BASIC IMMUNOLOGY doi: 10.1111/j.1365-3083.2008.02091.

x
..................................................................................................................................................................

Expression of CD68 in Non-Myeloid Cell Types

E. Gottfried*, L. A. Kunz-Schughart , A. Weberà, M. Rehli*, A. Peuker*, A. Müller*,
M. Kastenberger*, G. Brockhoff§, R. Andreesen* & M. Kreutz*

*Department of Hematology and Oncology,
University of Regensburg, Regensburg;
 OncoRay – Center for Radiation Research in
Oncology, University of Technology of Dresden,
Dresden; àPaediatric Clinic of the University of
Würzburg, Würzburg; and §Institute of
Pathology, University of Regensburg,
Regensburg, Germany
Received 16 October 2007; Accepted in revised
form 21 January 2008
Correspondence to: E. Gottfried, PhD,
University of Regensburg, Department of
Hematology and Oncology, 93042 Regensburg,
Germany. E-mail: eva.gottfried@klinik.
uni-regensburg.de
Abstract
CD68, the human homologue of macrosialin, is commonly regarded as a selec-
tive marker for human monocytes and macrophages. Its expression is thought
to be regulated by a macrophage-specific promoter. However, several immuno-
histochemical studies have indicated that CD68 antibodies also react with
other haematopoietic and non-haematopoietic cell types. We investigated the
expression of CD68 in various primary cells and carcinoma cell lines using
immunohistochemistry, flow cytometry, Western blot analysis and qRT-PCR.
Weak but significant immunoreactivity was detected in lymphocytes and sev-
eral tumour cell lines whereas staining of primary fibroblasts and endothelial
cells was comparable to macrophages. The intensity of CD68 staining in indi-
vidual cell types depended on the antibody clone and the fixation technique.
Anti-CD68 mAb KP1 should be used with great caution for frozen tissue sec-
tions due to its reactivity with a wide variety of cell types. Also, care should
be taken when distinguishing macrophages from fibroblasts ⁄ stromal cells in
paraffin sections after formalin fixation since both cell types are stained highly
positive for CD68. In accordance, mRNA expression of CD68 was not only
detected in macrophages and monocytes but also in fibroblasts as well as endo-
thelial cells and tumour cells, although with a varying intensity. Cloning of
full length 5¢-sequences and determination of transcription start sites shows
that macrophages and fibroblasts initiate transcription within the known pro-
moter region; however, from different start sites, indicating alternative pro-
moter architecture in myeloid versus non-myeloid cells. We suggest that
CD68 is not a selective macrophage marker but rather a lysosomal protein that
is enriched in macrophages.

In contrast to other members of the lamp-family of

Introduction
CD68 is a highly glycosylated type I transmembrane gly-
coprotein that is mainly localized within the endosomal
compartment. Because of its intracellular localization [1]
and its characteristic structure, CD68 belongs to the lamp
(lysosomal-associated membrane protein) – family of gly-
coproteins [2]. Its function is still unknown but its pre-
dominant location within late endosomes suggests a role
in antigen processing or in the protection of lysosomal
membranes against lysosomal hydrolases [2, 3]. In thio-
glycollate-elicited peritoneal macrophages, 10–15% of
mouse CD68 ⁄ macrosialin is found on the cell surface and
exchange with intracellular pools occurs at very high
rates [4]. Surface CD68 binds and internalizes oxidized
LDL (low-density lipoprotein) in monocytic THP-1 cells,
an effect that is inhibited by monoclonal antibodies
against CD68 [5].
glycoproteins which are ubiquitously expressed in all cell
types, CD68 appeared to have a more restricted, cell-type
specific pattern of expression. Indeed, it is described to be
selective for the monocyte ⁄ macrophage lineage and anti-
bodies against CD68 are widely applied in diagnostics
and research as monocyte ⁄ macrophage specific markers.
Several monoclonal antibodies recognize the CD68 anti-
gen and were grouped together on the basis of pan-mac-
rophage reactivity on tissue sections: Ki-M6 [6], Ki-M7
[7], Y2 ⁄ 131 and Y1 ⁄ 82A [8], EBM11 [9], KP1 [10], Ki-
M1P [11], PG-M1 [12]. The specificity of these antibod-
ies for CD68 was confirmed earlier by Western blotting
and transfection studies [11–13]. However, in contrast to
the generally accepted concept of monocyte ⁄ macrophage
specificity, we and others recorded some reactivity of
anti-human CD68 antibodies in various haematopoietic
and non-haematopoietic cells [14–17]. Considering that

 2008 The Authors

Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463 453
454 CD68 Expression in Non-Myeloid Cells E. Gottfried et al.
..................................................................................................................................................................

this reactivity is not unspecific background but due to Table 1 Cells and cell lines.
specific antibody binding, the application of such anti-
Cell lines Origin of cell lines Source
bodies for the identification and quantification of mono-
cytes ⁄ macrophages may be critical, in particular for A594 Lung carcinoma ECACC
diagnostic purposes. The present study therefore aimed to BT474 Breast carcinoma ATCC
assess a broad spectrum of cell types recognized by anti- T47D Breast carcinoma ATCC
Capan1 Pancreas carcinoma ECACC
CD68 antibodies and to verify that these cells indeed Panc1 Pancreas carcinoma ECACC
express CD68 on protein and RNA levels. HT29 Colon adenocarcinoma ATCC
J82 Bladder carcinoma ATCC
RT-4 Bladder carcinoma ATCC
Materials and methods RJ494 Renal cell carcinoma Prof. A. Mackensen,
University of Erlangen,
Isolation and culture of primary cells. Human monocytes Germany
and lymphocytes were obtained by leucapheresis from MelIm Melanoma Prof. A. Mackensen,
healthy donors, followed by density gradient centrifuga- University of Erlangen,
tion over Ficoll ⁄ Hypaque and separation by counter- Germany
current elutriation (J6M-E centrifuge; Beckmann, Munich, Laz388 B-cell lymphoma Prof. A. Mackensen,
University of Erlangen,
Germany) [18, 19]. Monocyte preparations showed a purity Germany
of >85% pure as determined by expression of the monocyte THP-1 Monocytic leucaemia ATCC
antigen CD14. Cytometric analysis of the isolated lympho- Jurkat T-cell lymphoma ATCC
cyte fraction consistently showed about 80% CD3-positive HMEC-1 Human microvascular Prof. G. Eissner, Department
and 8% CD20-positive cells using established protocols endothelial cell line of Hematology ⁄ Oncology
Regensburg, Germany
with monoclonal FITC-conjugated mouse-anti human
antibodies (Dako, Hamburg, Germany). Macrophages were Origin of primary
generated by culturing purified monocytes for 7 days on Primary cells cells Source
teflon foils (Heraeus Hanau, Germany) with a seeding cell
HUVEC Primary human Prof. G. Eissner, Department
density of 106 cells ⁄ ml in RPMI 1640 (Biochrom, Berlin, umbilical vein of Hematology ⁄ Oncology,
Germany) supplemented with antibiotics (50 IU ⁄ ml endothelial cells Regensburg, Germany
penicillin and 50 lg ⁄ ml streptomycin), L-glutamine N1 fibroblasts Primary fibroblasts Institute of Pathology,
(2 mmol ⁄ l, Gibco, Paisley, NJ, USA) and 2% pooled from normal skin University of Regensburg,
Germany
human AB-group serum (BioWhittaker, Walkersville,
OA112 Primary synovial Department of Rheumatology,
MD, USA) [20]. Dendritic cells (DC) were generated by tissue of University of Gießen,
culturing freshly isolated monocytes for 7 days in culture osteoarthritis patients Germany
flasks in RPMI 1640 (Biochrom) supplemented with anti- PFN2 Primary fibroblasts Institute of Pathology,
biotics (50 IU ⁄ ml penicillin and 50 lg ⁄ ml streptomycin), from normal breast University of Regensburg,
Germany
L-glutamine (2 mmol ⁄ l, 10% FCS (Pan-Biotech, Aiden-
PF28 Primary fibroblasts Institute of Pathology,
bach, Germany) and 500 U ⁄ ml IL-4 (Promocell, Heidel- from breast University of Regensburg,
berg, Germany) and 500 U ⁄ ml GM-CSF (Novartis AG, carcinoma Germany
Basel, Switzerland). Primary fibroblasts were isolated from Monocytes ⁄ Primary cells Isolation and generation from
normal breast (PFN2) and breast carcinoma (PF28) as macrophages healthy human donors (see
Material and Methods)
described previously [21]. Fibroblasts from normal skin
Lymphocytes Primary cells Isolation from healthy human
(N1) and synovial tissue of osteoarthritis patients (OA112) donors (see Material and
were kindly provided by the groups of Prof. Dr. Schmitz Methods)
(Department of Clin. Chem., Univ. of Regensburg) and
Prof. Dr U. Mueller-Ladner (Univ. of Gießen). Fibroblasts (both bladder carcinoma, ATCC), Panc1 and Capan1 (both
were cultured in DMEM containing 10% FCS and pancreas carcinoma, ECACC), Jurkat (T cell-lymphoma),
100 IU ⁄ ml penicillin ⁄ 100 lg ⁄ ml streptomycin (all from THP-1 (monocytic leucaemia cell line, ATCC); laz388
PAN-Biotech). Endothelial cells from umbilical veins were (B cell-lymphoma), RJ494 (renal cell carcinoma) and
isolated as described by Jaffé et al. [22] and subsequently MelIm (melanoma) were donated by Prof. A. Mackensen,
recultured as monolayers in complete EGM-2 medium Department of Hematology ⁄ Oncology Regensburg. All
(EC; Cambrex Bio Science, Verviers, Belgium). Details cell lines were cultured in RPMI 1640 (BioWhittaker) or
about the cells are given in Table 1. DMEM (Gibco BRL, Carlsbad, CA, USA) plus 10% FCS,
Culture of cell lines. Experiments were carried out with L-glutamine and antibiotics as detailed above. The human
the following human cell lines: A549 (lung carcinoma, microvascular endothelial cell line HMEC was cultured in
ECACC), HT29 (colon adenocarcinoma, ATCC), BT474 complete EGM-2 medium (EC; Cambrex Bio Science).
and T47D (both breast carcinoma, ATCC), RT4 and J82 Details about the cells are given in Table 1.

 2008 The Authors

Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
E. Gottfried et al. CD68 Expression in Non-Myeloid Cells 455
..................................................................................................................................................................

CD68 analysis by flow cytometry. For intracellular CD68
staining and flow cytometric analyses, adherent cells were
detached using a 0.05% trypsin ⁄ 0.02% EDTA (Gibco
BRL) solution in PBS. Cell suspensions were washed in
PBS and resuspended in ice-cold 0.25% paraformalde-
hyde ⁄ PBS for a 10 min incubation interval. Cells were
then washed in PBS and transferred into FACS buffer
[PBS, 0.1% sodium azide, 0.6 mg ⁄ ml Sandoglobulin (CSL
Behring, Hattersheim, Germany)] supplemented with
0.1% saponin (‘saponin buffer’) for 20 min on ice. Incuba-
tion with primary mouse-anti-human antibodies or respec-
tive IgG isotype controls (each 1–1.2 lg per 4 · 105 cells
in 100 ll buffer) was performed for 45 min before sus-
pensions were washed twice with saponin buffer and
incubated with saturating amounts of a secondary fluores-
cein-isothiocyanate-(FITC)-conjugated goat-anti-mouse
antibody (Dianova, Hamburg, Germany) for 20 min at
4 C. Cells were then again washed twice and resuspended
in PBS for immediate flow cytometric analysis using a
FACScan flow cytometer and the Cellquest software (both
BD Biosciences, San Jose, CA, USA). Instrumental set-
tings were defined and kept constant for each cell type.
Details about antibodies are given in Table 2.
A double staining of macrophages and fibroblasts was
performed. After washing, cells were preincubated with
either PE-conjugated anti-CD14 antibody (BD Bio-
sciences), or FITC-conjugated anti-CD90 (anti-fibroblast)
antibody (Dianova), APC-conjugated anti-HLA-DR anti-
body (BD Biosciences) or respective isotype controls for
30 min on ice. After two washes cells were permeabilized
as described above and then incubated with a FITC-con-
jugated antibody against CD68 (KP1, Dako) for
30–60 min on ice. Cells were washed twice with saponin
buffer and resuspended in 1% paraformaldehyde ⁄ PBS for
immediate analysis as described above. Details about the
antibodies are given in Table 2.
Immunohistochemistry. For immunohistochemical stain-
ing of a pancreatic carcinoma site, paraffin sections were
stained with anti-CD68 monoclonal KP-1 antibody
(Dako, see Table 2) using a routine peroxidase-based
labeling technique and diaminobenzidine (DAB; AbD
Serotec, Kidlington, UK) as chromogen.
For immunostaining of primary cells and cell lines,
cells were trypsinized, washed with PBS and resuspended
at 1 · 106 cells ⁄ ml medium. Cell suspensions were pip-
etted onto adhesion slides coated with polysiloxane
(Superior, Paul Marienfeld KG, Bad Mergentheim,
Germany). Adherent cells were fixed with glutaraldehyde
for 10 min and stored at 4 C in a humidified chamber.
Staining was performed with anti-CD68 antibodies and
isotype controls as previously described using an APAAP
method with FastRed as chromogen [23].
Immunohistochemistry was also carried out on paraf-
fin-embedded materials. For staining of cultured cells,
cells were harvested, washed with PBS, transferred into
1.5 ml cups and pelleted. Cell pellets were processed for
paraffin histology using a Hypercenter XP (Shandon,
Frankfurt ⁄ Main, Germany) following fixation in 4% buf-
fered formalin. Pellets were finally embedded in paraffin
and were sliced into 5 lm serial sections. Sections were
immunohistochemically stained with the KP1 mAb
(working dilution 4 lg ⁄ ml) via a routine peroxidase-
based technique using the NexEs ⁄ HC module (Ventana
Medical Systems, Tuczon,Arizona) and DAB for color
development.
Western blot analysis. 1 · 106 freshly isolated macro-
phages or N1 fibroblasts were washed in 3 ml washing
buffer (14 mM NaH2PO4, 20 mM Na2HPO4, 10 mM
EDTA pH 8.0). The cell pellet was incubated in 50 ll
lysis buffer (14 mM NaH2PO4, 20 mM Na2HPO4,
10 mM EDTA pH 8.0, 1% mercaptoethanol, 1% SDS)
for 5–10 min at 95 C. The lysate can be frozen at this

Table 2 Primary and secondary antibodies

used for immunohistochemical detection of Working
CD68 expression. concentration
Specificity ⁄ clone Antibody Distributor (lg ⁄ ml)

Primary antibodies (anti-human):

CD68 – KP1 Mouse monoclonal IgG1 Dako 12
CD68 – KiM6 Mouse monoclonal IgG1 BMA Biomedicals 12
CD68 – PG-M1 Mouse monoclonal IgG2a Dako 12
CD68 – EBM11 Mouse monoclonal IgG1 Dako 12
CD68 – 7B8a Mouse polyclonal IgG2b ** 12
CD68 – KiM7 Mouse monoclonal IgG1 AbD Serotec 10
CD68 – KP1 (FITC) Mouse monoclonal IgG1 Dako 10
CD14 – MFP9 (PE) Mouse monoclonal IgG2b BD Biosciences 1,25
CD90 – AS02 (FITC) Mouse monoclonal IgG1 Dianova 1:200
HLA-DR – L243 (APC) Mouse monoclonal IgG2a BD Biosciences 0,3
Secondary antibodies:
FITC-conjugated goat-anti mouse (Fab) Dianova 70
PE-conjugated goat-anti mouse Dako 70

**Generated by our laboratory in collaboration with Prof. Dr Daniela Maennel, Institute of Immu-
nology, University of Regensburg, Germany.

 2008 The Authors

Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
456 CD68 Expression in Non-Myeloid Cells
..................................................................................................................................................................
step. After thawing, the lysate was incubated with 25 ll
dodecylmaltoside (20%) to neutralize SDS. 1 U N-glyco-
sidase-F (Sigma) was added to each sample and incubated
over night at 37 C. Samples were separated on a 10%
SDS-PAGE and proteins transferred onto nitrocellulose
(Amersham Pharmacia Biotech, Freiburg, Germany).
Membranes were blocked in TBST (Tris-buffered saline
with 0.05% Tween 20, 5% albumin) overnight at 4 C.
After three washes in Tris-buffered saline with 0.1%
Tween 20, membranes were incubated with KP1 anti-
body (Dako, 1:1000) in 5% non-fat dry milk (in aqua
bidest) for 1 h. Blots were then washed three times and
incubated with the secondary antibody (goat-anti-mouse-
HRP) in 5% milk for 1 h. Immunoreactive bands were
developed using a chemoluminescence kit (ECL, Amer-
sham Pharmacia Biotech). Identical protein lysate aliqu-
ots stained with a polyclonal b-actin antiserum (Sigma,
1:2000) served as loading control.
Quantitative reverse transcription, PCR. RNA was
extracted from 5 · 106 adherent or pelleted cells using
the RNeasy isolation system from Qiagen (Hilden, Ger-
many) according to the manufacturer’s instructions. RNA
was quantified via UV ⁄ VIS spectrometry (Nanodrop,
Wilmington, DE, USA). Two micrograms of RNA were
reverse transcribed in final volumes of 40 ll using 2 ll
of Superscript II (Gibco BRL). qRT-PCR was per-
formed with 2 ll cDNA per 20 ll SYBR Green-Mix
(QuantiTect SYBR Green PCR Kit, Qiagen, Hilden,
Germany), 1 mM primers, 250 lM dNTPs each, 0.1 U ⁄ ll
Taq polymerase (Roche). For PCR with CD68 primers,
2 ll of the transcription product were used at an anneal-
ing temperature of 57 C. Aliquots were taken after 20,
25 and 30 cycles. For 18s RNA detection, 1 ll of the
transcription product was used at an annealing tempera-
ture of 57 C and aliquots were taken after 17, 19 and 21
cycles. Primer sets: CD68, forward: GCA ACT CGA
GCA TCA TTC TTT CAC C, CD68 reverse: GAT GAG
AGG CAG CAA GAT GGA C; 18s RNA, forward: ACC
GAT TGG ATG GTT TAG TGA G, 18sRNA reverse:
CCT ACG GAA ACC TTG TTA CGA C.
RNA ligase mediated RACE-PCR. Total RNA from
purified human blood monocytes, monocyte-derived mac-
rophages, monocyte-derived dendritic cells and N1 der-
mal fibroblasts was isolated by the method of
Chomczynski and Sacchi [24]. Ten micrograms of total
RNA was used for cDNA synthesis with the First-
ChoiceTM RLM-RACE Kit (Ambion, Austin, TX, USA).
The following specific primers were used to amplify full-
length 5¢ cDNA fragments of human CD68: hCD68-
OUT (5¢-AAG ATC CTG AGC TGC CCT TGC-3¢),
hCD68-IN (5¢-CCA AGA CCC ACA CCA TCC AC-3¢).
PCR products derived from fibroblast or macrophage
RNA were cloned into pCR2.1-TOPO (TOPO Cloning
Kit, Invitrogen) according to the manufacturer’s instruc-
tions. Seven individual, plasmid-containing bacterial
Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
E. Gottfried et al.
colonies were selected via ampicillin resistance plus LacZ-
gene expression. Plasmids were isolated via the plasmid
mini ⁄ midi or maxi Kits (Qiagen), inserts were amplified
by PCR and directly sequenced (GeneArt, Regensburg,
Germany).
Results
CD68 immunoreactivity in haematopoietic and
non-haematopoietic cell types
The expression of CD68 is widely used as a macrophage
marker. We started to extend reports on the expression
of CD68 in myeloid and non-myeloid cell populations.
First, we determined CD68 expression by intracellular
staining for flow cytometric analysis in a variety of cell
types using different monoclonal antibodies (mAb)
against various CD68 antigenic sites. Monocyte-derived
macrophages were always used as a positive control. As
expected, macrophages showed a strong reactivity with
all CD68 mAbs except for PG-M1 (Fig. 1A).
Freshly isolated human peripheral blood lymphocytes
were stained positive with the anti-CD68 mAbs KP1
and EBM11, but the staining intensity was much lower.
No reactivity was seen with the other antibody clones in
human lymphocytes (Fig. 1A). CD68 expression was also
analysed in two lymphocytic tumour cell lines (laz388
and Jurkat) and immunoreactivity was found with the
mAb KP1, but not with the other mAb clones (Fig. 2A).
The same was true for tumour cell lines of various
origins. Indeed, reactivity with the mAb KP1 was
detected in all tumour cell lines tested but the staining
intensity was variable. Capan1, Panc1 (both derived from
pancreas carcinomas) and RT4 (derived from urothelial
bladder carcinoma) exhibited the strongest reactivity
(Figs. 1A and 2B).
Further on, we documented the immunoreactivity of
anti-CD68 antibodies in cell types that contribute to the
stroma of solid tumours or play a role in chronic inflam-
matory diseases, such as fibroblasts and endothelial cells.
Primary fibroblasts isolated from breast carcinoma
(PF28), normal breast tissue (PFN2), normal skin (N1) as
well as from synovial tissue of osteoarthritis patients
(OA112) were also analysed. The anti-CD68 antibodies
KP1, KiM7, and to a lesser degree, the mAbs EBM11
and 7B8a bound to fibroblasts. No reactivity was found
with the other mAb clones against CD68 (Figs. 1A–C
and 2C). All endothelial cell types analysed, such as
endothelial cells, isolated from human umbilical veins
(HUVEC) and the microvascular endothelial cell line
HMEC showed a strong reactivity with mAb KP1 and
weak staining with EBM11 (Figs. 1A and 2C). Again, no
reactivity was detected with the other mAbs. The inten-
sity of staining with KP1 observed in fibroblasts and
endothelial cells was comparable to monocytes and
 2008 The Authors
E. Gottfried et al. CD68 Expression in Non-Myeloid Cells 457
..................................................................................................................................................................

A Macrophages Lymphocytes Capan1 HUVEC N1 fibroblasts

350

350

350

350

350
Counts
KP1

0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

350

350

350

350
350
Counts
EBM11

0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

350

350

350

350

350
Counts
KiM6

0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

350

350

350

350

350
Counts
7B8a

0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

350

350

350

350

350
Counts PGM1
0

0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104

B Macrophages
104 104

CD14 PE
103 103
Iso PE

102 102
101 101
100 0 100
10 101 102 103 104 100 10–1 10–2 10–3 10–4
Iso FITC CD68 FITC
104 104
HLADR APC

103 103
Iso APC

102 102
10 1 101
10 0 100
100 10–1 10–2 10–3 10–4 100 101 102 103 104
Iso FITC CD68 FITC

N1 Fibroblasts
104 104 104
CD14 PE
CD14 PE

10 3
10 3 103
Iso PE

10 2
10 2
102
Figure 1 Examples of CD68 immunoreactiv- 10 1
10 1
101
ity. (A) Intracellular staining for flow cyto- 10 0
10 0
100
metric analysis of CD68 immunostaining. 100 101 102 103 104 100 10–1 10–2 10–3 10–4 100 10–1 10–2 10–3 10–4
Different mAb clones (KP-1, EBM11, KiM6, Iso FITC CD68 FITC CD90 FITC
PG-M1, 7B8a) were used in various cell 104
104 104
HLADR APC

HLADR APC

types: macrophages, lymphocytes, Capan1

Iso APC

103
(pancreatic carcinoma cells), HUVEC (human 103 103
102 2
umbilical vein endothelial cells) and N1 10 102
fibroblasts. Histograms show the staining 101 1
10 101
intensity (filled) relative to the respective 100 0
10 100
isotype control (line). The signal intensity in 100 101 102 103 104 100 10–1 10–2 10–3 10–4 100 10–1 10–2 10–3 10–4
macrophages differs for all antibody clones. Iso FITC CD68 FITC CD90 FITC
(B) Purity of the cultured fibroblast popula-
tion was confirmed by double-staining CD68 C Macrophages N1 Fibroblasts
0 10 20 30 40 50 60 70 80
0 10 20 30 40 50 60 70 80

antibody (KP1) with a fibroblast marker

(anti-CD90), anti-HLA-DR antibody or anti-
Counts
Counts

CD14 antibody, respectively, to exclude the

contamination with myeloid cells. As a con-
trol, macrophages were stained with the same
antibody panel. (C) Macrophages and N1
fibroblasts were stained against CD68 with 100 101 102 103 104 100 101 102 103 104
KiM7 antibody. KiM7 KiM7

 2008 The Authors

Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
458 CD68 Expression in Non-Myeloid Cells E. Gottfried et al.
..................................................................................................................................................................

A monocytes ⁄ macrophages while CD14 and HLA-DR were

1000 KP1 stained as markers for monocytes ⁄ macrophages.
Mean fluorescence intensity

EBM11
800
Ki-M6
600 Immunohistochemical analysis of CD68 expression

400 Antibody reactivity often depends on the fixation tech-

nique and the method of detection. Therefore, monocytes
200
and the tumour cell line HT29 were analysed after
0 immunohistochemical staining using the APAAP tech-
MAC MO LY laz388 Jurkat nique. Cells were prepared as cytospins and fixation was
B performed with glutaraldehyde, which was known to be
1000
sufficient for CD68-positive cell detection (Fig. 3). APA-
Mean fluorescence intensity

KP1 AP staining showed a strong reactivity of both mono-

800
Ki-M6
cytes and tumour cells. Staining was mainly localized
600
within the cytoplasm, but subtle staining of the cell
400 membrane was also observed. In this immunohistochemi-
cal application, the EBM11 antibody resulted in staining
200
signals almost as intense as KP1. In addition, single cells
0 were also positive with the mAb Ki-M6.
A549 HT29 RJ494 BT474 T47D J82 RT4 Panc1 Capan1 KP1, which detects a fixation-resistant epitope [10], is
C 1000 most commonly used for the diagnosis of neoplasms in
paraffin-embedded material. Therefore, we applied KP1
Mean fluorescence intensity

KP1
800
EBM11 to stain formalin-fixed ⁄ paraffin-embedded fibroblasts and
Ki-M6 tumour cell pellets. N1 fibroblasts were strongly stained
600
as well as monocytes that were used as a positive control
400 (Fig. 4A). Among the tested tumour cell lines, individual
Capan1 (Fig. 4A) and RT4 (data not shown) cells were
200 slightly positive. All other tested tumour cell lines were
0
negative for KP1 reactivity in this experimental setting.
HMEC HUVEC PF N2 PF 28 OA112 N1 To confirm positive staining of tumour cells also in
routine paraffin-embedded material, we stained pancreatic
Figure 2 CD68 expression in different cell types. Haematopoietic and carcinoma biopsies using the KP1 antibody. Figure 4B
non-haematopoietic cells were analysed by flow cytometry for the intra- shows immunohistochemical staining of tumour cell
cellular expression of CD68 with different mAbs. The mean fluorescence
intensity is given as mean ± SD of at least three independent experi-
islets surrounded by stromal tissue. A strong staining of
ments, with the mean fluorescence intensity of the isotype control sub- tumour cells, as well as macrophages, in the stroma was
tracted. PFN2, PF28 and OA112 fibroblasts were only examined twice, detected.
as higher passage numbers change the phenotype. The CD68 expression To conclude, KP1 should only be used with great cau-
of (A) monocytes (MO), monocyte-derived macrophages (MAC) and tion for the distinction between monocytes ⁄ macrophages
lymphocytes was compared with (B) lymphocytic cell lines and various
tumour cell lines. (C) CD68 expression in endothelial cells (HMEC,
and tumour cells in formalin-fixed, paraffin-embedded
HUVEC) and different types of primary fibroblasts derived from breast tumour material.
tissue (PFN2, PF28), synovial tissue of osteoarthritis patients (OA112)
and skin (N1).
Western blot analysis of CD68 expression

To determine that the reactivity of KP1 with tumour

monocyte-derived macrophages. The data suggest that
the KP1 antibody reacts with a distinct epitope or a dis-
tinct antigen that is shared by macrophages and other
cells whereas other CD68 antibodies bind to a macro-
phage-specific epitope.
To confirm high purity of the primary fibroblasts and
avoid misinterpretation due to contaminating cell popu-
lations, double staining of N1 fibroblasts was performed
using anti-CD90, anti-CD14 and anti-HLA-DR antibod-
ies combined with the KP1 mAb (Fig. 1B). CD90 is
known to be highly expressed on fibroblasts but not on
cells, fibroblasts and endothelial cells is due to CD68
expression, we performed a Western blot analysis.
Fig. 5 shows that KP1 detects a band of approximately
110 kDa in both fibroblasts and macrophages. However,
the signal was much stronger in macrophages as com-
pared with fibroblasts. This is in line with mRNA
expression data (see Fig. 6) and flow cytometry data
using KiM7 as CD68 antibody (Fig. 1C). However, it
is contradictory to flow cytometry data where mono-
cytes ⁄ macrophages and fibroblasts showed a comparable
reactivity with KP1. One explanation could be that a

 2008 The Authors

Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
E. Gottfried et al. CD68 Expression in Non-Myeloid Cells 459
..................................................................................................................................................................

Monocytes (Isotype) Monocytes (KP1)

HT29 (Isotype) HT29 (KP1)

HT29 (EBM11) HT29 (Ki-M6)

Figure 3 Immunohistochemical analysis of

CD68 expression after fixation with glutaral-
dehyde. Cell suspensions of monocytes and
HT29 cell line were prepared as cytospins,
air dried and fixed with glutaraldehyde. AP-
AAP staining was performed with KP1,
EBM11, Ki-M6 or an isotype control as pri-
mary antibody.

cross-reactive epitope is dependent on native confirma-
tion and destroyed by SDS used for Western blot
analysis.
CD68 is a highly glycosylated protein and it is still
unclear whether KP1 detects a glycosylated protein rather
than a protein epitope. Therefore, we incubated macro-
phages and fibroblasts with N-glycosidase F, an enzyme
that cleaves N-glycans. As expected, this treatment
resulted in a reduction in the molecular mass of about
20 kDa, which is in line with the published data.
(Fig. 5).
However, mRNA expression could also be detected in
lymphocytes, fibroblasts, endothelial cells and tumour
cells of different origins, although to a varying extent
(Fig. 6).
The protein and mRNA data show that macrophages
express CD68 at higher levels than other positive cells,
but they also indicate that CD68 is not only restricted to
macrophages.
Transcriptional start sites of CD68 in monocytes,
macrophages, dendritic cells and fibroblasts

The proximal promoter of CD68 genes in mouse and

Expression of CD68 on mRNA level
In order to confirm the presence of CD68 mRNA in cells
that showed CD68 immunoreactivity, we performed
quantitative reverse transcription-PCR. As expected,
CD68 mRNA was highly expressed in macrophages and
monocytes as well as in the myeloid cell lines THP1.
man have been characterized and described as being mac-
rophage specific [25–27]. To clarify whether myeloid and
non-myeloid cell types use alternative or similar regula-
tory regions to control CD68 expression, we determined
the transcriptional start sites and the full-length sequence
of CD68 transcripts in human fibroblasts and compared

 2008 The Authors

Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
460 CD68 Expression in Non-Myeloid Cells E. Gottfried et al.
..................................................................................................................................................................

A HT29 Capan1

N1 fibroblasts Monocytes

Figure 4 Immunohistochemical analysis of CD68

expression in paraffin-embedded material. (A) Cell
suspensions were pelleted, fixed in 4% formalin
and embedded in paraffin. Five lm-sections were
stained with mAb KP1 or an isotype control.
(B) Immunohistochemical staining of a pancreatic
carcinoma site with mAb KP1 using routine
processing with a peroxidase-based labelling tech-
nique and diaminobenzidine (DAB) as chromogen
(brown). Upper panel: overview (original magnifi-
cation 50·). Lower panel: detailed view; arrows
indicate CD68 positive tumour cells and tumour
cell islets, arrowheads indicate CD68 positive
macrophages in the tumour stroma (original
magnification 200 · ).

 2008 The Authors

Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
E. Gottfried et al. CD68 Expression in Non-Myeloid Cells 461
..................................................................................................................................................................

Figure 5 Western blot analysis of CD68 expression. Pellets of mono-

cytes-derived macrophages (lanes 1, 2) or N1 fibroblasts (lanes 3,4) were
incubated with (lanes 2,4) or without (lanes 1,3) N-glycosidase F to
remove N-glycans. Western blot analysis was performed according to
process described in Material and Methods section. b-Actin staining is
shown as a loading control.

them to those obtained from human monocyte-derived

cell types. PCR fragments of full length 5¢-ends were
obtained by RNA ligase-mediated rapid amplification of Figure 7 CD68 transcription start sites and proximal promoter region.
cDNA ends (RLM-RACE), a method which allows the (A) CD68-specific 5¢-RLM-RACE products from N1 fibroblasts (N1),
selective amplification of capped, full-length transcripts. monocytes (MO), macrophages (MAK) and monocyte-derived dendritic
cells (DC) were separated by agarose gel electrophoresis along with
Each cell type produced a specific pattern of PCR frag-
molecular-weight markers and stained with ethidium bromide. (B)
ments, indicating differences in transcript initiation in Structure of the CD68 promoter. Start codon, intron I border and puta-
myeloid versus non-myeloid cell types (Fig. 7A): fibro- tive transcription factor binding sites are indicated. Transcription start
blasts mainly produced a longer 5¢-PCR-fragment, sites are marked by arrows below (for fibroblasts) and above (for macro-
whereas myeloid cell types produced equal amounts of phages) the sequence.
long and two short fragments (monocytes) or preferen-
tially made a short fragment (macrophages, dendritic
Discussion
cells). To determine the exact sequence of individual
5¢-ends, seven individual PCR-fragments from fibroblasts CD68 is widely used as a macrophage marker, and the
and macrophages were cloned and sequenced. The posi- CD68 promoter has been proposed for the generation of
tion and relative abundance of transcription start sites macrophage-specific transgenes [26]. In this paper, we
indicate that all fragments initiate within the published have confirmed and extended reports of CD68 expression
promoter region of the human CD68 gene (Fig. 7B). In in non-myeloid cell populations such as carcinomas, mel-
contrast to myeloid cell types, human fibroblasts prefer- anomas and other cell entities, as shown in tissue sections
entially initiate the transcription of CD68 from an and tumour cell lines [17, 28–35]. In our hands, espe-
upstream start site, indicating alternative promoter cially cells of the tumour stroma, i.e. fibroblasts and
architectures in myeloid versus non-myeloid cell types. endothelial cells, showed a strong staining with CD68
Relative mRNA expression (CD68/18s)

1.0000

0.1000
Figure 6 Expression of CD68 mRNA analy-
sed by qRT-PCR analysis. Different cell
types, such as monocytes, macrophages, 0.0100
THP1 (monocytic leucaemia cell line), lym-
phocytes, N1-fibroblasts, endothelial cells
0.0010
(HUVEC and HMEC) and various tumour
cell lines (HT-29, J82, MelIm, Capan1) were
used for analysis of CD68 mRNA expression 0.0001
by qRT-PCR. CD68 expression is given
1

es

es

EC

29

Im

1
P-

ge

st

J8

an
VE
yt

yt

T-

el
M
la
TH

relative to 18s RNA. The mean of three

ha

ap
oc

oc

M
U

H
H
ob

H
op

C
on

ph

br

independent runs of one representative exper-

r
m

m
ac

fi
1-
ly
m

iment is shown.
N

 2008 The Authors

Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
462 CD68 Expression in Non-Myeloid Cells
..................................................................................................................................................................
mAb KP1 which was comparable to the intensity found
in macrophages. The strong reactivity was independent of
the source of the primary fibroblasts. CD68 staining
intensity was dependent on the mAb clone, the fixation
technique and the method of detection. The antibody
KP1 showed the strongest reactivity independent of the
cell type examined.
In contrast to the strong immunoreactivity of CD68,
CD68 mRNA expression in non-myeloid cells was
detected at a variable but lower level as compared with
monocytes ⁄ macrophages. Specific expression, however,
was clearly observed in fibroblasts, endothelial cells and
in tumour cells. Cloning of full length 5¢-sequences and
determination of transcription start sites also confirmed
the presence of CD68 mRNA in monocytes, macrophages
and fibroblasts. Our data also suggest that fibroblasts
initiate CD68 transcription in the same promoter region
as macrophages; however, with a different start site pref-
erence. This may indicate cell-type specific cis-elements
and promoter architectures. A more detailed analysis of
the promoter in different cell populations, e.g. by trans-
fection of reporter constructs will be needed to define the
specific sequence elements that are required for CD68
expression in different cell types. Earlier studies already
demonstrated that the human CD68 promoter directs
reporter gene expression in both macrophage and non-
macrophage cell lines in transient transfections. Enhanced
reporter activity in macrophages actually depends on
regulatory sequences in the first intron of CD68 [25].
Macrophage-specific transgene expression in vivo was also
reported when using the CD68 promoter in combination
with the first intron, although transgene expression was
only analysed in isolated macrophages and not in other
cell populations [26]. Our results do not necessarily
contradict the observed macrophage specificity of CD68
promoter ⁄ intron constructs since distinct upstream or
downstream elements may be essential for non-macro-
phage expression of CD68 in vivo. A very recent publica-
tion showed that cell type specific expression of CD68 in
haematopoietic cells is regulated by local gene architec-
ture and associated with changes in Pol-II phosphoryla-
tion and mRNA splicing [36].
In line with the RNA expression data, Western blot
analysis in our hands revealed that the protein was
detected at higher levels in macrophages, but is not mac-
rophage restricted. CD68 protein was clearly detected in
fibroblasts. In this respect, CD68 is probably not differ-
ent from other lysosome-associated proteins or lysosomal
hydrolase, all of which are highly expressed in profes-
sional phagocytes. We acknowledge that there is a clear
need to avoid equating immunoreactivity of a single
mAb with the level of protein in a cell. Thus, we cannot
eliminate the possibility that KP1 reacts with a distinct
antigen that is sensitive towards SDS treatment and
therefore not detected on a Western blot. However, it is
Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
E. Gottfried et al.
equally likely that post-translational modifications alter
monoclonal antibody reactivity in macrophages as com-
pared with other cells. Regardless of the observed differ-
ences between mRNA or protein expression levels and
immunoreactivity, our findings clearly show that mAb
KP1 cannot be used reliably to identify macrophages in
tissue sections.
In conclusion, we suggest that CD68 is not a selective
macrophage marker, but is rather a lysosomal protein
that is enriched in macrophages.
Acknowledgments
We thank Prof. Dr David Hume, University of Queens-
land, for thorough discussion of the manuscript and
many helpful suggestions. We thank Prof. Dr F. Hof-
städter, University of Regensburg, for critically reading
the manuscript. This work was supported by DFG grant
Kr1418 ⁄ 6-1.
References
1 Saito N, Pulford KA, Breton-Gorius J, Masse JM, Mason DY, Cra-
mer EM. Ultrastructural localization of the CD68 macrophage-asso-
ciated antigen in human blood neutrophils and monocytes. Am J
Pathol 1991;139:1053–9.
2 Holness CL, da Silva RP, Fawcett J, Gordon S, Simmons DL. Mac-
rosialin, a mouse macrophage-restricted glycoprotein, is a member
of the lamp ⁄ lgp family. J Biol Chem 1993;268:9661–6.
3 Holness CL, Simmons DL. Molecular cloning of CD68, a human
macrophage marker related to lysosomal glycoproteins. Blood
1993;81:1607–13.
4 Kurushima H, Ramprasad M, Kondratenko N, Foster DM, Quehen-
berger O, Steinberg D. Surface expression and rapid internalization
of macrosialin (mouse CD68) on elicited mouse peritoneal macro-
phages. J Leukoc Biol 2000;67:104–8.
5 Ramprasad MP, Terpstra V, Kondratenko N, Quehenberger O,
Steinberg D. Cell surface expression of mouse macrosialin and
human CD68 and their role as macrophage receptors for oxidized
low density lipoprotein. Proc Natl Acad Sci U S A 1996;93:14833–
8.
6 Parwaresch MR, Radzun HJ, Kreipe H, Hansmann ML, Barth J.
Monocyte ⁄ macrophage-reactive monoclonal antibody Ki-M6 recog-
nizes an intracytoplasmic antigen. Am J Pathol 1986;125:141–51.
7 Kreipe H, Radzun HJ, Parwaresch MR, Haislip A, Hansmann ML.
Ki-M7 monoclonal antibody specific for myelomonocytic cell line-
age and macrophages in human. J Histochem Cytochem
1987;35:1117–26.
8 Davey FR, Cordell JL, Erber WN, Pulford KA, Gatter KC, Mason
DY. Monoclonal antibody (Y1 ⁄ 82A) with specificity towards
peripheral blood monocytes and tissue macrophages. J Clin Pathol
1988;41:753–8.
9 Kelly PM, Bliss E, Morton JA, Burns J, McGee JO. Monoclonal
antibody EBM ⁄ 11: high cellular specificity for human macrophages.
J Clin Pathol 1988;41:510–5.
10 Pulford KA, Rigney EM, Micklem KJ et al. KP1: a new monoclo-
nal antibody that detects a monocyte ⁄ macrophage associated antigen
in routinely processed tissue sections. J Clin Pathol 1989;42:414–
21.
11 Radzun HJ, Hansmann ML, Heidebrecht HJ et al. Detection of a
monocyte ⁄ macrophage differentiation antigen in routinely processed
 2008 The Authors
E. Gottfried et al.
..................................................................................................................................................................
paraffin-embedded tissues by monoclonal antibody Ki-M1P. Lab
Invest 1991;65:306–15.
12 Falini B, Flenghi L, Pileri S et al. PG-M1: a new monoclonal anti-
body directed against a fixative-resistant epitope on the macro-
phage-restricted form of the CD68 molecule. Am J Pathol
1993;142:1359–72.
13 Micklem K, Rigney E, Cordell J et al. A human macrophage-associ-
ated antigen (CD68) detected by six different monoclonal antibod-
ies. Br J Haematol 1989;73:6–11.
14 Kempf W, Adams V, Wey N et al. CD68 + cells of mono-
cyte ⁄ macrophage lineage in the environment of AIDS-associated
and classic-sporadic Kaposi sarcoma are singly or doubly infected
with human herpesviruses 7 and 6B. Proc Natl Acad Sci U S A
1997;94:7600–5.
15 Hakkinen T, Karkola K, Yla-Herttuala S. Macrophages, smooth
muscle cells, endothelial cells, and T-cells express CD40 and
CD40L in fatty streaks and more advanced human atherosclerotic
lesions. Colocalization with epitopes of oxidized low-density lipo-
protein, scavenger receptor, and CD16 (Fc gammaRIII). Virchows
Arch 2000;437:396–405.
16 Soilleux EJ, Morris LS, Trowsdale J, Coleman N, Boyle JJ. Human
atherosclerotic plaques express DC-SIGN, a novel protein found on
dendritic cells and macrophages. J Pathol 2002;198:511–6.
17 Kunisch E, Fuhrmann R, Roth A et al. Macrophage specificity of
three anti-CD68 monoclonal antibodies (KP1, EBM11, and PGM1)
widely used for immunohistochemistry and flow cytometry. Ann
Rheum Dis 2004;63:774–84.
18 Andreesen R, Brugger W, Scheibenbogen C et al. Surface phenotype
analysis of human monocyte to macrophage maturation. J Leukoc
Biol 1990;47:490–7.
19 Stevenson HC, Miller P, Akiyama Y et al. A system for obtaining
large numbers of cryopreserved human monocytes purified by leuka-
pheresis and counter-current centrifugation elutriation (CCE).
J Immunol Methods 1983;62:353–63.
20 Andreesen R, Picht J, Lohr GW. Primary cultures of human blood-
born macrophages grown on hydrophobic teflon membranes.
J Immunol Methods 1983;56:295–304.
21 Kunz-Schughart LA, Heyder P, Schroeder J, Knuechel R. A heterol-
ogous 3-D coculture model of breast tumor cells and fibroblasts to
study tumor-associated fibroblast differentiation. Exp Cell Res
2001;266:74–86.
22 Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human
endothelial cells derived from umbilical veins. Identification by mor-
phologic and immunologic criteria. J Clin Invest 1973;52:2745–56.
23 Cordell JL, Falini B, Erber WN et al. Immunoenzymatic labeling of
monoclonal antibodies using immune complexes of alkaline
 2008 The Authors
Journal compilation  2008 Blackwell Publishing Ltd. Scandinavian Journal of Immunology 67, 453–463
CD68 Expression in Non-Myeloid Cells 463
phosphatase and monoclonal anti-alkaline phosphatase (APAAP
complexes). J Histochem Cytochem 1984;32:219–29.
24 Chomczynski P, Sacchi N. Single-step method of RNA isolation by
acid guanidinium thiocyanate-phenol-chloroform extraction. Anal
Biochem 1987;162:156–9.
25 Greaves DR, Quinn CM, Seldin MF, Gordon S. Functional compari-
son of the murine macrosialin and human CD68 promoters in mac-
rophage and nonmacrophage cell lines. Genomics 1998;54:165–8.
26 Gough PJ, Gordon S, Greaves DR. The use of human CD68 tran-
scriptional regulatory sequences to direct high-level expression of
class A scavenger receptor in macrophages in vitro and in vivo. Immu-
nology 2001;103:351–61.
27 O’Reilly D, Quinn CM, El Shanawany T, Gordon S, Greaves DR.
Multiple Ets factors and interferon regulatory factor-4 modulate
CD68 expression in a cell type-specific manner. J Biol Chem
2003:278: 21909–19.
28 Pulford KA, Sipos A, Cordell JL, Stross WP, Mason DY. Distribu-
tion of the CD68 macrophage ⁄ myeloid associated antigen. Int
Immunol 1990;2:973–80.
29 Kaiser U, Hansmann ML, Papadopoulos I, Radzun HJ. Mono-
cyte ⁄ macrophage-directed antibodies Ki-M3 and Ki-M7 detect
renal-cell carcinomas. Int J Cancer Suppl 1988;3:45–9.
30 Doussis IA, Gatter KC, Mason DY. CD68 reactivity of non-macro-
phage derived tumours in cytological specimens. J Clin Pathol
1993;46:334–6.
31 Gloghini A, Rizzo A, Zanette I et al. KP1 ⁄ CD68 expression in
malignant neoplasms including lymphomas, sarcomas, and carcino-
mas. Am J Clin Pathol 1995;103:425–31.
32 Facchetti F, Bertalot G, Grigolato PG. KP1 (CD 68) staining of
malignant melanomas. Histopathology 1991;19:141–5.
33 Tos AP, Doglioni C, Laurino L, Fletcher CD. KP1 (CD68)
expression in benign neural tumours. Further evidence of its low
specificity as a histiocytic ⁄ myeloid marker. Histopathology 1993;
23:185–7.
34 Cassidy M, Loftus B, Whelan A et al. KP-1: not a specific marker.
Staining of 137 sarcomas, 48 lymphomas, 28 carcinomas, 7 malig-
nant melanomas and 8 cystosarcoma phyllodes. Virchows Arch
1994;424:635–40.
35 Hasegawa T, Hirose T, Seki K, Sano T, Hizawa K. Transforming
growth factor alpha and CD68 immunoreactivity in giant cell
tumours of bone: a study on the nature of stromal and giant cells,
and their interrelations. J Pathol 1993;170:305–10.
36 O’Reilly D, Greaves DR. Cell-type-specific expression of the human
CD68 gene is associated with changes in Pol II phosphorylation and
short-range intrachromosomal gene looping. Genomics 2007;90:407–
15.