Journal of Immunological Methods 396 (2013) 33–43

Contents lists available at ScienceDirect

Journal of Immunological Methods

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

A flow cytometric method for the analysis of macrophages inthe vascular wall

Jeffrey P. Moore a,1, Samy Sakkal b,1, Michelle L. Bullen a, Barbara K. Kemp-Harper a,Sharon D. Ricardo c, Christopher G. Sobey a, Grant R. Drummond a,⁎a Vascular Biology and Immunopharmacology Group, Department of Pharmacology, Monash University, Clayton, Victoria, Australiab School of Biomedical Sciences, Victoria University, St Albans, Victoria, Australiac Monash Immunology and Stem Cell Laboratories, Monash University, Clayton, Victoria, Australia

a r t i c l e i n f o

⁎ Corresponding author at: Department of PharmacoloClayton, Victoria 3800, Australia. Tel.: +61 3 9905 4869;

E-mail address: Grant.Drummond@monash.edu (G1 These authors contributed equally.

0022-1759/$ – see front matter © 2013 Elsevier B.V. Ahttp://dx.doi.org/10.1016/j.jim.2013.07.009

a b s t r a c t

Article history:Received 14 January 2013Received in revised form 3 May 2013Accepted 22 July 2013Available online 6 August 2013

Macrophages accumulate in the vascular wall during conditions such as hypertension andhypercholesterolemia and contribute to vascular remodelling. Here we describe a method forthe isolation and subsequent flow cytometric analysis of macrophages from aortas of mice. Cellsuspensions were prepared from thoracic aortas of male C57BL/6J mice by a combination ofmanual disruption, incubation in enzymatic digestion medium, and passage through a 70 μmcell strainer. Flow cytometric analysis of these suspensions revealed a high content of cellswith strong light scattering properties (i.e. SSChi) compared with suspensions derived frommouse blood, spleen, thymus or kidney. Unstained aortic cell suspensions also displayed apreponderance of autofluorescence in the B670, V560, V460, B525 and V610 channels of theflow cytometer, suggesting that these channels should be avoided for subsequent flowcytometric analyses, at least for initial gating steps. Thus, aortic preparations were labelledwith an APC-Cy7-conjugated antibody against the pan-leukocyte marker, CD45, as well as anAPC-conjugated antibody against the macrophage-specific antigen, F4/80, as these fluoro-chromes emit in channels that displayed relatively low levels of auto-fluorescence in our initialstudies (i.e. R780 and R660). Flow cytometric analysis of labelled aortic preparations revealeda distinct population of CD45+F4/80+ cells. Importantly, back-gating on this CD45+F4/80+

cell population showed it to be now virtually devoid of autofluorescence in all remaining openchannels, and hence an appropriate foundation for further detailed analysis of macrophagepolarization using multiple intra- and extra-cellular markers. Furthermore, we demonstratedthat angiotensin II-induced hypertension in C57BL6/J mice, and hypercholesterolemia inapolipoprotein E-deficient mice, each resulted in an approximate doubling of CD45+F4/80+

cells in the aortic wall, highlighting the utility of our new protocol for studying the impact ofdisease on macrophage accumulation in the vascular wall.

© 2013 Elsevier B.V. All rights reserved.

Keywords:Flow cytometryAortaMacrophageAutofluorescenceHypertensionHypercholesterolemi

1. Introduction

Cardiovascular risk states, including hypertension, hypercho-lesterolemia and diabetes are associated with the accumulation

gy, Monash University,fax: +61 3 9902 9500..R. Drummond).

ll rights reserved.

of macrophages in the vascular wall, especially in large conduitvessels such as the aorta and carotid arteries (Chan et al., 2012;Wenzel et al., 2011). These macrophages accumulate in thesubendothelial space and are a potential source of reactiveoxygen species (ROS) and cytokines, which are likely causes ofthe endothelial dysfunction and vascular inflammation that arehallmarks of hypertension, hypercholesterolemia and diabetes(Harrison et al., 2012). Moreover, macrophages have a propen-sity to engulf extracellular lipoproteins, particularly those that

34 J.P. Moore et al. / Journal of Immunological Methods 396 (2013) 33–43

have undergone oxidative modification, and in doing sobecome transformed into foam cells, one of the majorcellular constituents of atherosclerotic plaques (Ley et al.,2011).

To date, most studies investigating the role of macrophagesin vascular disease states have relied on immunohistochemical(IHC) and histological detection techniques (Bush et al., 2000;Clozel et al., 1991). Although such techniques may providevaluable information about the proximity of macrophages toother cells types and their location within the various layers ofthe vascular wall, the fact that they are generally performedonly on a small number of thin tissue sections means that datamay not be representative of the entire blood vessel understudy. In addition, such limited sampling may not detect cellspresent in very low numbers. IHC and histological techniquesalso have limitations in discerning between subtypes of closelyrelated cells, for example, differentmacrophage subpopulations.This is an important issue because identification of the differentsubpopulations of macrophages in the vascular wall in healthversus disease is critical for understanding their roles in vascularphysiology and pathophysiology (Gordon and Taylor, 2005).Such information will likely yield novel therapeutic strate-gies for minimizing the detrimental actions of macrophages(i.e. pro-oxidant, pro-inflammatory) without compromisingtheir beneficial effects (i.e. immunological, tissue repairing)on the vascular wall.

In contrast to IHC and histological techniques, flow cytom-etry allows rapid sampling of cell populations in relatively largeamounts of tissue (e.g. an entire blood vessel). Furthermore, thefact that most standard flow cytometric analysers have thecapacity to simultaneously measure 8 or more fluorescentsignals in a single sample, means that cells and their subpop-ulations can be defined on the basis of multiple surface and/orintracellular cell-specific markers.

However, in spite of its potential advantages, there areseveral factors that must be addressed before flow cytometrycan be employed to reliably quantify cell populations in a giventissue. One such factor is the quality of the cell suspension uponwhich flow cytometric analysis is performed. Specifically,samples should consist of fully dissociated, viable cells thatretain most (if not all) of the features (e.g. surface antigens)that distinguish them from other cell types in vivo. Forperipheral blood, and for primary and secondary immuneorgans such as the thymus and spleen, both of which are heldtogether by relatively diffuse extracellular matrices, prepara-tion of viable single cell suspensions can be achieved effectivelywith onlyminimal requirement (or no requirement at all in thecase of blood) formanual disruption of the tissue (Swirski et al.,2007). However, for tissues with a high content of high tensilestructural proteins such as collagen and elastin, more robusttechniquesmay be required to liberate cells, usually involving acombination of manual disruption and chemical digestionwithcollagenase enzymes. Depending on the incubation time withthe digestive enzymes, such protocols are usually very efficientat dissociating cells, even from tissues with a high content oftough matrix materials (Galkina et al., 2006; Butcher et al.,2011). However, increasing incubation time with collagenasesmay bringwith it a reduction in cell viability, as well as cleavageof antigens from the surface of cells that could otherwise havebeen used to characterise them. As such it is critical to optimizedissociation protocols such that an optimum balance between

single cell numbers versus cell viability and antigen expressionis achieved.

Another important consideration when analysing cell popu-lations by flow cytometry is the ability to reliably detectantigen-positive cells above a background of autofluorescence.Autofluorescence, which refers to the inherent fluorescentproperties of a cell suspension, occurswhen endogenous cellularcomponents become excited by the excitation source of the flowcytometer (Hulspas et al., 2009). Molecules such as NADPH andflavins are particularly prone to fluorescing, especially whenexcited with low frequency wavelengths of light (e.g. 488 nm)(Monici, 2005). Structural proteins such as collagen and elastinare also highly autofluorescent (Andersson-Engels et al., 1997;Georgakoudi et al., 2002; Tinker et al., 1983) and this can haveobvious implications for subsequent flow cytometric analysis ofcell populations in solid tissues. Hence, for flow cytometricdetection of antigen-positive cells in any given tissue, it is crucialto have a thorough knowledge of its autofluorescent properties,so that appropriate gating strategies and thresholds abovebackground can be employed.

Here we describe a method for isolation and subsequentflow cytometric analysis of macrophages from aortas of micethat is associated with a minimum degree of cell death, andvirtually no reduction in the expression levels of leukocyte-and macrophage-specific surface antigens. Furthermore, wehighlight the fact that cell suspensions generated from aortasare highly auto-fluorescent and provide a flow-cytometricgating strategy that allows macrophages to be reliablyidentified despite these high background signals.

2. Materials and methods

2.1. Animals

Male C57BL/6J (8–19 weeks-old) and apolipoproteinE-deficient (ApoE−/−; 26 weeks-old) mice were used in thisstudy. Mice were obtained from the Monash University CentralAnimal Research Facility and all procedures were approved bythe Monash Animal Research Platform (MARP) Animal EthicsCommittee and conducted in compliance with the NationalHealth and Medical Research Council of Australia's (NHMRC)Guidelines for the Ethical and Humane Use of Animals inresearch. Mice were housed at 25 °C on a 12 h light/dark cyclein micro-isolator boxes under specific pathogen-free conditionsin littermate groups of up to four individuals. Mice had access tofood and water ad libitum and were maintained either onnormal chow or a “Western” diet (22% fat, 0.12% cholesterol;Specialty Feeds, Australia) from the time of weaning.

Hypertension was induced in a subgroup of C57BL/6J miceby chronic treatment with angiotensin II (Ang II). Theseanimals underwent surgery to implant an osmotic minipump(Alzet Model 2002; Alzet Corp), which delivered either Ang II(0.7 mg/kg/d, s.c.) or, in the case of the control animals,vehicle (0.9% saline, s.c.), at a constant infusion rate of0.11 μL/h for 14 d. Briefly, mice were anesthetized byisofluorane inhalation (2–4% with oxygen via a nose-cone)and a small lateral incision (5 mm) was made in the skin inthe scapula region. A subcutaneous pocket was then made viablunt dissection, intowhich an osmoticminipumpwas inserted.The woundwas closed with a stainless steel clip andmice were

35J.P. Moore et al. / Journal of Immunological Methods 396 (2013) 33–43

allowed to recover on a heating pad (~10 min) before beingreturned to their home boxes.

ApoE−/−mice utilised for hypercholesterolemia studieswere maintained on a “Western” diet (described above)normal chow diet until 6 weeks of age and then placed on a"Western" diet (described above) for 21 weeks. For mea-surement of total plasma cholesterol levels, blood from wildtype and ApoE−/− mice, was collected from the inferior venacava into heparinised tubes and plasma isolated via centri-fugation (4000 ×g at 4 °C for 10 min). Plasma total choles-terol levels were then determined using a Roche MODULAR917 enzymatic colorimetric array (Roche Diagnostics, CastleHill, NSW, Australia).

2.2. Blood pressure measurements

Blood pressurewasmeasured at days 0, 3, 7, 10 and 14 afterosmotic minipump implantation by tail-cuff plethysmographyusing a MC4000 Multichannel System (Hatteras Instruments,USA). Prior to the surgery, allmice had undergone “training” onthe tail cuff system for at least 3 consecutive days.

2.3. Preparation of cell suspensions

Mice were killed via CO2 asphyxiation and a length ofthoracic aorta, extending from proximal to the branch-pointof the left subclavian artery to just distal to the point atwhich the vessel passes through the diaphragm, wasexcised. The perivascular adipose tissue was retained on allspecimens. Aortas were placed in a microcentrifuge tubecontaining a digestion cocktail comprising collagenase type IX(125 U/ml), collagenase type I–S (450 U/ml) andhyaluronidaseIS (60 U/ml) (Sigma-Aldrich) dissolved in PBS supplementedwith Ca2+ and Mg2+ (Galkina et al., 2006; Guzik et al., 2007).Tissues were then subjected to manual disruption for 3 minusing fine scissors and incubated at 37 °C for either 20, 40 or60 min, while being gently agitated on a rocking platform.Following enzymatic digestion, cells were washed by tworounds of centrifugation at 1200 rpm (283 ×g) for 10 minand re-suspension in PBS buffer supplemented with 5 mMEDTA, 1% bovine serum albumin and 0.02% sodium azide.The suspension was then passed through a 70 μm cell

Aorta Kidney Spl

FS

SSC

Fig. 1. Flow cytometric analysis of light scattering properties of single cell suspensiopreponderance of cells with high side scatter (SSC) and low forward scatter (FSC) pronote that only the aorta and kidney cell preparations were prepared using enzymati(FSC-height vs FSC-area).

strainer (Falcon, BD) to remove undigested tissue/debris,and a small aliquot (10 μL) was taken and stained withTrypan blue for assessment of cell numbers and viabilityusing a Countess Cell Counter (Invitrogen, USA).

Kidneys, spleens, thymuses and blood were also removedfrom a subset of C57BL/6J mice for subsequent flowcytometric analysis. Kidney cell suspensions were preparedby a combination of manual disruption, enzymatic digestion(for 60 min) and passage through a 70 μm cell strainer, asdescribed for aortas above. Spleen and thymus cell prepara-tions were either processed using the conventional glass-slidemethod (Hammond et al., 1998) or, to enable direct compar-isons with aorta and kidney preparations, via the samecollagenase-based digestion procedure described above. Final-ly, whole blood samples were mixed with the anticoagulantClexane (400 U/mL) and analysed on the flow cytometerwithout the need for further processing.

2.4. Antibody labelling

A minimum of 5 × 105 cells (based on cell counts obtainedabove) from each aorta were incubated with the appropriateantibody cocktail for 25 min at 4 °C protected from light. Cellswere then washed and resuspended in supplemented PBSbuffer for immediate flow cytometric analysis. The viability dye7-Amino Actinomycin D (7-AAD) was added to each cellsuspension at a final concentration of 5 μg/mL, 3 min prior toanalysis on the flow cytometer. For isotype, fluorescenceminusone (FMO) and single-colour compensation controls, a similarnumber of aortic cells were labelled as described above withappropriate antibody cocktails. Sample data from 2 × 105

events were acquired on a LSR II Flow Cytometer (BDBiosciences, USA) using up to 9 fluorescence channels. Datawere exported as FCS3.0 files and analysed using FlowJosoftware 7.6.5 (Treestar Software, USA). Antibodies and isotypecontrols included APC Cy7 anti-CD45 (clone 30-F11), APC Cy7Rat IgG2b k isotype control (clone RTK4530), PE anti-NK1.1(clone PK136), PE anti-B220 (clone RA3-6B2), PE anti-CD49b(clone HMα2), PE anti-CD90.2 (clone 30-H12; BioLegend), PEanti-TER119 (clone TER-119; BD Biosciences), APC anti-F4/80(clone BM8), APC Rat IgG2a k isotype control (clone eBR2a;eBioscience) and 7-AAD (Molecular Probes).

een Thymus Blood

C

ns derived from aorta, kidney, spleen, thymus and peripheral blood. Note theperties in aortic cell suspensions compared with the other preparations. Alsoc digestion protocols. All dot plots are presented following doublet exclusion

Aorta

Spleen

Kidney

Thymus

Blood

A B

C D

EB

670

YG780YG585R660 R780B

670

V610V460V560 B525

B67

0

YG780YG585R660 R780

B67

0

V610V460V560 B525

B67

0

YG780YG585R660 R780

B67

0

V610V460V560 B525

B67

0

YG780YG585R660 R780

B67

0

V610V460V560 B525

B67

0

YG780YG585R660 R780

B67

0

V610V460V560 B525

Autofluoroescence

Fig. 2. Flow cytometric analysis of autofluorescence in unstained (A) aortic, (B) kidney, (C) splenic, (D) thymic and (E) blood cell suspensions across commonlyused fluorescence channels. Unstained cell suspensions were analysed for autofluorescence using the following bandpass filters: B670 (used on the Y axis in allgraphs); V610, V460, V560, B525, YG585, YG780, R780 and R660. Note the substantially higher levels of autofluorescence detected in all channels for aortic versusthe other cell suspensions. All dot plots are presented following doublet exclusion (FSC-height vs FSC-area).

36 J.P. Moore et al. / Journal of Immunological Methods 396 (2013) 33–43

2.5. Statistical analyses

All data are expressed as mean ± S.E.M. unless otherwisespecified. For statistical comparisons involving more than twoexperimental groups, one-way analysis of variance (ANOVA)was used. For comparisons between two data sets, unpaired ttests were employed. For blood pressure measurements, whichinvolved repeated measures on individual animals over differ-ent days, a repeated-measuresANOVAwas performed. A P valueof less than 0.05 was considered significant. All statisticalanalyses were performed using Graphpad Prism software(version 6.0b).

3. Results and discussion

3.1. Light scattering and autofluorescence properties of aorticcell suspensions

The light scattering properties of aortic cell suspensionswere compared with those of cell suspensions derived fromlymphoid organs and blood (Fig. 1). Whereas the vast

majority of cells present in blood and lymphoid organs displayedlow side scatter properties (SSClow), aortic cell suspensions bycomparison contained a large number of cells that were SSChi.These cells are likely to include the stromal cells of the aorticwallsuch as endothelial cells, vascular smooth muscle cells, fibro-blasts and adipocytes, which are present in far higher numbersthan leukocytes, at least under physiological conditions.We alsoexamined the side scatter properties of a cell suspension derivedfrom enzymatic digestion of another non-lymphoid organ, thekidney, and obtained a profile that was intermediate betweenthe aorta and lymphoid organs. Like the aorta, the kidney iscomprised of a complex population of leukocytes and stromalcells. However, notably, the kidneys were largely devoid ofadipose tissue (removed with the kidney capsule). Given thatadipocytes have previously been described as a SSChi cell-type(Le and Cheng, 2009; Lee et al., 2004), the lack of adipocytesmayexplain why there were fewer SSChi cells present in cellsuspensions derived from the kidney as compared with theaorta. Overall, these observations highlight the inherent difficul-ties in identifying leukocytes amongst the high background ofheterogeneous, SSChi stromal cells in tissues such as the aorta.

A B

C D

B67

0

YG780YG585R660 R780

B67

0

V610V460V560 B525

B67

0

YG780YG585R660 R780

B67

0

V610V460V560 B525

B67

0

YG780YG585R660 R780

B67

0

V610V460V560 B525

B67

0

YG780YG585R660 R780

B67

0

V610V460V560 B525

Spleen

Thymus

Enzymatic Non-enzymatic

Spleen

Thymus

Fig. 3. Flow cytometric analysis of autofluorescence in unstained (A) enzymatically- and (B) non-enzymatically-prepared thymus cell suspensions, and in(C) enzymatically- and (D) non-enzymatically-prepared splenic cell suspensions across commonly used fluorescence channels. Unstained cell suspensions wereanalysed for autofluorescence using the following bandpass filters: B670 (used on the Y axis in all graphs); V610, V460, V560, B525, YG585, YG780, R780 andR660. All dot plots are presented following doublet exclusion (FSC-height vs FSC-area).

37J.P. Moore et al. / Journal of Immunological Methods 396 (2013) 33–43

Cells with high light scattering properties (SSChi and/orforward scatter-high [FSChi]) often display a high degree ofautofluorescence, which can interfere with efforts to subse-quently detect cell-specific antigens using fluorescently-conjugated antibodies (Hulspas et al., 2009). Therefore,after excluding doublets (FSC-height vs FSC-area), weexamined the autofluorescent properties of unstained aorticcell suspensions in nine fluorescence channels that arecommonly used for flow cytometric analyses. In all cases, B670was used as the Y-axis parameter, as this channel is renownedfor its tendency to exhibit a high degree of autofluorescenceirrespective of the cell type examined (Mitchell et al., 2010).Even with this consideration, the levels of autofluorescencedetected in the B670 channelwere substantially higher for aorticcell suspensions (Fig. 2A) than for other cell preparations(Fig. 2B–E). Aortic cell suspensions also exhibited a high levelof autofluorescence in all other channels examined (Fig. 2A).Levels of autofluorescence were particularly high when cellswere analysed through the V560, V460, B525 and V610channels (upper panels in Fig. 2A). Although less autofluores-cence was detected in channels R660, YG585, R780 and YG780(lower panels in Fig. 2A), the levelswere still greater than thoseobserved in cell suspensions derived from the kidney, lym-phoid organs and blood (Fig. 2B–E).

To determine whether the enzymatic digestion protocolused for the aorta and kidney preparations may havecontributed to their higher levels of autofluorescence, wenext examined the impact of this preparation method onautofluorescence levels in spleen and thymus preparations.Autofluorescence levels in all channels examined were moreor less identical regardless of whether preparations were

prepared by enzymatic or non-enzymatic means (Fig. 3A–D).These observations suggest that the autofluorescence observedin the aorta and kidney cell suspensions is an intrinsic propertyof the tissues fromwhich they are derived. Hence, these findingshighlight the need for well-designed voltage-setting, com-pensation, fluorophore selection and gating strategies whenanalysing leukocyte subsets in aortic (and kidney) preparations.

The series of experiments depicted in Fig. 4 were carriedout on cell preparations labelled with only 7-AAD and anAPC-Cy7-conjugated CD45 antibody. 7-AAD is a fluorescentcompound that binds with high affinity to double-strandedDNA. Because it does not readily penetrate intact plasmamembranes, it only accumulates in dead or dying cells andcan thus be used to eliminate these cells from subsequentflow cytometric analysis. 7-AAD fluoresces most strongly inthe B670 channel which, based on our findings above, couldpotentially pose problems when working with aortic cellsuspensions (Lecoeur et al., 2002; Philpott et al., 1996). Ascan be seen in Fig. 4A, 7-AAD-positive cell populations wereclearly visible in the kidney, spleen and thymus preparations.By contrast, in aortic cell preparations, the 7-AAD-positivecell population was less clearly defined due to the high levelof background fluorescence in the B670 channel (Fig. 4A).Therefore, to ensure that only the 7-AAD-positive cells wereremoved from subsequent analyses, gating was set based onan unstained aortic sample (i.e. as per Fig. 2A).

Having selected for viable cells, the pan-leukocyte marker,CD45, was then used to distinguish leukocytes from thestromal cells of the aorta. An APC-Cy7-conjugated antibodywas selected for this critical first step in isolating leukocytesbecause this fluorochrome emits in the R780 channel, which,

Aorta Kidney Spleen Thymus

C

A

B

V610

B67

0

FSC

R78

0B

670

R660

Fig. 4. Flow cytometric analysis of the viable leukocyte populations present in single cell suspensions derived from aorta, kidney, spleen, and thymus. Panelsdepicted in Row A highlight the relative difficulty in excluding 7-AAD-positive cells in aortic preparations due to the high levels autofluorescence detected in theB670 channel. Panels in Row B demonstrate the well-defined cell populations detected following exclusion of dead cells in aortic and non-aortic preparationsalike using an APC-Cy7-conjugated antibody against the pan-leukocyte marker CD45. In Row C, back-gating onto open channels (in this example R660) revealsthat the APC-Cy7-conjugated anti-CD45 antibody effectively eliminates the autofluorescent cells that were initially present in the aortic preparation, opening upthe potential for detailed leukocyte subset characterisation using multiple fluorochromes. All dot plots are presented following doublet exclusion (FSC-height vsFSC-area).

38 J.P. Moore et al. / Journal of Immunological Methods 396 (2013) 33–43

in our earlier studies on unstained aortic cell suspensions,displayed a relatively low level of autofluorescence com-pared to other channels. The dot plots presented in Fig. 4Bshow clear populations of CD45-positive cells in all thepreparations examined. Not surprisingly, the relative pro-portion of CD45-positive to CD45-negative (i.e. stromal)cells was markedly lower in aorta and kidney (b1%), ascompared with spleen (~50–60%) and thymus (~70–80%).Importantly, when the CD45-positive aortic cell populationwas back-gated on the remaining open channels, autofluo-rescence was dramatically reduced in all cases, and virtuallyabsent in channels R660, YG585 and YG780 (Fig. 4C; see alsoSupplementary Fig. 1). Hence, these studies demonstrate theutility of the APC-Cy7-conjugated CD45 antibody for elim-inating most of the autofluorescent cell populations of theaortic wall. Furthermore, these studies highlight the R660,YG585 and YG780 (and R780) as the most appropriate

channels for further characterisation of leukocyte subsets inthe aorta.

The previous information was used to devise a gatingstrategy to analyse the macrophage population in aortas. Inaddition to 7-AAD and CD45-labelling, cell preparations werelabelled with an F4/80-APC monoclonal antibody for positivedetection of macrophages (Hume et al., 1983), as well as witha cocktail of PE-conjugated antibodies against markers ofseveral non-myeloid cell types including: CD49b and NK1.1(NK and NKT cells); B220 (B cells); CD90 (T cells); and Ter119(erythroid cells). Isotype and FMO controls were alsoprepared and used to set thresholds for positive expressionof the respective antigens. APC and PE were selected asfluorochromes because they emit in the R660 and YG585channels, respectively, which were identified as being two ofthe least autofluorescent channels in our analysis of aorticleukocytes (see Fig. 4C; see also Supplementary Fig. 1). After

Dead Cells

CD45+

Dump Gate F4/80+

CD45IsotypeControl

Viable Cells

F4/80IsotypeControl

Doublet Exclusion

Dump FMO7-AAD FMO

FSC-area

FSC

-hei

ght

V610

B67

0

FSC

R78

0

R780

YG

585

R780

R66

0

FSC

R78

0

V610

B67

0

R780

YG

585

R780R

660

Fig. 5. Application of flow cytometric gating strategy to quantify macrophage numbers in the aortic wall of mice. Following exclusion of doublets and dead cells,leukocytes were identified using an APC-Cy7-conjugated anti-CD45 antibody. Non-myeloid cells were then excluded from further analysis using a PE-conjugatedantibody cocktail against CD49b, NK1.1, B220; CD90 and Ter119, and finally macrophages were identified with an APC-conjugated anti-F4/80 antibody. Isotype orFMO controls for each of the above antibodies were used to determine positive population (see bottom row).

39J.P. Moore et al. / Journal of Immunological Methods 396 (2013) 33–43

omission of doublets (FSC-H vs FSC-A) and dead cells(7-AAD-positive), the aortic leukocyte population was iden-tified by gating on CD45, as described above (Fig. 5). Next,non-myeloid cells were excluded from further analysis bygating on the PE-negative population (Fig. 5). Finally, theremaining cells were defined as either macrophages ornon-macrophage myeloid cells (e.g. dendritic cells and/orgranulocytes) based on their positive or negative expressionof the F4/80 antigen (Fig. 5).

3.2. Impact of tissue digestion time on cell yield, viability andantigen expression

To assess the impact of enzymatic digestion time onleukocyte viability and yield, manually disrupted aortic homog-enates were incubated in the collagenase-based digestioncocktail for either 20, 40 or 60 min, stained with 7-AAD andantibodies, and analysed on the flow cytometer. There was aslight trend (albeit non-significant) for the 60 min incubationtime to liberate more total leukocytes and macrophages fromthe aortic wall, than either the 40 or 20 min incubationprotocols (Fig. 6A) while varying incubation times appeared tohaveminimal impact on leukocyte viability (Fig. 6B). Enzymatic

digestion may also result in the total or partial cleavage ofcertain cell surface proteins. Such cleavage could potentiallyaffect the expression/presentation of epitopes used forantibody-specific identification of leukocytes in flow cytom-etry. Although CD45 is reported to be relatively resistant toenzymatic cleavage (Gray et al., 2002, 2008), we neverthe-less assessed the impact of the different digestion times usedherein on expression of this antigen by measuring themedian fluorescence intensity of the signal emitted fromthe cell population detected in the R780 channel. Impor-tantly, varying the incubation time had no impact on themedian fluorescence intensity of the signal emitted from theCD45+ cell population. This suggests that there was minimaleffect on extracellular presentation of the CD45 antigen,even with the 60 min exposure time (Fig. 6C). Likewise,median fluorescence intensity of the F4/80 signal was notaltered by changing the digestion time, again indicating thatexpression of this antigen was not affected by our digestionprotocols (Fig. 6C). Finally, we also found that there was noeffect of varying the digestion time on the median fluores-cence intensity of the signal emitted from cells detected inthe ‘dump gate’ using the PE-conjugated antibody cocktail(Supplementary Fig. 2). This indicates that the digestion

A Cell ViabilityB C

% V

iabl

e C

ells

0

100

60

20

40

80

% V

iabl

e C

ells

Med

ian

Fluo

resc

ence

Int

ensi

ty

0

25000

15000

10000

2000060 min40 min20 min

CD45

0

4000

2000

3000

1000

F4/80

% M

ax%

Max

Antigen ExpressionCell Number

Cel

l num

bers

(per

2x1

05 ev

ents

)

0

500

2000

0

400

100

200

300

1500

1000

500

n=6n=6n=6

60 min40 min20 min

5000

n=6n=6n=6

60 min40 min20 min

n=6n=6n=6

60 min40 min20 min

n=6n=6n=6

60 min40 min20 min0

100

60

20

40

80

n=6n=6n=6

60 min40 min20 min

n=6n=6n=6

60 min40 min20 min

CD45

F4/80

CD45

F4/80

CD45

F4/80

Med

ian

Fluo

resc

ence

Int

ensi

ty

60 min40 min20 min

Cel

l num

bers

(per

2x1

05 ev

ents

)

100

80

60

40

20

0

0 102 103 104 105

100

80

60

40

20

0

0 102 103 104 105

Fig. 6. Effect of varying incubation times in a collagenase/hyaluronidase-based digestion medium on (A) total numbers and (B) percentage viability of leukocytes (CD45+) and macrophages (CD45+F4/80+) in aortic cellsuspensions. (C) Shows the lack of effect of the varying incubation times on expression levels of the CD45+ and F4/80+, as assessed by median fluorescence intensity (MFI).

40J.P.M

ooreet

al./JournalofIm

munologicalM

ethods396

(2013)33

–43

F480

+C

ell n

umbe

rs

(per

2x1

05 ev

ents

)F4

80+

Cel

l num

bers

(p

er 2

x105

even

ts)

400

saline Ang II

200

600

800

0 3 7 10 14Days of Ang II / saline Infusion

Syst

olic

BP

(mm

Hg)

100

110

120

130

140

150

160

170

saline

Ang IIA

*

*

B

Saline Ang II

D

0

250

100

WT ApoE-/-

50

200

150

*WT ApoE-/-

Hypertension

Hypercholesterolaemia

CD45

F4/8

0

CD45

F4/8

0

C

0

WT ApoE-/-

10

20

30

40

Plas

ma

Cho

lest

erol

(m

mol

/L)

*

Fig. 7. Application of the aortic digestion and flow cytometric gating strategy to assess the impact of experimentally-induced cardiovascular risk factors on macrophage numbers in the aortic wall. Panels (A) and (B)illustrate the effect of 14 d of Ang II infusion on systolic BP and accumulation of macrophages (CD45+F4/80+) in the aortic wall of mice, respectively. Panel (C) shows plasma cholesterol levels of hypercholesterolemic ApoE−/− versus the normocholesterolemic wild type strain. Panel (D) shows macrophage numbers in the aortic wall of hypercholesterolemic ApoE−/− versus the normocholesterolemic wild type strain. In B and D, left handpanels show group data (mean ± S.E.M.; n ≥ 6), while right hand panels show representative flow cytometric dot plots. *P b 0.05 by 2-way repeated measures ANOVA or unpaired t test.

41J.P.M

ooreet

al./JournalofIm

munologicalM

ethods396

(2013)33

–43

42 J.P. Moore et al. / Journal of Immunological Methods 396 (2013) 33–43

protocol also had minimal effects on presentation of thevarious antigens used to exclude non-myeloid leukocytesubsets in our gating strategy.

3.3. Application of aortic flow cytometry protocol to detectchanges in macrophage numbers during experimentally-induceddisease conditions

Hypertension and hypercholesterolemia are two majorrisk factors for the development of atherosclerosis, which isthe primary cause of myocardial infarctions and strokes. Oneof the earliest processes in the development of atheroscle-rotic plaques is the accumulation of macrophages in thesub-endothelial compartment of large arteries (Libby et al.,2011; Murray and Wynn, 2011). Hence, we used the flowcytometric method detailed above to quantify macrophageaccumulation in the aortic wall in experimental modelsof hypertension and hypercholesterolemia. Ang II is anocta-peptide produced in the kidneys that regulates bloodpressure by activating the sympathetic nervous system,promoting Na+/H2O reabsorption in the kidneys, and viadirect constrictor effects on the vasculature (Harrison et al.,2011). The treatment of mice with Ang II for 14 d caused amarked elevation in systolic BP compared to saline infusion(Fig. 7A). Consistent with previous findings using alternativeapproaches such as immunohistochemistry (Bush et al.,2000), 14 d of Ang II-induced hypertension was associatedwith a significant increase (~2-fold) in macrophage numbers inthe aortic wall (Fig. 7B), along with a similar change in totalleukocyte numbers (i.e. a 2-fold increase from 722 ± 111 to1433 ± 316 CD45+ cells per 2 × 105 events in normotensiveversus hypertensive aortas, respectively; P b 0.05). Likewise, in26 week-old ApoE−/−mice, which are genetically pre-disposedto the development of hypercholesterolemia, there was asignificant increase in plasma cholesterol (Fig. 7C) along with a2-fold increase inmacrophagenumbers in the aorticwall, whichis also consistent with previous reports (Glass and Witztum,2001; Judkins et al., 2010) (Fig. 7D). In addition, compared towild-type mice, there was a trend towards an increase in totalleukocyte numbers in the aortic wall of ApoE−/− mice (1100 ±226 to 1360 ± 279 CD45+ cells per 2 × 105 events,respectively).

3.4. Concluding remarks

Conditions such as hypertension and hypercholesterolemiaare associatedwith an influx of leukocytes into the arterial wall;a process that likely underpins vascular sequelae such asremodelling, fibrosis and formation of atherosclerotic plaques.Hence, a better understanding of the leukocyte subsets that arepresent, and their activation state in diseased versus healthyblood vessels may give insights into novel therapeutic inter-ventions to prevent/treat vascular disease. In the present studywe have described a protocol that allows robust and reproduc-ible flow cytometric analysis of leukocyte populations in theaortic wall of mice under steady state conditions and also inrelevantmodels of hypertension and hypercholesterolemia. Wehave shown that enzymatic digestion of aortic homogenates forup to 60 min yields single cell suspensions containing a highpercentage of viable cells. Moreover, this digestion protocolappears to have minimal impact on the expression levels of key

cell-specific antigens on the surface of the leukocytes within thesuspension.We have also demonstrated that the high content ofstromal cells (i.e. smooth muscle, fibroblasts, adipocytes)present in aortic cell preparations makes them highlyautofluorescent, especially when excited with wavelengthsin the violet or blue region of the light spectrum. However,by using the flow cytometric gating strategy devised herein,which avoids these short wavelength excitation channels forthe initial gating steps, the impact of this autofluorescenceon detection of antigen-positive leukocytes can be largelyavoided, highlighting flow cytometry as a powerful tool forfuture studies aimed at investigating the pathogenesis ofvascular disease.

Supplementary data to this article can be found online athttp://dx.doi.org/10.1016/j.jim.2013.07.009.

Conflicts of interest

None.

Acknowledgements

This work was supported by project grants from theNHMRC (1041326) and the National Heart Foundation ofAustralia (GM11M5825). GRD and CGS are supported bySenior Research Fellowships from the NHMRC, while JM andMB receive PhD stipends from Monash University. None ofthe funding sources outlined above had any role in studydesign; in the collection, analysis and interpretation of data;in the writing of the report; or in the decision to submit thearticle for publication.

References

Andersson-Engels, S., Klinteberg, C., Svanberg, K., Svanberg, S., 1997. In vivofluorescence imaging for tissue diagnostics. Phys. Med. Biol. 42, 815.

Bush, E., Maeda, N., Kuziel, W.A., Dawson, T.C., Wilcox, J.N., DeLeon, H.,Taylor, W.R., 2000. CC chemokine receptor 2 is required for macrophageinfiltration and vascular hypertrophy in angiotensin II-induced hyper-tension. Hypertension 36, 360.

Butcher, M.J., Herre, M., Ley, K., Galkina, E., 2011. Flow cytometry analysis ofimmune cells within murine aortas. J. Vis. Exp. 53, e2848.

Chan, C.T., Moore, J.P., Budzyn, K., Guida, E., Diep, H., Vinh, A., Jones, E.S., Widdop,R.E., Armitage, J.A., Sakkal, S., Ricardo, S.D., Sobey, C.G., Drummond, G.R.,2012. Reversal of vascularmacrophage accumulation and hypertension by aCCR2 antagonist in deoxycorticosterone/salt-treatedmice. Hypertension 60,1207.

Clozel, M., Kuhn, H., Hefti, F., Baumgartner, H.R., 1991. Endothelial dysfunctionand subendothelial monocyte macrophages in hypertension effect ofangiotensin converting enzyme inhibition. Hypertension 18, 132.

Galkina, E., Kadl, A., Sanders, J., Varughese, D., Sarembock, I.J., Ley, K., 2006.Lymphocyte recruitment into the aortic wall before and duringdevelopment of atherosclerosis is partially L-selectin dependent. J. Exp.Med. 203, 1273.

Georgakoudi, I., Jacobson, B.C., Müller, M.G., Sheets, E.E., Badizadegan, K.,Carr-Locke, D.L., Crum, C.P., Boone, C.W., Dasari, R.R., Van Dam, J., Feld,M.S., 2002. NAD(P)H and collagen as in vivo quantitative fluorescentbiomarkers of epithelial precancerous changes. Cancer Res. 62, 682.

Glass, C.K., Witztum, J.L., 2001. Atherosclerosis. The road ahead. Cell 104,503.

Gordon, S., Taylor, P.R., 2005. Monocyte and macrophage heterogeneity. Nat.Rev. Immunol. 5, 953.

Gray, D.H.D., Chidgey, A.P., Boyd, R.L., 2002. Analysis of thymic stromal cellpopulations using flow cytometry. J. Immunol. Methods 260, 15.

Gray, D.H.D., Fletcher, A.L., Hammett, M., Seach, N., Ueno, T., Young, L.F.,Barbuto, J., Boyd, R.L., Chidgey, A.P., 2008. Unbiased analysis, enrichmentand purification of thymic stromal cells. J. Immunol. Methods 329, 56.

Guzik, T.J., Hoch, N.E., Brown, K.A., McCann, L.A., Rahman, A., Dikalov, S.,Goronzy, J., Weyand, C., Harrison, D.G., 2007. Role of the T cell in the

43J.P. Moore et al. / Journal of Immunological Methods 396 (2013) 33–43

genesis of angiotensin II induced hypertension and vascular dysfunction.J. Exp. Med. 204, 2449.

Hammond, K., Cain, W., van Driel, I., Godfrey, D., 1998. Three day neonatalthymectomy selectively depletes NK1.1+ T cells. Int. Immunol. 10,1491.

Harrison, D.G., Guzik, T.J., Lob, H.E., Madhur, M.S., Marvar, P.J., Thabet, S.R.,Vinh, A., Weyand, C.M., 2011. Inflammation, immunity, and hyperten-sion. Hypertension 57, 132.

Harrison, D.G., Marvar, P.J., Titze, J.M., 2012. Vascular inflammatory cells inhypertension. Front. Physiol. 3, 128.

Hulspas, R., O'Gorman, M.R.G., Wood, B.L., Gratama, J.W., Sutherland, D.R.,2009. Considerations for the control of background fluorescence inclinical flow cytometry. Cytometry B Clin. Cytom. 76, 355.

Hume, D.A., Robinson, A.P., MacPherson, G.G., Gordon, S., 1983. Themononuclear phagocyte system of the mouse defined by immunohisto-chemical localization of antigen F4/80. Relationship between macro-phages, Langerhans cells, reticular cells, and dendritic cells in lymphoidand hematopoietic organs. J. Exp. Med. 158, 1522.

Judkins, C.P., Diep, H., Broughton, B.R.S., Mast, A.E., Hooker, E.U., Miller, A.A.,Selemidis, S., Dusting, G.J., Sobey, C.G., Drummond, G.R., 2010. Directevidence of a role for Nox2 in superoxide production, reduced nitricoxide bioavailability, and early atherosclerotic plaque formation inApoE−/− mice. Am. J. Physiol. Heart Circ. Physiol. 298, H24.

Le, T.T., Cheng, J.-X., 2009. Single-cell profiling reveals the origin ofphenotypic variability in adipogenesis. PLoS One 4, e5189.

Lecoeur, H., de Oliveira-Pinto, L.M., Gougeon, M.-L., 2002. Multiparametricflow cytometric analysis of biochemical and functional events associatedwith apoptosis and oncosis using the 7-aminoactinomycin D assay.J. Immunol. Methods 265, 81.

Lee, Y.-H., Chen, S.-Y., Wiesner, R.J., Huang, Y.-F., 2004. Simple flow cytometricmethod used to assess lipid accumulation in fat cells. J. Lipid Res. 45, 1162.

Ley, K., Miller, Y.I., Hedrick, C.C., 2011. Monocyte and macrophage dynamicsduring atherogenesis. Arterioscler. Thromb. Vasc. Biol. 31, 1506.

Libby, P., Ridker, P.M., Hansson, G.K., 2011. Progress and challenges intranslating the biology of atherosclerosis. Nature 473, 317.

Mitchell, A.J., Pradel, L.C., Chasson, L., Van Rooijen, N., Grau, G.E., Hunt, N.H.,Chimini, G., 2010. Technical advance: autofluorescence as a tool formyeloid cell analysis. J. Leukoc. Biol. 88, 597.

Monici, M., 2005. Cell and tissue autofluorescence research and diagnosticapplications. Biotechnol. Annu. Rev. 11, 227.

Murray, P.J., Wynn, T.A., 2011. Protective and pathogenic functions ofmacrophage subsets. Nat. Rev. Immunol. 11, 723.

Philpott, N.J., Turner, A.J., Scopes, J., Westby, M., Marsh, J.C., Gordon-Smith,E.C., Dalgleish, A.G., Gibson, F.M., 1996. The use of 7-amino actinomycinD in identifying apoptosis: simplicity of use and broad spectrum ofapplication compared with other techniques. Blood 87, 2244.

Swirski, F.K., Libby, P., Aikawa, E., Alcaide, P., Luscinskas, F.W., Weissleder, R.,Pittet, M.J., 2007. Ly-6Chi monocytes dominate hypercholesterolemia-associated monocytosis and give rise to macrophages in atheromata.J. Clin. Invest. 117, 195.

Tinker, D.H., Rucker, R.B., Tappel, A.L., 1983. Variation of elastin fluorescencewith method of preparation: determination of the major fluorophore offibrillar elastin. Connect. Tissue Res. 11, 299.

Wenzel, P., Knorr, M., Kossmann, S., Stratmann, J., Hausding, M., Schuhmacher,S., Karbach, S.H., Schwenk, M., Yogev, N., Schulz, E., Oelze, M., Grabbe, S.,Jonuleit, H., Becker, C., Daiber, A., Waisman, A., Münzel, T., 2011. LysozymeM-positive monocytes mediate angiotensin II-induced arterial hyperten-sion and vascular dysfunction. Circulation 124, 1370.