Characterization of CD45�/CD31þ/CD105þ Circulating Cellsin the Peripheral Blood of Patients with GynecologicMalignancies
Hyun-Kyung Yu1,2, Ho-Jeong Lee1,5, Ha-Na Choi1, Jin-Hyung Ahn1, Ji-Young Choi3, Haeng-Seok Song3,Ki-Heon Lee3, Yeup Yoon1, Lee S. H. Yi2, Jang-Seong Kim4, Sun Jin Kim5, and Tae Jin Kim3
AbstractPurpose: Circulating endothelial cells (CEC) have been widely used as a prognostic biomarker and
regarded as a promising strategy for monitoring the response to treatment in several cancers. However, the
presence andbiologic roles of CECshave remained controversial for decades because technical standards for
the identification and quantification of CECs have not been established. Here, we hypothesized that CECs
detected by flow cytometry might be monocytes rather than endothelial cells.
Experimental Design: The frequency of representative CEC subsets (i.e., CD45�/CD31þ, CD45�/CD31þ/CD146þ, CD45�/CD31þ/CD105þ) was analyzed in the peripheral blood of patients with gyne-
cologic cancer (n ¼ 56) and healthy volunteers (n ¼ 44). CD45�/CD31þ cells, which are components of
CECs, were isolated and the expression of various markers (CD146, CD105, vWF, and CD144 for
endothelial cells; CD68 and CD14 for monocytes) was examined by immunocytochemistry.
Results: CD45�/CD31þ/CD105þ cells were significantly increased in the peripheral blood of patients
with cancer, whereas evaluation of CD45�/CD31þ/CD146þ cells was not possible both in patients with
cancer and healthy controls due to the limited resolution of the flow cytometry. Immunocytochemistry
analyses showed that these CD45�/CD31þ/CD105þ cells did not express vWF and CD146 but rather
CD144. Furthermore, CD45�/CD31þ/CD105þ cells uniformly expressed the monocyte-specific markers
CD14 and CD68. These results suggest that CD45�/CD31þ/CD105þ cells carry the characteristics of
monocytes rather than endothelial cells.
Conclusions: Our data indicate that CD45�/CD31þ/CD105þ circulating cells, which are significantly
increased in the peripheral bloodof patientswith gynecologic cancer, aremonocytes rather than endothelial
cells. Further investigation is required to determine the biologic significance of their presence and function
in relation with angiogenesis. Clin Cancer Res; 19(19); 5340–50. �2013 AACR.
IntroductionOvercoming resistance to therapy is the ultimate goal of
the development of novel treatment modalities in cancer(1). The biologic heterogeneity and genetic instability ofcancer cells are significant barriers for the design of effectivetherapies. Therefore, relatively more homogenous andgenetically stable host factors have been suggested as alter-native targets (2). Angiogenesis, which is one of the com-mon and crucial steps in the development and progressionof solid tumors, is a host-dependent process and, conse-quently, has been introduced as an attractive target of cancertreatment (3). A significant number of drugs designed tointerrupt the establishment of tumor-associated vasculatureby neutralizing vasculogenic factors is currently in clinicaltrials and some of them have been approved for clinical usein patients with cancer (4, 5). However, understanding themechanisms of angiogenesis and establishing validatedmarkers that accurately reflect the pharmacologic effects ofantiangiogenic therapeutics remain major challenges (6).
Authors' Affiliations: 1Mogam Biotechnology Research Institute, Yongin;2Department of Biological Science, Sungkyunkwan University, Suwon;3Department of Obstetrics and Gynecology, Cheil General Hospital andWomen's Healthcare Center, Kwandong University College of Medicine,Seoul; 4Biomedical Translational Research Center, Korea Research Insti-tute of Bioscience and Biotechnology, Daejeon, Republic of Korea; and5Department of Cancer Biology, The University of Texas MD AndersonCancer Center, Houston, Texas
Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).
Corresponding Authors: Jang-Seong Kim, Biomedical TranslationalResearch Center, Korea Research Institute of Bioscience and Biotechnol-ogy, 111 Gwahangno, Yuseong-gu, Daejeon 305-806, Republic of Korea.Phone: 82-42-860-4270; Fax: 82-42-879-8498; E-mail:[email protected]; Sun Jin Kim, Department of Cancer Biology, TheUniversity of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd,Houston, TX 77030. Phone: 1-713-563-4653; Fax: 1-713-563-5489;E-mail: [email protected]; and Tae Jin Kim, Department of Obstet-rics and Gynecology, Cheil General Hospital and Women's HealthcareCenter, Kwandong University College of Medicine, 1-19 Mukjeong-dong,Jung-gu, Seoul 100-380, Republic of Korea. Phone: 82-2-2000-7577; Fax:82-2-2000-7183; E-mail: [email protected]
Because circulating endothelial cells (CEC) are likely tocontribute tonewvessel formation (7) and their levels in theblood change in response to pro- or antiangiogenic drugs(8–10), the measurement of CECs (total CECs includingprogenitor cells) has been regarded as a promising strat-egy for monitoring tumor angiogenesis. Several studiesreported a significant increase in the number of CECs inpatients with cancer with progressive disease (i.e., lym-phoma, breast cancer, renal cancer, etc.; refs. 11–13).In addition, studies have shown that CEC kinetics andviability correlate well with clinical outcomes of patientswith cancer undergoing antiangiogenic treatment (14,15).However, the presence and role of CECs have been
controversial for decades because trials using CECs as adiagnostic parameter or a therapeutic target did notproduce consistent results. Moreover, markers that accu-rately identify CECs have not been established becausemany circulating cells such as monocytes express many ofthe same markers as endothelial cells (16). Consequently,the detection and estimation of CECs have remainedchallenging as they comprise a small proportion ofperipheral blood cells, and there is no consensus in theimmunophenotype of CEC (17). The application of var-ious combinations of markers or different enumerationtechniques to the detection of CECs has produced dis-crepant results in the amount and immunophenotype ofCECs (18–20). Furthermore, certain authors have ques-tioned whether CECs detected by flow cytometry areauthentic endothelial cells and their function in angio-genesis, if any (17, 21). Recent studies have suggestedthat the actual angiogenic cell types incorporated intonewly formed vessels are myeloid cells such as monocytes(22–24) rather than CECs and/or their progenitor cells.
Therefore, establishing a method to validate the identityof CECs detected by flow cytometry and to assess theirbiologic significance is critical before expanding theirclinical use.
Considering the angiogenic role of monocytes and thetechnical hurdles of flow cytometry, we hypothesized thata subset of circulating cells detected by flow cytometryusing conventional CECmarkers might be monocytes rath-er than endothelial cells. To show this, we first evaluatedthe flow-cytometric techniques and markers currently inuse and analyzed the frequency of representative CECsubsets (i.e., CD45�/CD31þ, CD45�/CD31þ/CD146þ,CD45�/CD31þ/CD105þ; refs. 25–28) in the peripheralblood of patients with gynecologic cancer and healthyvolunteers. To identify the genuine lineage of those cells,we isolated CD45�/CD31þ cells (a common denominatorof CECs) and assessed the expression of variousmarkers forendothelial cells or monocytes by immunocytochemistry.
Materials and MethodsSubjects
Peripheral blood samples (1–2 mL) were collected from44 healthy donors (12 men and 32 women; age, 28–54years) and 56 patients with gynecologic cancer including 8patients with endometrial cancer (age, 39–59 years), 24with cervical cancer (age, 30–71 years), and 24with ovariancancer (age, 23–67 years; Supplementary Table S1). Allhealthy volunteers were free of any medications and hadno cardiovascular disease. The Institutional ReviewBoard atKwandongUniversity College ofMedicine (Seoul, Republicof Korea) approved all protocols, and informed consentwasobtained from all subjects.
Antibodies for flow cytometryThe following monoclonal antibodies directly conjugat-
ed with fluorescein isothiocyanate (FITC), phycoerythrin(PE), peridinin chlorophyll A protein (PerCP), or allophy-cocyanin (APC) were used for flow-cytometric analysis:anti-CD31 FITC (WM-59 clone), anti-CD61 FITC (VI-PL2clone), anti-CD3 PE (SK7 clone), anti-CD19 PE (HIB19clone), anti-CD31 PE (WM-59 clone), anti-CD41a PE(HIP8 clone), anti-CD56 PE (MY31 clone), anti-CD146 PE(P1H12 clone), anti-CD45 PerCP (2D1 clone), and anti-CD14 APC (M5E2 clone). Isotype-matched FITC-, PE-,PerCP-, and APC-conjugated control antibodies were pur-chased from BD Biosciences. Anti-CD105 PE (SN6 clone)and isotype-matched PE-conjugated control antibodieswere purchased from Serotec. Anti-CD31 APC (WM-59clone) and isotype-matched APC-conjugated control anti-bodies were purchased from eBioscience.
Preparation of peripheral blood mononuclear cellsPeripheral blood was collected from healthy volunteers
and patients with cancer using EDTA as an anticoagulantand processed within several hours after collection as fol-lows: whole blood was diluted 1:1 (vol/vol) with PBScontaining 0.5% bovine serum albumin (BSA) and 2mmol/L EDTA and overlaid onto an equal volume of Ficoll
Translational RelevanceCirculating endothelial cells (CEC) are recognized as a
marker of tumor angiogenesis and a predictor of prog-nosis, as well as a target for therapy. However, theirpresence in the peripheral blood of patients with cancerand role in angiogenesis remain controversial. Ourresults showed that CD45�/CD31þ/CD105þ cells weresignificantly increased, whereas CD45�/CD31þ/CD146þ cells were not detected in the peripheral bloodof patients with gynecologic cancer. In addition, theexpression of monocyte-specific markers such as CD14and CD68 in CD45�/CD31þ/CD105þ cells suggestedthat they are monocytes rather than endothelial cells.Collectively, our data suggest that the accuracy of con-ventional flow-cytometric analyses for identifying CECsshould be meticulously reevaluated in its technical andbiologic aspects. Moreover, further investigation is nec-essary to establish the biologic significance of the pres-ence of CD45�/CD31þ/CD105þ monocytes and theirfunction in relation to angiogenesis.
Characterization of Circulating Endothelial Cells
www.aacrjournals.org Clin Cancer Res; 19(19) October 1, 2013 5341
Paque (GE healthcare). Samples were centrifuged at 1,800rpm for 25minutes at room temperature with no brake. Themononuclear cell layer was carefully collected and washedtwice with cold PBS containing 0.5% BSA and 2 mmol/LEDTA at 4�C. Red blood cells were lysed with 0.38%ammonium chloride solution. The final mononuclear cellpreparation was resuspended with PBS containing 0.5%BSA and 2 mmol/L EDTA and then subjected to flow-cytometric analysis. The viability of the mononuclear cellsused for the analyses was determined by the dye exclusiontest, and cells with a viability of 99% or more were used forfurther experiments.
Flow cytometry and cell sortingIsolated peripheral bloodmononuclear cells (PBMC; 107
cells per mL of blood) were pretreated with FcR blockingreagent (Miltenyi Biotec) to block nonspecific antibodybinding and incubated on ice for 25 minutes with a panelof monoclonal antibodies (summarized in Table 1). Cellswere washed with PBS containing 0.5% BSA and 2 mmol/LEDTA and fixed with 4% paraformaldehyde (PFA; ElectronMicroscopy Sciences). The antibody-labeled cells were ana-lyzed using a FACS Aria flow cytometer (BD Biosciences)equipped with 2 lasers (488 nm and 633 nm). Data wereanalyzed with FlowJo software (Tree Star, Inc.,) or FACSDiva (BD Biosciences). For the analysis of CEC candidates,at least 100,000 singlet lymphocytes were isolated andthe frequencies of CD45�/CD31þ, CD45�/CD31þ/CD146þ, or CD45�/CD31þ/CD105þ cells were analyzedand expressed as a percentage of the singlet lymphocytepopulation. For sorting of CD45�/CD31þ cells, PBMCsprepared from patients with cancer were pretreated with
FcR blocking reagent and stained with fluorescence-labeledmonoclonal antibodies against CD45 and CD31. The cellswere fixed with 4% PFA and then sorted with a FACS Ariaflow cytometer. A 70-mmnozzle (BD Biosciences), a sheathpressure of 20 to 25 pounds per square inch, and anacquisition rate of 2,000 to 3,000 events per second wereused according to the guidelines for FACS Aria users (BDBiosciences).
Immunofluorescence stainingImmunocytochemical fluorescence labeling of cells was
conducted as previously described (24). Briefly, CD45�/CD31þ cells isolated by flow cytometry were cytospun ontoglass slides and washed 3 times with PBS for 3 minutes. Tostain intracellular antigens, cells were permeabilized with0.5% Triton X-100 (Sigma Chemical Co.). To prevent cross-reaction with antibodies used to stain cells in the flow-cytometric analysis, Fab-fragment blocking was conductedovernight at 4�C with an antibody from the same hostspecies of antibody, which was a F(ab0)2 fragment fromgoat anti-mouse immunoglobulin G (IgG) (Jackson Immu-noResearch Laboratories) in this study. After the blockingstep, cells were stained with the following antibodies:mouse anti-human CD105 monoclonal antibody (mAb;1:100, Serotec), mouse anti-human CD146 mAb (1:100dilution, Chemicon), mouse anti-human CD144 mAb(1:100, Reliatech GmbH), mouse anti-human CD68 mAb(1:100 dilution, DAKO), polyclonal rabbit anti-humanvWF Ab (1:100 dilution, DAKO), or mouse anti-humanCD14 mAb (1:20 dilution, DAKO), followed by labelingwith the corresponding secondary antibodies conjugatedwith FITC or Texas Red. For double-staining experiments,
Table 1. Antibody panels used for flow-cytometric analysis and isolation of CD45�/CD31þ cells
the protein and fragment blocking steps were repeatedbefore treating with the second primary antibody toprevent cross-reaction. Cell nuclei were counterstainedwith Hoechst 33342 dye (Invitrogen). The slides werethen washed 3 times with PBS for 3 minutes each andmounted in Vecta-shield (Vector Laboratories). Imageswere acquired with a LSM510 Meta DuoScan confocalsystem (Zeiss).
Statistical analysisThe Mann–Whitney U test was used to determine the
statistical significance of differences in the frequencies ofCEC candidates between the peripheral blood collectedfrom patients with cancer and healthy volunteers. All sta-tistical tests were two-sided. P values less than 0.05 wereconsidered significant.
ResultsEstablishment of flow cytometry gating strategies forthe measurement of circulating endothelial cellsFirst, we found that monocytes showed higher levels of
autofluorescence than lymphocytes (data not shown), indi-cating that they should be analyzed separately for fluores-cence compensation. In flow-cytometric analysis, the detec-tion of equal levels of autofluorescence in positive andnegative populations for each single stain indicates that thefluorescent compensation is appropriate (29). As an initialstep to establish the efficient gating strategies for the detec-tion of CECs by flow cytometry, we attempted to determinethe subset(s) of PBMCs expressing the CD45�/CD31þ
phenotype. PBMCs were plotted according to the forwardscatter (FSC) versus side scatter (SSC) profiles and FSClow
/SSClow (Fig. 1A, left), FSChigh/SSCmid (Fig. 1B, left), andFSCmid/SSChigh (data not shown) fractions were gated aslymphocytes, monocytes, or granulocyte subpopulations,respectively. Subpopulations of cells with different CD45and CD31 expression patterns were further analyzed for theexpression of CD146 or lineage-specific markers includingCD3 (T lymphocyte), CD14 (monocyte), CD19 (B lym-phocyte), and CD56 (NK cells). The FSClow/SSClow subsetwas mostly composed of CD45þ/CD31� cells expressingCD3, CD19, or CD56 antigens (Fig. 1A), whereas theFSChigh/SSCmid subset mainly included CD45þ/CD31þ
cells expressing the CD14 antigen (Fig. 1B). CD45�/CD31þ
cells were detected only in the FSClow/SSClow fraction (Fig.1A) but neither in the FSChigh/SSCmid (Fig. 1B) nor FSCmid/SSChigh fractions (data not shown).In the polychromatic flow-cytometric analysis used in the
present study, adequate threshold was assessed by fluores-cence-minus-one (FMO) gating, which consists of analyz-ing cells stained with all antibodies except the one beingtested (30). To investigate the effects of the gating controlson the actual event frequencies, the flow-cytometric analysisof the singlet lymphocyte fraction was conducted usingeither the isotype control or the FMO control (Fig. 2A).Because the negative threshold of the FMO control (Fig. 2A,b) was higher than that of the isotype control (Fig. 2A, a),gating with FMO controls could decrease the false-positive
event frequencies. In line with the study by Cui and collea-gues (31), gating with FMO controls was shown to be amore efficient method to increase the accuracy and speci-ficity of the positive signals in polychromatic flow-cyto-metric detection of rare events, such asCECs, than the use ofisotype controls.
On the basis of these results, we established a gatingstrategy to determine the frequencies of CECs in PBMCs,as described in Fig. 2B. In brief, cells were stained witha panel of antibodies in parallel with FMO controls. Thesinglet lymphocyte population was identified on a FSC/SSC plot and subgated onto a bivariant antigen plot toidentify CD45�/CD31þ cells. These cells were further sub-gated to identify the corresponding CD146þ or CD105þ
subpopulation.
Flow-cytometric analysis of CECs in PBMCsA number of protein markers including CD31, CD34,
CD105, CD146, CD144, and VEGF receptor-2 have beenused to define CECs. However, there is no truly specificmarker to identify CECs because those markers are alsoexpressed in other type of cells (32). A generally accepteddefinition of CECs is CD45�/CD31þ cells expressingCD146 or CD105 (11), but this definition needs to bemodified. According to the gating strategy described above,we examined the frequencies of CD45�/CD31þ, CD45�/CD31þ/CD146þ, and CD45�/CD31þ/CD105þ cells in thePBMCs of patients with cancer (n ¼ 56) and healthyvolunteers (n ¼ 44) by flow cytometry. The frequency ofCD45�/CD31þ cells was significantly higher in the singletFSClow/SSClow population of patients with cancer (median,1.365%; range, 0.110–26.85%) than in that of healthyvolunteers (median, 0.183%; range, 0.027–3.980%; P <0.0001) as shown in Fig. 3A. In contrast with a previousreport (11), the frequency of CD45�/CD31þ/CD146þ cellsin healthy volunteers (median, 0%; range, 0–0.003%) andpatients with cancer (median, 0.001%; range, 0–0.016%)was lower than the cutoff values of the FMO control group(median, 0.007%; range, 0–0.021%), indicating that esti-mation of the frequency of those cells is not possible both incancer patients and healthy controls due to the limitedresolution of the flow cytometry. Actually, when the isotypecontrol was used, the frequencies of CD45�/CD31þ/CD146þ cells were significantly higher (median, 0.041%;range, 0.007–0.132%; Fig. 2C), underscoring the impor-tance of using FMO controls.
Meanwhile, CD146 expression was detected only in theCD45þ/CD31� subpopulation of the FSClow/SSClow popu-lations (Fig. 1A). These CD146-positive cells also expressedCD3 (data not shown). These results, together with those ofprevious studies showing that CD146 is present in a subsetof activated T lymphocytes (21, 33), indicate that CD146þ
cells in theCD45þ/CD31� lymphocyte subpopulation are Tlymphocytes.
The frequency of CD45�/CD31þ/CD105þ cells inhealthy volunteers (median, 0.003%; range, 0–0.027%)showed no significant differences statistically when com-pared with FMO control (P > 0.05). On the other hand, the
Characterization of Circulating Endothelial Cells
www.aacrjournals.org Clin Cancer Res; 19(19) October 1, 2013 5343
frequency of CD45�/CD31þ/CD105þ cells in patientswith cancer (median, 0.012%; range, 0–0.461%) was signi-ficantly higher than in healthy volunteers (P < 0.0001)as shown in Fig. 3B.
There was no significant statistical difference in the fre-quency of CD45�/CD31þ, CD45�/CD31þ/CD146þ, andCD45�/CD31þ/CD105þ cells between healthy male andfemale volunteers.
A
CD3
CD
45
CD
45+/C
D31
–C
D45
+/C
D31
+C
D45
– /C
D31
+C
D45
+/C
D31
–C
D45
+/C
D31
+C
D45
– /C
D31
+
CD45+/CD31–
CD45+/CD31+
CD45–/CD31+
CD45+/CD31–
CD45+/CD31+
CD45–/CD31+
CD
56
CD
45
CD
45
CD
45
CD
56
CD
45
CD
45
CD
45
CD
56
CD
45
CD
45
CD
45
CD
56
CD
45
CD
45
CD
45
CD
56
CD
45
CD
45
CD
45
CD
56
CD
45
CD
45
CD14 CD19 CD146
CD3 CD14 CD19 CD146
CD3 CD14 CD19 CD146
CD3 CD14 CD19 CD146
CD3 CD14 CD19 CD146
CD3 CD14 CD19 CD146
B
Figure 1. CD45 andCD31 expression in PBMCs obtained frompatients with cancer. PBMCswere initially gated into (A) FSClow/SSClow and (B) FSCmid/SSCmid
fractions to includemainly lymphocytes andmonocytes, respectively, and singlet cells were selected on the basis of FSC-Height versus FSC-Area plots. Thecells were then subdivided intoCD45þ/CD31�, CD45þ/CD31þ, andCD45�/CD31þ cells. The cellswere further analyzed for the expressionof lineage-specificmarkers (CD3, CD14, CD19, and CD56) and CD146. Lymphocytes gated on FSC/SSC plots were mostly CD45þ/CD31� (A, left), whereas monocytesweremostly CD45þ/CD31þ (B, left). In addition, CD45þ/CD31� cells were composed of CD3þ, CD19þ, CD56þ, andCD146þ cells (A andB, top right), whereasCD45þ/CD31þ cells were mostly CD14þ (A and B, middle right), regardless of lymphocyte and monocyte gating based on the FSC/SSC plot. In particular,FSClow/SSClow gating showed CD45�/CD31þ cells (A and B, left) that did not express lineage-specific markers or CD146 (A and B, bottom right).
Yu et al.
Clin Cancer Res; 19(19) October 1, 2013 Clinical Cancer Research5344
Characterization of anucleated CD45�/CD31þ cellsTo date, several methods have been used to quantify
CECs. However, there is currently no consensus on themost accurate markers for their identification. Moreover,CEC quantification methods have not been adequatelyvalidated or standardized. For these reasons, there has beena significant variation in theCECnumbers reported, includ-ing those of the present study. To determine whether theCEC candidate cells identified by flow cytometry have trueendothelial phenotypes, we isolated CD45�/CD31þ cellsfrom the PBMCs of patients with cancer by FACSAria sorterand examined their morphologic and immunologic char-acteristics. CD45þ/CD31� cells (mostly lymphocytes) andCD45þ/CD31þ cells (mostlymonocytes) were also isolatedand used as controls.Two different populations of CD45�/CD31þ cells were
observed by confocal microscopy and scanning electron
microscopy, anucleated cells (2–6 mm in diameter), whichwere smaller than lymphocytes, and nucleated cells (8–10mm in diameter; data not shown). Strijbos and colleaguesreported that the vast majority of CD45�/CD31þ cells are,in fact, large platelets rather than endothelial cells (17). Toinvestigate the possibility of false-positive results of theflow-cytometric quantification of CECs, we assessedCD45�/CD31þ cells for the expression of platelet markerssuch as CD41 and CD61 by flow cytometry and found thatmost CD45�/CD31þ cells (more than 98%) stained posi-tive for CD41 and CD61 (Fig. 4A and B, respectively).Furthermore, anucleated CD45�/CD31þ cells stained neg-ative for anti-CD146, anti-CD105, and anti-CD144 anti-bodies but positive for vWF, a common marker of endo-thelial cells and platelets (Fig. 5B, top) as assessed byimmunocytochemical staining and confocal microscopy.These results together with the morphologic phenotypes of
A Isotype control
IgG1-FITC
FSC-A
SS
C-A
FSC-A
FS
C-H
CD
146-
PE
CD
105-
PE
% o
f ce
lls
CD
45-P
erC
P
CD45–/CD31+ cells
IgG
1-P
E
IgG1-P
E
CD31-FITC
CD31-FITC
CD31-FITC CD31-FITC
FMO Control
Usingisotype control
UsingFMO control
B
C 0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
Figure 2. Gating strategy for thedetection of CEC candidates. A,quantile contour plots showedan increased fluorescencebackground in the FMO control(right) compared with the isotypecontrol (left). B, gating strategy forthe detection of CD45�/CD31þ,CD45�/CD31þ/CD146þ, andCD45�/CD31þ/CD105þ cells byflow cytometry. First, FSClow
/SSClow fractions were gated onthe basis of FSC-A/SSC-A plots,and singlet cells were selected onthe basis of FSC-H/FSC-A plots.Next, CD45�/CD31þ cells wereinitially gated. CD45�/CD31þ/CD146þ or CD45�/CD31þ/CD105þ cells were furtheranalyzed on the basis of FMOcontrol gating. C, graphs showedthat the percentage of false-positive cells was significantlyincreased in using isotype controlas compared with FMO control.Median value was indicated as ahorizontal bar.
Characterization of Circulating Endothelial Cells
www.aacrjournals.org Clin Cancer Res; 19(19) October 1, 2013 5345
these cells suggest that anucleated CD45�/CD31þ cells aremainly platelets.
Characterization of nucleated CD45�/CD31þ cellsTo determine whether endothelial cells are included
in the nucleated CD45�/CD31þ cell population, CD45�/CD31þ cells were isolated as described previously andstained with antibodies against CD146 (34), CD105 (35),
vWF (36), and CD144 (37). Consistent with the flow-cytometric results, most nucleated CD45�/CD31þ cellsexpressed CD105 but not CD146 (Fig. 5A, top). TheseCD45�/CD31þ/CD105þ cells also expressed CD144 (Fig.5A, bottom).
An increasing body of evidence indicates that monocytesshare several functional and immunophenotypic character-istics with endothelial cells (38, 39). Moreover, endothelialprogenitor cells or circulating angiogenic cells derived frommonocyte/macrophage lineage were reported to promoteangiogenesis by secreting angiogenic growth factors (40).To determine whether CD45�/CD31þ/CD105þ cells pos-sess monocyte/macrophage characteristics in addition tothose of endothelial cells, cells were costained with an anti-CD68 antibody (macrophage marker) and the anti-vWFantibody. Interestingly, CD45�/CD31þ/CD105þ cellsstained positive for CD68 (Fig. 5B, bottom) but negativefor vWF (Fig. 5B, top). Furthermore, these cells expressedthe monocyte-specific antigen CD14 (Fig. 5C). CD45þ/CD31� cells (mostly lymphocytes) were used as a negativecontrol for CD68, vWF, and CD14 expression, whereasCD45þ/CD31þ cells (mostly monocytes) were used as apositive control for CD68 andCD14 expression (Fig. 5C forCD14 expression). Human umbilical vein endothelial cellsand human dermal microvascular endothelial cells wereused as positive controls for vWF, CD146, CD105, andCD144 (data not shown).
The specificity of sorted CD45�/CD31þ cell populationcould be confirmed and the possibility of contamination ofsorted cell population with CD45þ/CD31þ cells could beexcluded because CD45þ/CD31þ cells were not detectedwhen the sorted cells by CD45�/CD31þ gate were analyzedby the CD45/CD31 expression (Supplementary Fig. S1).
Collectively, these data indicate that anucleated CD45�/CD31þ cells were mainly platelets, and CD45�/CD31þ/CD105þ cells were derived from monocytes/macrophagesrather than endothelial cells.
DiscussionCECs in the peripheral blood have been widely recog-
nized as a marker of angiogenesis. Most studies haverelied on multiparametric flow cytometry to identifyendothelial cells because of the limited specificity of themarkers used for CEC detection, which can also beexpressed by other hematopoietic cells (16, 39). CECsare currently defined as cells that express CD31 and othermarkers such as CD146, CD34, or CD105, but not CD45,a pan-leukocyte marker (7, 11). However, their rarenessand phenotypic overlap with other hematopoietic cellshave led to controversies about the identification anddetermination of CECs in peripheral blood. Therefore,the present study focused on assessing the accuracy ofcurrent flow-cytometric techniques for the identificationof CECs and the immunofluorescence phenotyping ofCECs using different markers. Our results suggest thatthose cells that meet the conventional criteria to bedefined as CECs are derived from a monocyte lineagerather than having an endothelial origin.
A
30
20
100.050
0.025
0.000
0.50.40.30.20.1
P < 0.0001
% o
f cel
ls
% o
f cel
ls
P < 0.0001
P < 0.0001
5.5
3.0
0.5
B
Healthy subjects
Cancerpatients
FMOControl
Healthysubjects
Cancerpatients
Figure 3. Differential levels of CEC candidates in patients withgynecologic cancer (n¼56) andhealthy volunteers (n¼44). PBMCswereanalyzed by three-color flow cytometry followed by the gating strategydescribed in Fig. 2B. The frequencies of (A) CD45�/CD31þ and (B)CD45�/CD31þ/CD105þ cells in singlet FSClow/SSClow populationsfrompatientswith cancer andhealthy volunteers aredepictedasa scatterplot. The numbers of CD45�/CD31þ and CD45�/CD31þ/CD105þ cellswere significantly higher in patientswith cancer than in healthy volunteers(A and B, respectively). Differences in CEC candidate levels betweenpatients with cancer and healthy volunteers were assessed by theMann–Whitney U test. Median value was indicated as a horizontal bar.
A
B
CD
45-P
erC
PC
D45
-Per
CP
CD31-FITC
CD
61-F
ITC
CD31-PE CD61-PE
CD31-FITC
CD
41-P
E
Figure 4. CD41 and CD61 expression in CD45�/CD31þ cells. Theexpressionof theplateletmarkersCD41andCD61 inCD45�/CD31þcellswas analyzed by three-color flow cytometry, followed by the gatingstrategy shown in Fig. 2. Quantile contour plots indicated that mostCD45�/CD31þ cells expressed CD41 and CD61.
Yu et al.
Clin Cancer Res; 19(19) October 1, 2013 Clinical Cancer Research5346
Autofluorescence is an important consideration whenconducting polychromatic flow-cytometric analyses.Because monocytes have higher levels of autofluorescencethan lymphocytes, these 2 types of cells should be analyzedseparately (29). The frequency of CECs has mostly beenanalyzed by gating the lymphocyte and monocyte popula-tions as awhole without separation, which can lead to false-negative or false-positive results. To overcome this potentialdefect, we first gated FSClow/SSClow fraction (mostly lym-phocytes) and then subgated CD45�/CD31þ cells for amore detailed and accurate characterization. In addition,we also showed that FMO is a more accurate control to setthe boundaries for the analysis of rare cells such as CECs asdescribed previously (31).
Several studies have shown evidences that CECs, definedas DNAþ cells with CD34þ/CD45�/CD146þ or CD31þ/CD45�/CD146þ immunophenotypes, are significantlyelevated in the peripheral blood of patients with cancerthan healthy subjects (41, 42). The origin of the cells wasconfirmed as endothelial cells by morphology, immuno-histochemistry, gene expression, and the presence of Wei-bel–Palade bodies. However, there have been scientificissues to be improved that the effects of various fluorescentcompensation methods on the detection of CECs were notconsidered.Moreover, expression of several lineagemarkersother than endothelial cell origin should be fully investi-gated to identify the genuine origin of those cells, becausethe immunophenotypes can be overlapped among cells
Figure 5. Immunocytochemicalstaining of CD45�/CD31þ cells.CD45�/CD31þ cells sorted byflow cytometry were cytospun onglass slides, fixed with 4% PFA,permeabilized with 0.1% TritonX-100 for CD68 and vWF staining,and then labeled with antibodiesagainst the endothelial cellmarkersCD105, CD146, CD144, and vWFor the monocyte markers CD68and CD14. A, double staining ofCD45�/CD31þ cells with anti-CD146 or CD144 and anti-CD105Abs. CD45�/CD31þ cellsexpressed CD105 and CD144(bottom) but not CD146 (top).Nuclei were counterstained withHoechst 33342 dye (originalmagnification, �630). B, doublestaining of CD45�/CD31þ cellswith anti-vWF or CD68 and anti-CD105 Abs. Nucleated CD45�/CD31þ cells expressed CD68(bottom) but not vWF (top),whereas anucleated CD45�/CD31þ cells expressed onlyvWF. Nuclei were counterstainedwith Hoechst 33342 dye (originalmagnification, �630). C, singlestaining of CD45�/CD31þ cellswith anti-CD14 Ab. NucleatedCD45�/CD31þ cells expressedCD14. CD45þ/CD31� and CD45þ/CD31þ cells were used as negativeand positive controls for thedetection of CD14 expression,respectively. The control slide wasstained with only fluorochrome-conjugated secondary Ab. Nucleiwere counterstained with Hoechst33342 dye (original magnification,�630).
Characterization of Circulating Endothelial Cells
www.aacrjournals.org Clin Cancer Res; 19(19) October 1, 2013 5347
derived from endothelial cells, hematopoietic progenitors,or monocytes, etc.
On the basis of the cell sorting conditions used, ourresults did not agreewith previous studies (11) inwhich fewCD45�/CD31þ/CD146þ cells, which are known to defineCECs, were present in the peripheral blood, even in that ofpatients with cancer. Although the frequency of CD45�/CD31þ/CD146þ cells seemed to be significant if isotypecontrols were used, they remained at background levelsin the presence of FMO controls. In addition, our immu-nocytochemical data showed that none of the isolatedCD45�/CD31þ cells expressed CD146. CD146 expressionhas been reported on activated T-cell subsets in healthyindividuals (33). Previously, Duda and colleagues reportedthat CD146þ marks endothelial cells in normal and neo-plastic tissues, as well as a subset of T cells (21). Theseauthors reported that CD146 expression was frequentlydetected on pericytes, and CD45þ/CD146þ cells were occa-sionally contained in the massive hematopoietic cell infil-tration in tumor tissues. Our results were in agreement withthose of Duda and colleagues in that CD146þ cells weredetected exclusively in the CD3þ cell population.
We could not detect the CD45�/CD31þ/CD146þ cells inthe peripheral blood of patients with gynecologic cancer;however, it remains to be further tested and comparedamong patients with different types or sites ofmalignancies.For instance, endothelial cells in hemangioma tissues sho-wed negative immunoreactivity for CD146 (43), whereasthe expression of CD146 was highly increased in the bloodvessels of breast carcinoma (44).
Because flow cytometry is not sensitive enough to obtainreproducible results when analyzing rare cells such as CECs,the results of these analyses should be interpreted withcaution (45). The frequency ofCECs in the peripheral bloodof the healthy population is between 1� 10�7 and 1� 10�5
per leukocyte (0–20 cells/mL of venous blood; refs. 20, 46).This level is below the detection threshold for conventionalflow cytometry, which is approximately 1 � 10�4 (45). Wetherefore assumed that the frequency of CECs may be toolow for detection by flow cytometry, despite their presenceamong PBMCs.
On the other hand, the frequency of CD45�/CD31þ/CD105þ cells (another potential CEC candidate) was sig-nificantly higher in patients with cancer than in the healthypopulation. Furthermore, the frequency ofCD45�/CD31þ/CD105þ cells was significantly increased by up to approx-imately 8-fold in patients with gynecologic cancer com-pared with healthy volunteers. CD105 (endoglin), a 180kDa homodimeric integral membrane glycoprotein and acommonly used marker for the detection of CECs next toCD146, was mainly expressed on endothelial cells ofcapillaries, veins, and arteries (35) but was also detectableon activated monocytes, macrophages, erythroid precur-sors, fibroblasts, mesangial cells, follicular dendritic cells,and syncytiotrophoblasts (47). Interestingly, our immuno-cytochemistry data showed that these CD45�/CD31þ/CD105þ cells did not express vWF (36), whereas they werepositive for CD144 (VE-cadherin; ref. 37). Actually, mono-
cytes express not only CD31 constitutively (48) but alsoother markers including CD144, KDR, Tie-2, and CD105when they are activated (49). Therefore, we further analyzedCD45�/CD31þ/CD105þ cells with antibodies againstmonocyte-specific markers such as CD14 and CD68, andshowed that CD45�/CD31þ/CD105þ cells uniformlyexpressed both CD14 and CD68. This result indicates thatCD45�/CD31þ/CD105þ cells have the characteristics ofmonocytes rather than those of endothelial cells (Supple-mentary Fig. S2). Prokopi and colleagues showed that theendothelial phenotype may arise in mononuclear cellsthrough the uptake of platelet microparticles (derived fromthe disintegration of platelets during mononuclear cellpreparation) that abundantly containmarker proteins suchasCD31andvWF(50).More importantly,monocytes in theperipheral blood are known to exhibit versatile and flexibledifferentiation potentials and functions, as they can differ-entiate into macrophages, dendritic cells, osteoclasts,microglia, or even endothelial-like cells (51). Myeloid line-age cells including monocytes were reported to expressmarkers that were expressed in both endothelial cells andmonocytes (22, 49, 52–54). Furthermore, circulatingCD31þ cells, which can contribute to the development ofnew blood vessels, were shown to be monocytes in anelaborate animal model (24).
In summary, our results showed that CD45�/CD31þ/CD105þ circulating cells detected by flow cytometry undergating conditions we established, which were significantlyincreased in the peripheral blood of patients with gyneco-logic cancer, were not CECs but rather monocytes, suggest-ing that the conventional flow-cytometric techniques usedfor the identification of cell subpopulations could beimproved by adjusting gating conditions. Further study isneeded to identify thebiologic significanceof these cells andtheir function in relation to angiogenesis.
Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.
Authors' ContributionsConception and design: H.-K. Yu, J.-S. Kim, S.J. Kim, T.J. KimDevelopment of methodology: H.-K. Yu, J.-H. Ahn, L.S.H. Yi, J.-S. KimAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): H.-K. Yu, H.-J. Lee, H.-N. Choi, J.-Y. Choi, H.-S.Song, K.-H. Lee, S.J. Kim, T.J. KimAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): H.-K. Yu, H.-N. Choi, J.-Y. Choi, H.-S.Song, J.-S. Kim, S.J. Kim, T.J. KimWriting, review, and/or revision of the manuscript: H.-K. Yu, L.S.H. Yi,J.-S. Kim, S.J. Kim, T.J. KimAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): H.-N. Choi, H.-S. Song, J.-S. Kim,S.J. KimStudy supervision: Y. Yoon, L.S.H. Yi, J.-S. Kim, S.J. Kim, T.J. Kim
Grant SupportThis work was financially supported by a grant from the Korea Health 21
R&D Project, Ministry for Health, Welfare and Family Affairs, Republic ofKorea (A050905), by a grant from the KRIBB Research Initiative Program,and by Basic Science Research Programs through the National ResearchFoundation of Korea funded by the Ministry of Education, Science andTechnology (2009-0073284 and 2012R1A1A2007994).
The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby marked
Yu et al.
Clin Cancer Res; 19(19) October 1, 2013 Clinical Cancer Research5348
and implications. Science 1982;217:998–1003.2. Jung YD, Ahmad SA, Akagi Y, Takahashi Y, Liu W, Reinmuth N, et al.
Role of the tumor microenvironment in mediating response to anti-angiogenic therapy. Cancer Metastasis Rev 2000;19:147–57.
3. Folkman J. Tumor angiogenesis: therapeutic implications. N Engl JMed 1971;285:1182–6.
4. Hurwitz H, Fehrenbacher L, Novotny W, Cartwright T, Hainsworth J,Heim W, et al. Bevacizumab plus irinotecan, fluorouracil, andleucovorin for metastatic colorectal cancer. N Engl J Med 2004;350:2335–42.
5. Goodman VL, Rock EP, Dagher R, Ramchandani RP, Abraham S,Gobburu JV, et al. Approval summary: sunitinib for the treatment ofimatinib refractory or intolerant gastrointestinal stromal tumors andadvanced renal cell carcinoma. Clin Cancer Res 2007;13:1367–73.
6. SessaC, Guibal A, Del ConteG, RueggC. Biomarkers of angiogenesisfor the development of antiangiogenic therapies in oncology: tools ordecorations? Nat Clin Pract Oncol 2008;5:378–91.
7. Bertolini F, Shaked Y, Mancuso P, Kerbel RS. The multifacetedcirculating endothelial cell in cancer: towards marker and targetidentification. Nat Rev Cancer 2006;6:835–45.
8. Li N, Zheng D, Wei X, Jin Z, Zhang C, Li K. Effects of recombinanthuman endostatin and its synergy with cisplatin on circulating endo-thelial cells and tumor vascular normalization in A549 xenograft murinemodel. J Cancer Res Clin Oncol 2012;138:1131–44.
9. Shaked Y, Bertolini F, Man S, Rogers MS, Cervi D, Foutz T, et al.Genetic heterogeneity of the vasculogenic phenotype parallels angio-genesis; Implications for cellular surrogate marker analysis of anti-angiogenesis. Cancer Cell 2005;7:101–11.
10. Monestiroli S,Mancuso P, Burlini A, Pruneri G, Dell'Agnola C, Gobbi A,et al. Kinetics and viability of circulating endothelial cells as surrogateangiogenesis marker in an animal model of human lymphoma. CancerRes 2001;61:4341–4.
11. Mancuso P, Burlini A, Pruneri G, Goldhirsch A, Martinelli G, Bertolini F.Resting and activated endothelial cells are increased in the peripheralblood of cancer patients. Blood 2001;97:3658–61.
12. Furstenberger G, von Moos R, Lucas R, Thurlimann B, Senn HJ,Hamacher J, et al. Circulating endothelial cells and angiogenic serumfactors during neoadjuvant chemotherapy of primary breast cancer. BrJ Cancer 2006;94:524–31.
13. Bhatt RS, Zurita AJ, O'Neill A, Norden-Zfoni A, Zhang L, Wu HK, et al.Increased mobilisation of circulating endothelial progenitors in vonHippel-Lindau disease and renal cell carcinoma. Br J Cancer 2011;105:112–7.
14. Ronzoni M, Manzoni M, Mariucci S, Loupakis F, Brugnatelli S, Ben-cardino K, et al. Circulating endothelial cells and endothelial progeni-tors as predictive markers of clinical response to bevacizumab-basedfirst-line treatment in advanced colorectal cancer patients. Ann Oncol2010;21:2382–9.
15. Mancuso P, Colleoni M, Calleri A, Orlando L, Maisonneuve P, PruneriG, et al. Circulating endothelial-cell kinetics and viability predict sur-vival in breast cancer patients receiving metronomic chemotherapy.Blood 2006;108:452–9.
16. Schmeisser A, Strasser RH. Phenotypic overlap between hematopoi-etic cells with suggested angioblastic potential and vascular endo-thelial cells. J Hematother Stem Cell Res 2002;11:69–79.
17. Strijbos MH, Kraan J, den Bakker MA, Lambrecht BN, Sleijfer S,Gratama JW. Cells meeting our immunophenotypic criteria of endo-thelial cells are large platelets. Cytometry B Clin Cytom 2007;72:86–93.
18. DomeB, Timar J, Ladanyi A, PakuS, Renyi-Vamos F, KlepetkoW, et al.Circulating endothelial cells, bone marrow-derived endothelial pro-genitor cells and proangiogenic hematopoietic cells in cancer: frombiology to therapy. Crit Rev Oncol Hematol 2009;69:108–24.
19. Goon PK, Watson T, Shantsila E, Boos CJ, Lip GY. Standardization ofcirculating endothelial cell enumeration by the use of human umbilicalvein endothelial cells. J Thromb Haemost 2007;5:870–2.
20. Erdbruegger U, Haubitz M, Woywodt A. Circulating endothelialcells: a novel marker of endothelial damage. Clin Chim Acta2006;373:17–26.
21. Duda DG, Cohen KS, di Tomaso E, Au P, Klein RJ, Scadden DT, et al.Differential CD146 expression on circulating versus tissue endothelialcells in rectal cancer patients: implications for circulating endothelialand progenitor cells as biomarkers for antiangiogenic therapy. J ClinOncol 2006;24:1449–53.
22. Conejo-Garcia JR,BuckanovichRJ,BenenciaF,CourregesMC,RubinSC, Carroll RG, et al. Vascular leukocytes contribute to tumor vascu-larization. Blood 2005;105:679–81.
23. Urbich C, Heeschen C, Aicher A, Dernbach E, Zeiher AM, DimmelerS. Relevance of monocytic features for neovascularization capacityof circulating endothelial progenitor cells. Circulation 2003;108:2511–6.
24. KimSJ, Kim JS, Papadopoulos J,Wook KimS,MayaM, Zhang F, et al.Circulating monocytes expressing CD31: implications for acute andchronic angiogenesis. Am J Pathol 2009;174:1972–80.
25. Duda DG, Cohen KS, Scadden DT, Jain RK. A protocol for phenotypicdetection and enumeration of circulating endothelial cells and circu-lating progenitor cells in human blood. Nat Protoc 2007;2:805–10.
26. Norden-Zfoni A, Desai J, Manola J, Beaudry P, Force J, Maki R, et al.Blood-based biomarkers of SU11248 activity and clinical outcome inpatients with metastatic imatinib-resistant gastrointestinal stromaltumor. Clin Cancer Res 2007;13:2643–50.
27. Starlinger P, Brugger P, Reiter C, Schauer D, Sommerfeldt S,Tamandl D, et al. Discrimination between circulating endothelialcells and blood cell populations with overlapping phenotype revealsdistinct regulation and predictive potential in cancer therapy. Neo-plasia 2011;13:980–90.
28. Wierzbowska A, Robak T, Krawczynska A, Pluta A, Wrzesien-Kus A,Cebula B, et al. Kinetics and apoptotic profile of circulating endothelialcells as prognostic factors for induction treatment failure in newlydiagnosedacutemyeloid leukemiapatients. AnnHematol 2008;87:97–106.
29. Robinson JP. Current protocols in cytometry. New York: Wiley; 199830. Tung JW, Heydari K, Tirouvanziam R, Sahaf B, Parks DR, Herzenberg
LA. Modern flow cytometry: a practical approach. Clin Lab Med2007;27:453–68.
31. Cui YX, Johnson T, BaumbachA, ReevesBC, Rogers CA, Angelini GD,et al. Stepwise optimization of the procedure for assessment ofcirculating progenitor cells in patients withmyocardial infarction. PLoSONE 2012;7:e30389.
32. McLean K, Buckanovich RJ. Myeloid cells functioning in tumorvascularization as a novel therapeutic target. Transl Res 2008;151:59–67.
33. Pickl WF, Majdic O, Fischer GF, Petzelbauer P, Fae I, Waclavicek M,et al. MUC18/MCAM (CD146), an activation antigen of human Tlymphocytes. J Immunol 1997;158:2107–15.
34. George F, Brisson C, Poncelet P, Laurent JC, Massot O, Arnoux D,et al. Rapid isolation of human endothelial cells fromwhole blood usingS-Endo1 monoclonal antibody coupled to immuno-magnetic beads:demonstration of endothelial injury after angioplasty. Thromb Hae-most 1992;67:147–53.
35. Strijbos MH, Verhoef C, Gratama JW, Sleijfer S. On the origin of(CD105þ) circulating endothelial cells. Thromb Haemost 2009;102:347–51.
36. Mannucci PM. Platelet von Willebrand factor in inherited and acquiredbleeding disorders. Proc Natl Acad Sci U S A 1995;92:2428–32.
37. Berx G, van Roy F. Involvement of members of the cadherin super-family in cancer. Cold Spring Harb Perspect Biol 2009;1:a003129.
Characterization of Circulating Endothelial Cells
www.aacrjournals.org Clin Cancer Res; 19(19) October 1, 2013 5349
39. Schmeisser A, Graffy C, Daniel WG, Strasser RH. Phenotypic overlapbetween monocytes and vascular endothelial cells. Adv Exp Med Biol2003;522:59–74.
40. Rehman J, Li J, Orschell CM, March KL. Peripheral blood "endothelialprogenitor cells" are derived from monocyte/macrophages andsecrete angiogenic growth factors. Circulation 2003;107:1164–9.
41. Kraan J, Strijbos MH, Sieuwerts AM, Foekens JA, den Bakker MA,Verhoef C, et al. A new approach for rapid and reliable enumeration ofcirculating endothelial cells in patients. J Thromb Haemost 2012;10:931–9.
42. Mancuso P, Antoniotti P, Quarna J, Calleri A, Rabascio C, Tacchetti C,et al. Validation of a standardized method for enumerating circulatingendothelial cells and progenitors: flow cytometry and molecular andultrastructural analyses. Clin Cancer Res 2009;15:267–73.
43. Li Q, Yu Y, Bischoff J, Mulliken JB, Olsen BR. Differential expression ofCD146 in tissues and endothelial cells derived from infantile haeman-gioma and normal human skin. J Pathol 2003;201:296–302.
44. Li W, Yang D, Wang S, Guo X, Lang R, Fan Y, et al. Increasedexpression of CD146 and microvessel density (MVD) in invasivemicropapillary carcinoma of the breast: comparative study with inva-sive ductal carcinoma-not otherwise specified. Pathol Res Pract2011;207:739–46.
45. Craig FE, Foon KA. Flow cytometric immunophenotyping for hema-tologic neoplasms. Blood 2008;111:3941–67.
46. Woywodt A, Streiber F, de Groot K, Regelsberger H, Haller H, HaubitzM.Circulatingendothelial cells asmarkers forANCA-associated small-vessel vasculitis. Lancet 2003;361:206–10.
47. Fonsatti E, Nicolay HJ, Altomonte M, Covre A, Maio M. Targetingcancer vasculature via endoglin/CD105: a novel antibody-based diag-nostic and therapeutic strategy in solid tumours. Cardiovasc Res2010;86:12–9.
48. Newman PJ. The biology of PECAM-1. J Clin Invest 1997;100:S25–9.49. SchmeisserA,GarlichsCD,ZhangH,EskafiS,GraffyC, Ludwig J, et al.
Monocytes coexpress endothelial and macrophagocytic lineage mar-kers and form cord-like structures in Matrigel under angiogenic con-ditions. Cardiovasc Res 2001;49:671–80.
50. Prokopi M, Pula G, Mayr U, Devue C, Gallagher J, Xiao Q, et al.Proteomic analysis reveals presence of platelet microparticles inendothelial progenitor cell cultures. Blood 2009;114:723–32.
51. Gordon S, Taylor PR. Monocyte and macrophage heterogeneity. NatRev Immunol 2005;5:953–64.
52. Seta N, Kuwana M. Human circulating monocytes as multipotentialprogenitors. Keio J Med 2007;56:41–7.
54. Li B, Pozzi A, Young PP. TNFalpha accelerates monocyte to endo-thelial transdifferentiation in tumors by the induction of integrin alpha5expression and adhesion to fibronectin. Mol Cancer Res 2011;9:702–11.
Yu et al.
Clin Cancer Res; 19(19) October 1, 2013 Clinical Cancer Research5350
2013;19:5340-5350. Published OnlineFirst August 6, 2013.Clin Cancer Res Hyun-Kyung Yu, Ho-Jeong Lee, Ha-Na Choi, et al. Peripheral Blood of Patients with Gynecologic Malignancies
Circulating Cells in the+/CD105+/CD31−Characterization of CD45
Updated version
10.1158/1078-0432.CCR-12-3685doi:
Access the most recent version of this article at:
