, we noted that assaysensitivity and reproducibility were markedly improved bypre-incubation of macrophages in suspension for 24h prior toassay set-up. Because harvesting from suspension is lesstraumatic to the macrophages than harvesting from adherentculture, it seems likely that this judicious pre-incubation stepprecluded ‘pre-activation’ of themacrophages by dying or deadcells generated during harvesting. Significantly, the ESDMgenerated according to the current protocol were found to beof markedly greater purity than either BMDM or freshlyisolated peritoneal macrophages – both commonly used asmodel murine macrophage populations in research applica-tions in vitro – and did not contain contaminating ESCs ordendritic cells.

Although the primary aim of the current work was toestablish an optimised protocol for the production of ESDMsuitable for migratory and phagocytic responses of macro-phages to apoptotic cells, the results of these studies alsorevealed reciprocal differences betweenmacrophages' abilitiesto migrate towards, versus phagocytose, apoptotic cells. Weobserved that mature macrophages were better able tophagocytose apoptotic cells whereas their precursors werebetter able tomigrate towards them.While it will be importantto understand the underlying molecular mechanisms, theseobservations suggest that the more mature ESDM may reflectthe activation state of resident tissue macrophages that neednot be required to navigate over relatively long distances priorto engulfing apoptotic cells. By contrast, the less mature ESDMappear reminiscent of monocytes or immature macrophages –which do not phagocytose apoptotic cells effectively (Newmanet al., 1982) – that are recruited to sites of apoptosis andmatureen route or following arrival. Candidate macrophage receptorsunderlying the sensing of, and migration towards, apoptoticcells include G2A, CX3CR1 and P2Y2, G-protein-coupled re-ceptors respectively engaging with the chemoattractantslysophosphatidylcholine, fractalkine (CX3CL1) and ATP/UTP(Lauber et al., 2003; Peter et al., 2008; Truman et al., 2008;Elliott et al., 2009) that are released from apoptotic cells. Inlight of the chemoattractive power of pre-apoptotic cells forESDM demonstrated here, it will be important to determinewhether these factors are released during high-density culturestress since this could have important implications, forexample, in the recruitment of macrophages to malignanttumours in which such stress occurs. Candidate macrophagereceptors and macrophage-derived bridging molecules in-volved in docking and engulfment of apoptotic cells includeCD14, CD36, αvβ3/5 integrins and MFG-E8 (Gregory andPound, 2010). Interestingly, fractalkine has been shown toenhance the apoptotic cell clearance capacity of mononuclearphagocytes via up-regulation of MFG-E8 (Miksa et al., 2007)and extracellular ATP is known to increase the apoptotic-cellbinding power of macrophages (Marques-da-Silva et al.,2011). It will be informative to determine whether analogous

mechanisms operate in the ESDM model described here. Weare currently testing the hypothesis that exposure of immatureESDM to apoptotic-cell derived chemoattractants activates theclearance mechanisms that are functional in mature ESDM.These results will be of interest not only in characterising inmolecular terms the apoptotic cell interaction mechanismsoperating in ESDM but also in comparing these to mechanismsthat may also be active in ESCs, since these too have beenreported to display some macrophage functions, including thecapacity to engulf apoptotic cells (Charriere et al., 2006). It isconceivable that the latter capacity may be via ‘amateur’phagocyte mechanisms shared by multiple cell types includingfibroblasts, endothelial cells and muscle cells (Gregoryand Pound, 2010) that have the ability to engulf apoptoticneighbours.

In conclusion, the optimised protocol described hereallows for the provision of valuable functional murinemacrophage populations in high purity and at varying statesof maturation that have the potential for use in a wide varietyof macrophage research applications. For example, in addi-tion to the migratory and phagocytic responses presented,the ESDM described here produce cytokines in response tomultiple stimuli (our unpublished observations). As well asproviding cells in high purity and in limitless quantities, theESDM approach also provides convenient opportunities forthe manipulation of macrophage genes. Macrophages arenotoriously difficult to manipulate by transfection, notleast because they possess important mechanisms to detectinvading cytoplasmic DNA and undergo apoptosis in re-sponse (Stacey et al., 1993; Roberts et al., 2009). By contrast,many sophisticated genetic engineering strategies have beendeveloped for use in ESCs, including conditional gene target-ing as well as constitutive and inducible over-expression andsiRNA knockdown. These strategies have been used widely toanalyse the molecular processes involved in the productionand function of specific cell types in vivo and in cell typesproduced from ESCs in vitro (Chambers et al., 2007; Loh et al.,2007; Iacovino et al., 2009; Jackson et al., 2012). We thereforeenvisage that the system described here can be used to assessthe molecular mechanisms involved in macrophage functionby targeting or over-expressing genes in undifferentiated ESCsthen differentiating these genetically manipulated ESCs intofunctional macrophages carrying the appropriate geneticalteration. Given that some genes are likely to have effects onmacrophage differentiation or maturation per se, conditionalmutants seem likely to be particularly valuable.

Acknowledgements

This work was supported by the China ScholarshipCouncil and Leukaemia and Lymphoma Research, UK. Wethank Julie Wilson for technical assistance.

References

Burgess, A.W., Metcalf, D., Kozka, I.J., Simpson, R.J., Vairo, G., Hamilton, J.A.,Nice, E.C., 1985. Purification of two forms of colony-stimulating factorfrom mouse L-cell-conditioned medium. J. Biol. Chem. 260, 16004.

Chambers, I., Silva, J., Colby, D., Nichols, J., Nijmeijer, B., Robertson, M., Vrana,J., Jones, K., Grotewold, L., Smith, A., 2007. Nanog safeguards pluripo-tency and mediates germline development. Nature 450, 1230.


Charriere, G.M., Cousin, B., Arnaud, E., Saillan-Barreau, C., Andre, M.,Massoudi, A., Dani, C., Penicaud, L., Casteilla, L., 2006. Macrophagecharacteristics of stem cells revealed by transcriptome profiling. Exp.Cell Res. 312, 3205.

Chawla, A., Barak, Y., Nagy, L., Liao, D., Tontonoz, P., Evans, R.M., 2001. PPAR-gamma dependent and independent effects on macrophage-geneexpression in lipid metabolism and inflammation. Nat. Med. 7, 48.

Dorshkind, K., 1990. Regulation of hemopoiesis by bone marrow stromalcells and their products. Annu. Rev. Immunol. 8, 111.

Elliott, M.R., Chekeni, F.B., Trampont, P.C., Lazarowski, E.R., Kadl, A., Walk,S.F., Park, D., Woodson, R.I., Ostankovich, M., Sharma, P., Lysiak, J.J.,Harden, T.K., Leitinger, N., Ravichandran, K.S., 2009. Nucleotides releasedby apoptotic cells act as a find-me signal to promote phagocyticclearance. Nature 461, 282.

Ghosn, E.E., Cassado, A.A., Govoni, G.R., Fukuhara, T., Yang, Y., Monack, D.M.,Bortoluci, K.R., Almeida, S.R., Herzenberg, L.A., Herzenberg, L.A., 2010.Two physically, functionally, and developmentally distinct peritonealmacrophage subsets. Proc. Natl. Acad. Sci. U. S. A. 107, 2568.

Gilchrist, D.S., Ure, J., Hook, L., Medvinsky, A., 2003. Labeling of hematopoi-etic stem and progenitor cells in novel activatable EGFP reporter mice.Genesis 36, 168.

Gregory, C.D., Pound, J.D., 2010. Microenvironmental influences of apoptosisin vivo and in vitro. Apoptosis 15, 1029.

Handyside, A.H., O'Neill, G.T., Jones, M., Hooper, M.L., 1989. Use of BRL-conditioned medium in combination with feeder layers to isolate adiploid embryonal stem cell line. Roux's Arch. Dev. Biol. 198, 48.

Hirsch, S., Austyn, J.M., Gordon, S., 1981. Expression of the macrophage-specific antigen F4/80 during differentiation of mouse bone marrowcells in culture. J. Exp. Med. 154, 713.

Iacovino, M., Hernandez, C., Xu, Z., Bajwa, G., Prather, M., Kyba, M., 2009. Aconserved role for Hox paralog group 4 in regulation of hematopoieticprogenitors. Stem Cells Dev. 18, 783.

Jackson, M., Taylor, A.H., Jones, E.A., Forrester, L.M., 2010. The culture ofmouse embryonic stem cells and formation of embryoid bodies.Methods Mol. Biol. 633, 1.

Jackson, M., Axton, R.A., Taylor, H., Wilson, J.A., Gordon-Keylock, S.A.M.,Kokkaliaris, K., Brickman, J., Schulz, H., Hummel, O., Hubner, N.,Forrester, L.M., 2012. HOXB4 can enhance the differentiation ofembryonic stem cells by modulating the haematopoietic niche. StemCells 30, 150.

Jersmann, H.P., Ross, K.A., Vivers, S., Brown, S.B., Haslett, C., Dransfield, I.,2003. Phagocytosis of apoptotic cells by human macrophages: analysisby multiparameter flow cytometry. Cytometry A 51, 7.

Karlsson, K.R., Cowley, S., Martinez, F.O., Shaw, M., Minger, S.L., James, W.,2008. Homogeneous monocytes and macrophages from human embry-onic stem cells following coculture-free differentiation in M-CSF and IL-3. Exp. Hematol. 36, 1167.

Lauber, K., Bohn, E., Krober, S.M., Xiao, Y.J., Blumenthal, S.G., Lindemann, R.K.,Marini, P., Wiedig, C., Zobywalski, A., Baksh, S., Xu, Y., Autenrieth, I.B.,Schulze-Osthoff, K., Belka, C., Stuhler, G., Wesselborg, S., 2003. Apoptoticcells induce migration of phagocytes via caspase-3-mediated release of alipid attraction signal. Cell 113, 717.

Loh, Y.H., Zhang, W., Chen, X., George, J., Ng, H.H., 2007. Jmjd1a and Jmjd2chistone H3 Lys 9 demethylases regulate self-renewal in embryonic stemcells. Genes Dev. 21, 2545.

Marques-da-Silva, C., Burnstock, G., Ojcius, D.M., Coutinho-Silva, R., 2011.Purinergic receptor agonists modulate phagocytosis and clearance ofapoptotic cells in macrophages. Immunobiology 216, 1.

Miksa, M., Amin, D., Wu, R., Ravikumar, T.S., Wang, P., 2007. Fractalkine-induced MFG-E8 leads to enhanced apoptotic cell clearance bymacrophages. Mol. Med. 13, 553.

Moore, K.J., Fabunmi, R.P., Andersson, L.P., Freeman, M.W., 1998. In vitro-differentiated embryonic stem cell macrophages: a model system forstudying atherosclerosis-associated macrophage functions. Arterioscler.Thromb. Vasc. Biol. 18, 1647.

Napoli, I., Kierdorf, K., Neumann, H., 2009. Microglial precursors derivedfrom mouse embryonic stem cells. GLIA 57, 1660.

Newman, S.L., Henson, J.E., Henson, P.M., 1982. Phagocytosis of senescentneutrophils by human monocyte-derived macrophages and rabbitinflammatory macrophages. J. Exp. Med. 156, 430.

Nussenzweig, M.C., Steinman, R.M., Unkeless, J.C., Witmer, M.D., Gutchinov,B., Cohn, Z.A., 1981. Studies of the cell surface of mouse dendritic cellsand other leukocytes. J. Exp. Med. 154, 168.

Odegaard, J.I., Vats, D., Zhang, L., Ricardo-Gonzalez, R., Smith, K.L., Sykes, D.B.,Kamps, M.P., Chawla, A., 2007. Quantitative expansion of ES cell-derivedmyeloid progenitors capable of differentiating into macrophages. J.Leukoc. Biol. 81, 711.

Pease, S., Braghetta, P., Gearing, D., Grail, D., Williams, R.L., 1990. Isolation ofembryonic stem (ES) cells in media supplemented with recombinantleukemia inhibitory factor (LIF). Dev. Biol. 141, 344.

Peter, C., Waibel, M., Radu, C.G., Yang, L.V., Witte, O.N., Schulze-Osthoff, K.,Wesselborg, S., Lauber, K., 2008. Migration to apoptotic “find-me”signals is mediated via the phagocyte receptor G2A. J. Biol. Chem. 283,5296.

Ralph, P., Nakoinz, I., 1975. Phagocytosis and cytolysis by a macrophagetumour and its cloned cell line. Nature 257, 393.

Raschke, W.C., Baird, S., Ralph, P., Nakoinz, I., 1978. Functional macrophagecell lines transformed by Abelson leukemia virus. Cell 15, 261.

Roberts, T.L., Idris, A., Dunn, J.A., Kelly, G.M., Burnton, C.M., Hodgson, S.,Hardy, L.L., Garceau, V., Sweet, M.J., Ross, I.L., Hume, D.A., Stacey, K.J.,2009. HIN-200 proteins regulate caspase activation in response toforeign cytoplasmic DNA. Science 323, 1057.

Stacey, K.J., Ross, I.L., Hume, D.A., 1993. Electroporation and DNA-dependentcell death in murine macrophages. Immunol. Cell Biol. 71 (Pt 2), 75.

Steinman, R.M., Cohn, Z.A., 1973. Identification of a novel cell type inperipheral lymphoid organs of mice. I. Morphology, quantitation, tissuedistribution. J. Exp. Med. 137, 1142.

Subramanian, A., Guo, B., Marsden, M.D., Galic, Z., Kitchen, S., Kacena, A.,Brown, H.J., Cheng, G., Zack, J.A., 2009. Macrophage differentiation fromembryoid bodies derived from human embryonic stem cells. J. StemCells 4, 29.

Truman, L.A., Ogden, C.A., Howie, S.E., Gregory, C.D., 2004. Macrophagechemotaxis to apoptotic Burkitt's lymphoma cells in vitro: role of CD14and CD36. Immunobiology 209, 21.

Truman, L.A., Ford, C.A., Pasikowska, M., Pound, J.D., Wilkinson, S.J., Dumitriu,I.E., Melville, L., Melrose, L.A., Ogden, C.A., Nibbs, R., Graham, G.,Combadiere, C., Gregory, C.D., 2008. CX3CL1/fractalkine is releasedfrom apoptotic lymphocytes to stimulate macrophage chemotaxis.Blood 112, 5026.

Wang, H., Grand, R.J.A., Milner, A.E., Armitage, R.J., Gordon, J., Gregory, C.D.,1996. Repression of apoptosis in human B-lymphoma cells by CD40-ligand and Bcl-2: relationship to the cell-cycle and role of theretinoblastoma protein. Oncogene 13, 373.

Way, K.J., Dinh, H., Keene, M.R., White, K.E., Clanchy, F.I., Lusby, P., Roiniotis,J., Cook, A.D., Cassady, A.I., Curtis, D.J., Hamilton, J.A., 2009. Thegeneration and properties of human macrophage populations fromhemopoietic stem cells. J. Leukoc. Biol. 85, 766.

Young, D.A., Lowe, L.D., Clark, S.C., 1990. Comparison of the effects of IL-3,granulocyte-macrophage colony-stimulating factor, and macrophagecolony-stimulating factor in supporting monocyte differentiation inculture. Analysis of macrophage antibody-dependent cellular cytotoxicity.J. Immunol. 145, 607.

journal of immunological methods -...

Documents