RESEARCH ARTICLE | SEPTEMBER 01 2003

Downloaded from http://journals.aai.org/jimmunol/article-pdf/171/5/2262/1168994/2262.pdf by guest on 29 April 2024

TNF Skews Monocyte Di erentiation from Macrophages to Dendritic Cells 1

Pascale Chomarat; ... et. al
J Immunol (2003) 171 (5): 2262–2269.
https://doi.org/10.4049/jimmunol.171.5.2262

Related Content
Understanding the Interplay of the Human Immune System with Cancer in Humanized Mice (49.17)
J Immunol (April,2007)

Adjuvant skewing of host immune response to vaccination and subsequent infection with in uenza virus
J Immunol (May,2023)

Intestinal Regulation of C. albicans by PP Langerin+ DCs

J Immunol (May,2017)
 The Journal of Immunology

TNF Skews Monocyte Differentiation from Macrophages to

Dendritic Cells1

Pascale Chomarat,2,3 Carole Dantin,2 Lynda Bennett, Jacques Banchereau, and

A. Karolina Palucka4
Monocytes represent a large pool of circulating precursors of APCs, both macrophages and dendritic cells (DCs). It is thus
important to identify the mechanisms by which microenvironment regulates monocyte differentiation. We have previously shown
that, upon contact with resting stromal cells such as fibroblasts, monocytes differentiate into macrophages in an IL-6/M-CSF-
dependent fashion. Yet, in the inflamed tissue, monocytes need to yield DCs for the adaptive immunity to be induced. Inasmuch
as TNF and IL-1 are present at the site of inflammation, we tested their capacity to modulate monocyte differentiation into either
macrophages or DCs. TNF, but not IL-1, induce monocytes to become DCs despite the presence of fibroblasts. TNF-induced DCs

Downloaded from http://journals.aai.org/jimmunol/article-pdf/171/5/2262/1168994/2262.pdf by guest on 29 April 2024

contain Langerin-positive cells and are able to induce allogenic T cell proliferation. Then, TNF was found to decrease the
expression and internalization of the M-CSF receptor, thus overriding the IL-6/M-CSF pathway. Thus, TNF facilitates the in-
duction of adaptive immunity by promoting DC differentiation not only from CD34ⴙ progenitors but also from CD14ⴙ blood
precursors. The Journal of Immunology, 2003, 171: 2262–2269.

M onocytes represent a large pool of circulating precur- as a major acute inflammation cytokine (9, 10), two other cyto-
sors that can differentiate into macrophages (M␾)5 or kines play a very important role in the inflammatory process, i.e.,
dendritic cells (DCs) (1– 4). M␾, scavengers of the IL-1 (11) and TNF (12, 13). Both cytokines are implicated in acute
immune system, are essential to eliminate the burden of pathogens. and chronic inflammation, likewise pathogen-mediated, for exam-
We have shown previously that in a resting stromal microenviron- ple sepsis, and autoimmune, for example rheumatoid arthritis (9,
ment, resembling a noninflammatory situation, monocytes differ- 14). They share several functional properties, as both are able to
entiate into M␾ in a process regulated by fibroblast-derived IL-6 induce IL-6, up-regulate the expression of adhesion molecules, and
(5). IL-6 up-regulates expression of the M-CSF receptor and fa- increase the synthesis of acute phase proteins (12, 13, 15). Yet, we
cilitates M-CSF internalization, resulting in predominant M⌽ phe- show here that TNF, but not IL-1, is able to redirect the differen-
notype. However, M␾ are not sufficient for the development of tiation of monocytes toward DCs by offsetting the IL-6/M-CSF
pathogen-specific protective immunity. This can only be initiated pathway. Therefore, TNF is essential for the induction of adaptive
by DCs, the inducers of the immune response. The inflammatory immunity.
reaction occurring upon pathogen invasion, “danger” signal, leads
to DC activation, migration, and maturation, culminating in the
Materials and Methods
induction of immunity and pathogen elimination (3, 4, 6).
Cell cultures
Tissue injury can cause dramatic changes throughout the micro-
environment (7), including release of inflammatory mediators (cy- Mononuclear cells were isolated from peripheral blood of healthy volun-
tokines, chemokines) (8) and cellular infiltrates of activated leu- teers by Ficoll-Paque density gradient centrifugation. Monocytes were pu-
rified by depletion using mAbs and Dyna beads, routinely resulting in
kocytes, neutrophils, and mast cells. Therefore, by determining ⬎90% purity. Monocytes were seeded at 1 ⫻ 106 cells/well in six-well
precursor differentiation, the cytokines in the local environment plates in 3 ml of RPMI 1640 medium supplemented with 10% FBS (Life
shape protective immune responses. It is thus important to under- Technologies, Rockville, MD), 2 mM L-glutamine, 200 IU/ml penicillin,
stand the regulation of monocyte differentiation. Whereas IL-6 acts and 200 ␮g/ml streptomycin (Sigma-Aldrich, St. Louis, MO). Normal skin
fibroblasts (CD1074SK; American Type Culture Collection, Manassas,
VA) were added to the six-well plates at a density of 2.5 ⫻ 105 cells/well,
unless otherwise indicated. Cultures were fed with cytokines every 3 days.
Baylor Institute for Immunology Research, Dallas, TX 75204 Recombinant human cytokines (R&D Systems, Minneapolis, MN) were as
Received for publication February 7, 2003. Accepted for publication June 25, 2003. follows: IL-1␣ (50 ng/ml), IL-1␤ (25 ng/ml), IL-6 (200 ng/ml) and TNF
The costs of publication of this article were defrayed in part by the payment of page (20 –100 ng/ml unless otherwise indicated, batch-related variability), IL-4
charges. This article must therefore be hereby marked advertisement in accordance (25 ng/ml), and GM-CSF (100 ng/ml, Leukine; Immunex, Seattle, WA).
with 18 U.S.C. Section 1734 solely to indicate this fact. Neutralizing TNF, IL-6, and IL-6R mAbs were purchased from R&D
1 Systems.
This work was supported by National Institutes of Health (Grant RO1 CA78846 to
J.B. and Grant RO1 CA89440 to A.K.P.). J.B. is the recipient of the Max & Gayle
Clampitt Chair for Immunology Research. Allogeneic MLR assay
2
P.C. and C.D. are equal contributors to this study.
After 5 days of culture, the CD1⫹ cells generated from unstimulated or
3
Current address: Institut de Recherche, SERVIER, 125 chemin de Ronde, 78290 TNF-treated monocyte/fibroblast cocultures were FACS sorted (FACS-
Croissy sur Seine, France. E-mail address: pascale.chomarat@fr.netgrs.com Vantage; BD Biosciences, Mountain View, CA) and used as stimulators for
4
Address correspondence and reprint requests to Dr. A. Karolina Palucka (E-mail purified naive (CD45RA⫹CD45RO⫺) allogenic CD4⫹ T cells. Graded
address: karolinp@baylorhealth.edu) or Dr. Jacques Banchereau (E-mail address: doses of stimulator cells were seeded with 1 ⫻ 105 naive CD4⫹ T cells in
jacquesb@baylorhealth.edu), Baylor Institute for Immunology Research, 3434 Live round-bottom microtest tissue culture plates in complete RPMI 1640 with
Oak, Dallas, TX 75204 TX. 10% human AB serum (Gemini Bio-Products, Woodland, CA). After 4
5
Abbreviations used in this paper: M␾, macrophage; DC, dendritic cell; LC, Lang- days of incubation, cells were pulsed overnight with 1 ␮Ci of [3H]thymi-
erhans cell; RA, rheumatoid arthritis. dine to determine T cell proliferation.

Copyright © 2003 by The American Association of Immunologists, Inc. 0022-1767/03/$02.00

The Journal of Immunology 2263

Cytokine detection 1/25; BioSource International). For the kinetic studies, cells were double
stained with PE-HLA-DR (BD Biosciences) to allow the analysis of M-
Cytokines were detected in cell-free culture supernatants using commercial CSF receptors on myeloid cells.
quantitative sandwich immunoassay kits: IL-6 (BD PharMingen, San Di-
ego, CA) and M-CSF (R&D Systems).

FACS analysis Real-time RT-PCR

At indicated time points, cells were processed for double staining (30 min Monocytes were incubated with GM-CSF and IL-4 for 5 days with or
at 4oC) using FITC-CD1a (recently reclassified as a CD1b/c mAb; Bio- without IL-6 (200 ng/ml), TNF (100 ng/ml), or a combination. Cells were
Source International, Camarillo, CA) and PE-CD14 or PE-CD83 (BD Bio- harvested at day 5 and RNA was extracted using an RNeasy kit from
sciences) mAbs. Surface M-CSF receptors were detected using a rat anti- Qiagen (Valencia, CA). Real-time PCR was conducted using an Applied
M-CSFR mAb 3-4A4 (Santa Cruz Biotechnology, Santa Cruz, CA) Biosystems Prism 7700 Sequence Detector (Applied Biosystems, Foster
revealed in a second step by FITC-labeled goat F(ab⬘)2 anti-rat Ig (dilution City, CA). Probes and primers for the IL-6R and M-CSFR genes were

Downloaded from http://journals.aai.org/jimmunol/article-pdf/171/5/2262/1168994/2262.pdf by guest on 29 April 2024

FIGURE 1. TNF skews monocyte differentiation toward DCs in fibroblast/monocyte cocultures. Monocytes were cultured for 5 days in GM-CSF and
IL-4 with or without stromal cells (ratio 4:1). a, IL-1␣ (50 ng/ml), IL-1␤ (25 ng/ml), and TNF (30 ng/ml) were added at the onset of the cultures and
differentiation was monitored using expression of CD1b/c and CD14. The percentages of each myeloid cell subset in the total cell population are indicated.
Double-negative cells are fibroblasts. b, Dose-response curve. TNF was added to cocultures at increasing concentrations. Percentages (vertical axis) of
CD14⫹CD1b/c⫺ (- - - -) and CD14⫺CD1b/c⫹ (—) cells gated on HLA-DR expression at day 5 of culture. DC differentiation was observed already at the
dose of 10 ng/ml. However, because of batch-to-batch variation TNF was used at 100 ng/ml in the majority of experiments unless otherwise indicated. c,
Skewing of monocyte differentiation toward DCs is dependent on added TNF. Monocytes are cultured without (upper left plot) or with (upper right plot)
TNF. TNF-neutralizing mAbs are added at day 0 (lower plots). Monocyte differentiation is determined by flow cytometry at day 5. d, Phenotypic
characterization. Day 5 HLA-DR⫹ cells are gated on forward scatter (FSC)/side scatter plots to distinguish monocytes from fibroblasts (left panel). Surface
expression of DC-SIGN (middle panel) and Langerin (CD207, right panel) by HLA-DR⫹ cells.
2264 TNF SKEWS MONOCYTES TOWARD DCs

FIGURE 2. TNF, but not IL-1,

skews monocyte differentiation toward
DCs in fibroblast/monocyte cocultures.
Giemsa staining of cells retrieved from
cultures supplemented with GM-CSF
and IL-4 supplemented or not with TNF
or IL-1. Note an abundant highly vacu-
olated lightly stained cytoplasm and a
centrally positioned nucleus in cells cul-
tured without TNF addition (left and
middle panels). Cells from cultures with
TNF (right panel) display a homoge-
nous smooth intensely stained cyto-
plasm, cytoplasmic protrusions, and a
laterally positioned nucleus, consistent
with DC morphology.

designed using the computer software Primer Express (Applied Biosys- Confocal analysis
tems). Duplicate target and endogenous control reactions were set up in

Downloaded from http://journals.aai.org/jimmunol/article-pdf/171/5/2262/1168994/2262.pdf by guest on 29 April 2024

separate tubes. Reactions were prepared using the TaqMan One Step RT- Day 5 DCs were incubated for 24 h in GM-CSF and IL-4 with or without
PCR Master Mix and the human 18S RNA Pre-Developed Assay Reagent 200 ng/ml IL-6, 100 ng/ml TNF, or a combination. Cells were then col-
kit (Applied Biosystems) according to the manufacturer’s instructions. lected and incubated for 30 min at 37oC in medium alone or with M-CSF
Primers and probes were used at final concentrations of 900 and 50 nM, (50 ng/ml). After ice-cold washes, cells were stained for surface M-CSFR
respectively. The Sequence Detector was programmed for an initial step of expression and analyzed by flow cytometry. For immunofluorescence and
30 min at 48°C and 10 min at 95°C, followed by 40 thermal cycles of 15 s confocal microscopy, DCs were adhered to polylysine-coated glass slides
at 95°C and 1 min at 60°C. Fold expression was calculated using both the for 4 h at 4oC. Cells were fixed in 4% paraformaldehyde and permeabilized
Relative Standard Curve and the Comparative CT methods, where 18S in 0.05% saponin in PBS (pH 7.4). M-CSFR was detected using a non-
ribosomal RNA was used as the endogenous control for normalization blocking rat mAb 3-4A4 revealed by FITC-conjugated goat F(ab⬘)2 anti-rat
(Applied Biosystems TaqMan User Bulletin 2). Ig (BioSource International). Slides were mounted using a fluoromount

FIGURE 3. TNF skews monocyte differentiation toward partially mature DCs. a, MLR: day 5 DCs and monocyte/fibroblast cocultures treated or not with
TNF (100 ng/ml) were sorted using CD1b/c marker and incubated at graded doses with allogenic naive (CD45RA⫹CD45RO⫺) CD4⫹ T cells. Thymidine
incorporation is expressed as mean ⫾ SD of triplicates. Results are representative of three separate experiments. b, Monocytes were cultured for 5 days
with or without fibroblasts in GM-CSF and IL-4 in the presence or not of TNF. At day 5, the cells were stained with FITC-CD1b/c and PE-CD14 and then
sorted. CD14⫺CD1b/c⫹ cells were then restained with PE-CD83 and analyzed by flow cytometry. Each plot shows CD83 expression in cultures without
adding TNF (dotted line) and TNF-stimulated cultures (solid line). c, Flow cytometry plots of HLA-DR (ordinate) and CD83 (vertical axis) expression in
DCs generated by culturing monocytes with GM-CSF/IL-4 and TNF (day 5, left plot) and further exposed for 48 h to LPS 100 ng/ml (day 7, right plot).
The Journal of Immunology 2265

TNF-mediated skewing of monocyte differentiation toward DCs

could be observed already at a concentration of 1 ng/ml TNF and
reached a peak at 10 ng/ml (Fig. 1b). The observed effect on mono-
cyte differentiation was TNF dependent rather than due to contam-
ination such as LPS, as it could be prevented by the addition of
neutralizing TNF Ab (Fig. 1c). Monocyte/fibroblast cocultures in
the presence of TNF yield DCs that express surface CD1 (Fig. 1a)
and DC-SIGN (43 ⫾ 9% of HLA-DR⫹ DC-SIGN⫹ cells, n ⫽ 9)
(Fig. 1d). However, a discrete fraction of cells displays surface
staining with Ab recognizing Langerin (9 ⫾ 4% of HLA-DR⫹
CD207⫹ cells, n ⫽ 9; Fig. 1d).
Contrary to TNF, adding either IL-1-␣ or IL-1␤ to monocyte/
fibroblast cultures did not promote DC differentiation since hardly
any CD1⫹ cells could be seen (Fig. 1a). Rather, M␾ differentiation
prevailed as the cells retained high CD14 expression and displayed
M␾ morphology (Fig. 2). Indeed, the analysis of Giemsa-stained
cells sorted from cultures supplemented with GM-CSF and IL-4
(with or without IL-1) revealed an abundant highly vacuolated,

Downloaded from http://journals.aai.org/jimmunol/article-pdf/171/5/2262/1168994/2262.pdf by guest on 29 April 2024

FIGURE 4. TNF converts early M␾ into DCs in stromal cell/monocyte
cocultures. a, Monocytes were cultured for 5 days in GM-CSF and IL-4 lightly stained, cytoplasm and a centrally positioned nucleus, a
with normal skin fibroblasts. TNF (10 ng/ml) was added from day 0 to day phenotype consistent with M␾ differentiation (Fig. 2). On the con-
4 cocultures. At day 5, differentiation was monitored using expression of trary, CD1⫹ cells isolated from cultures performed with TNF dis-
CD1b/c and CD14. Similar results were obtained when the differentiation play a homogenous smooth intensely stained cytoplasm, cytoplas-
was measured at day 7. b, Monocytes were cultured in GM-CSF and IL-4 mic protrusions, and a laterally positioned nucleus, a morphology
with normal skin fibroblasts. From days 3 to 5, cultures were pulsed with
typical of DCs. Thus, TNF and IL-1 differentially modulate mono-
TNF (100 ng/ml) and blocking IL-6 and IL-6R mAbs (15 ␮g/ml) for 3
consecutive days. Forty-eight hours after the last pulse, cells were stained cyte differentiation.
for CD1b/c and CD14 expression. Results are representative of three to six
separate experiments.
TNF-induced cells display DC function
Next, we determined whether the cells generated in monocyte/
medium containing 7-aminoactinomycin D (Sigma-Aldrich). Confocal la- fibroblast cultures in the presence of TNF were able to prime naive
ser scanning microscopy was performed with a confocal microscope (TCS- T cells, a property of DCs. To this end, CD1⫹ cells from mono-
SP; Leica, Brussels, Belgium) equipped with ⫻63 oil objective.
cyte/fibroblast cultures or from control cultures without fibroblasts
Statistical analysis were sorted and used as stimulators in allogenic MLR with naive
CD4⫹CD45RA⫹ T cells. As shown in Fig. 3a, CD1⫹ cells sorted
A paired Student t test was used to determine the statistical significance of
the data. In all cases, values were log transformed before statistical anal- from day 5 TNF-treated cultures with or without fibroblasts induce
ysis. A value of p ⬍ 0.05 was chosen for rejection of the null hypothesis. a strong proliferation of naive allogenic CD4⫹ T cells consistent
with DC differentiation. Furthermore, fibroblasts do not inhibit the
Results function of differentiated DCs as cells retrieved from cultures with
Adding TNF to monocytes cultured with fibroblasts promotes fibroblasts are as potent as DCs generated in the absence of fibro-
monocyte differentiation toward DCs blasts (Fig. 3a). When compared with CD1⫹ cells isolated from
Under “steady-state” conditions fibroblasts skew monocyte differ- non-TNF cultures, DCs generated in the presence of TNF induce
entiation toward M␾ (5). We tested whether an inflammatory en- higher T cell proliferation, a finding consistent with the TNF-in-
vironment that contains IL-1 and/or TNF would alter such a dif- duced maturation.
ferentiation pathway. Adding TNF at the onset of the monocyte/ Indeed, TNF is a well-established DC maturation factor. How-
fibroblast cocultures in the presence of GM-CSF/IL-4 resulted in ever, DCs generated in TNF-stimulated monocyte/fibroblast cul-
the generation of cells with a DC phenotype. Thus, the analysis of tures were only partially mature as ⬃50% of the cells expressed
10 independent experiments revealed that ⬎95% of HLA-DR⫹ increased levels of CD83 (Fig. 3b), CD86, and, to a lesser extent,
cells generated in the presence of TNF express CD1 but not CD14. CD80 (data not shown). TNF-induced DCs were not blocked in
In contrast, ⬎90% of HLA-DR⫹ cells generated without adding their maturation capacity and could undergo further maturation
TNF express CD14 (Fig. 1a). Dose-response analysis revealed that when further cultured with CD40 ligand, TNF, or LPS (Fig. 3c and

FIGURE 5. TNF increases IL-6 and M-CSF

levels. Monocytes were cultured for 5 days with
or without fibroblasts in GM-CSF and IL-4 with
or without TNF (100 ng/ml) or IL-1␣␤ (50 ng/
ml). IL-6 and M-CSF levels were tested in day
5 supernatants. Results are expressed as
mean ⫾ SD of three independent experiments.
2266 TNF SKEWS MONOCYTES TOWARD DCs

Downloaded from http://journals.aai.org/jimmunol/article-pdf/171/5/2262/1168994/2262.pdf by guest on 29 April 2024

FIGURE 6. TNF down-regulates surface M-CSFR expression. Monocytes were cultured for 5 days with or without fibroblasts in GM-CSF and IL-4 with
or without TNF (100 ng/ml) or IL-1␣␤ (50 ng/ml). a, Cells were double stained for M-CSFR and HLA-DR. Expression of M-CSFR is shown on HLA-DR⫹
cells. Results are representative of three independent experiments. b, Kinetics of M-CSFR expression. Each day of culture, cells were double stained for
M-CSFR and HLA-DR. Expression of M-CSF receptors is shown on HLA-DR⫹ cells. c, Expression of M-CSFR genes on TNF-treated DCs. Monocytes
were cultured for 5 days in GM-CSF and IL-4 with or without TNF (100 ng/ml) and/or IL-6. At day 5, cells were harvested, RNA was extracted, and gene
expression was quantified by real-time RT-PCR. d, Expression of M-CSFR protein on TNF-treated DCs. Monocytes were cultured for 5 days in GM-CSF
and IL-4, washed, and cultured without cytokines (left) or with TNF (right) for 24 h.

data not shown). Thus, TNF promotes monocyte differentiation the high levels of fibroblast- derived IL-6 found in day 3 cultures
toward partially mature DCs. may predominate over TNF and maintain the CD14high phenotype.
Thus, day 3 monocyte/fibroblast cocultures were pulsed daily (for
TNF converts activated monocytes/early M␾ into DCs 3 days) with TNF along with Abs neutralizing IL-6/IL-6R until the
M␾ generated by culturing monocytes with M-CSF can convert cells were analyzed at day 7. As shown in Fig. 4b, blocking IL-6
into DCs upon exposure to GM-CSF and IL-4, even at late stages reversed resistance to TNF as determined by the disappearance of
of differentiation (16). Furthermore, as shown above, monocyte/ CD14⫹ cells and the presence of CD14⫺CD1⫹ DCs that consti-
fibroblast cocultures made in the presence of GM-CSF and IL-4 tuted ⬎90% of the HLA-DR⫹ population in monocyte/fibroblast
yield CD14high cells displaying M␾ morphology (Fig. 2). There- cultures. These results demonstrate the importance of the IL-6/
fore, we next determined whether delayed addition of TNF could TNF balance in determining the differentiation of monocytes into
lead to conversion of activated monocytes/early M␾ into DCs. As either DCs or M␾.
shown in Fig. 4a, adding TNF to monocyte/fibroblast cultures at
day 1 or day 2 resulted in the generation of DCs. However, when TNF offsets the IL-6/M-CSF pathway that promotes M␾
TNF was added at day 3 or later, the cells retained CD14 expres- differentiation
sion and did not acquire CD1. This resistance to TNF was not due Since blocking IL-6 facilitated TNF-induced monocyte/M␾ differ-
to a loss of TNFRs (data not shown). entiation into DCs, we surmised that TNF might interfere in the
We therefore wondered whether the resistance to TNF reflected monocyte response to IL-6/M-CSF. To this end, levels of IL-6 and
an irreversible commitment to M␾ differentiation. Alternatively, M-CSF were measured in culture supernatants. As shown in Fig. 5,
The Journal of Immunology 2267

Downloaded from http://journals.aai.org/jimmunol/article-pdf/171/5/2262/1168994/2262.pdf by guest on 29 April 2024

FIGURE 7. Inhibition of M-CSFR internalization by TNF. Day 5 DCs were incubated in GM-CSF and IL-4 with or without IL-6 (200 ng/ml), TNF (100
ng/ml), or a combination for 24 h. Cells were then exposed or not to M-CSF (50 ng/ml), washed, and stained for M-CSFR for flow cytometry (a) and
confocal analysis (b). Confocal photographs represent sections of cells from each culture condition labeled for M-CSFR (green) and nuclei (red). Results
are representative of three independent experiments.

in three experiments, monocyte/fibroblast cocultures spiked with M-CSF receptors. Adding TNF, but not IL-1␣ or IL-1␤, at the
TNF contained high levels of IL-6 (average 4-fold increase as onset of monocyte/fibroblast cocultures made with GM-CSF/IL-4
compared with GM-CSF/IL-4 cocultures, p ⬍ 0.0001) and of M- resulted in down-regulation of M-CSFR protein expression (Fig.
CSF (average 8-fold increase, p ⬍ 0.0001; Fig. 5). Supernatants of 6a). Such down-regulation was independent of the presence of
cultures spiked with IL-1 contained very high levels of IL-6 (up to stromal cells (Fig. 6b). The analysis of the kinetics of M-CSFR
250 ng/ml, average 12-fold increase as compared with control cul- expression showed that moderate down-regulation of its expres-
tures, n ⫽ 3) but only low levels of M-CSF (Fig. 5). These results sion could be observed already at the early stages of culture (days
suggest that TNF-driven DC differentiation may result from the 1 and 2; Fig. 6b).
ability of TNF to prevent the utilization of IL-6 and/or M-CSF. The down-regulation of M-CSFR expression was confirmed at
Conversely, IL-1 may promote M␾ differentiation by increasing the gene transcription level (Fig. 6c). In these experiments, M-
the levels of IL-6 and facilitating the utilization of M-CSF. CSFR mRNA expression (by real-time PCR) in monocytes was
Therefore, we investigated whether the apparent lack of cyto- analyzed either directly after isolation from blood or after 5 days
kine utilization in TNF cultures was due to down-regulation of of culture with GM-CSF and IL-4 with and without the addition of
2268 TNF SKEWS MONOCYTES TOWARD DCs

IL-6 and/or TNF at the onset of culture (without fibroblasts). M- We have previously shown that monocytes yield DCs with
CSFR mRNA expression in cultured cells was normalized for com- Langerhans cell (LCs) properties when cultured with GM-CSF and
parison to the expression in resting, i.e., uncultured, monocytes. IL-15 (25). In this study, we show that in the presence of fibro-
Adding TNF to GM-CSF/IL-4 DC cultures resulted in ⬎2-fold blasts and TNF, monocytes yield CD1a⫹DC-SIGN⫹ cells as well
decrease in the M-CSFR mRNA expression (Fig. 6c), while the as a subset of cells expressing surface Langerin, a marker of LCs.
expression of IL-6-R␣ chain was not affected (data not shown). TNF has been shown to up-regulate Langerin expression on the
The effect of TNF was prevailing even in the presence of exoge- surface of immature LCs obtained by culturing monocytes with
nous IL-6 (Fig. 6c). These results suggest that TNF acts in the GM-CSF and IL-4 supplemented with TGF-␤ (26). Whether the
early stages of monocyte differentiation and renders monocytes acquisition of Langerin expression in monocyte/fibroblast cocul-
unresponsive to IL-6/M-CSF via down-regulation of M-CSFR ex- tures represents a direct effect of TNF or whether it is mediated by
pression consistent with earlier observations in the mouse (17, 18). TNF-induced cytokines remains to be determined. Indeed, dermal
Furthermore, TNF down-regulates the expression of M-CSFR fibroblasts express cytoplasmic IL-15, which upon long-term (over
on already differentiated DCs (Fig. 6d). Indeed, when monocytes 3 days) exposure to TNF can be relocated to the cell surface (27).
cultured for 5 days with GM-CSF and IL-4 are exposed for 24 h to This membrane-bound form of IL-15, particularly when presented
TNF a considerable decrease of surface M-CSFR expression could to neighboring cells in the complex with IL-15R␣ (28), may dis-
be observed (Fig. 6d). Even though M-CSFR expression may be play a potent biological activity as demonstrated for example by
barely detectable by Ab staining and flow cytometry, a very low sustained T cell proliferation (27).
number (below detection limit by flow cytometry) number of re- Our results show that DCs generated in the presence of GM-

Downloaded from http://journals.aai.org/jimmunol/article-pdf/171/5/2262/1168994/2262.pdf by guest on 29 April 2024

ceptor molecules present on the cell surface might suffice for the CSF and TNF are “semimature” because only a fraction of cells
internalization of the receptor-ligand complex (12). To address express CD83 and they can undergo full maturation upon subse-
this, day 5 DCs generated with GM-CSF and IL-4 were cultured quent exposure to soluble CD40 ligand or LPS. The role of TNF
for 24 h with either IL-6 and/or TNF, washed, and subsequently in DC maturation is not entirely clear and the extent of TNF-
exposed to M-CSF for 30 min. M-CSFR expression on the cell induced DC maturation is likely to be dependent on the presence
surface was monitored by flow cytometry and its internalization of additional signals. Thus, while TNF maintains the viability of
was monitored by intracytoplasmic staining and confocal micros- freshly isolated murine LCs, GM-CSF is necessary for their func-
copy. As shown in Fig. 7, only IL-6-treated DCs displayed intense tional maturation (29). Furthermore, in vitro-generated human LCs
intracytoplasmic staining, suggesting receptor-ligand internaliza- exposed to TNF, in the presence of TGF-␤, up-regulate surface
tion. Contrary to IL-6, TNF-induced down-regulation of M-CSFR Langerin and CCR7 expression and migrate in response to CCR7
expression was accompanied by the inhibition of receptor inter- ligands (26). Such TNF-dependent acquisition of CCR7 expression
nalization. Thus, TNF stabilizes the phenotype of already differ- provides an explanation for the presence of immature LCs in the
entiated DCs and acts through down-regulation of M-CSFR ex- draining lymph node of the skin area involved by dermatopathic
pression and internalization. lymphadenitis (26). However, the classical DC maturation mark-
ers, DC-lysosome-associated membrane glycoprotein and co-
Discussion stimulatory molecules expression, are absent and the cells display
Our studies demonstrate the importance of IL-6 and TNF in the low T cell stimulatory activity, findings consistent with only partial
regulation of monocyte differentiation into M␾ or DCs. Our pre- maturation (26, 30). Finally, in a murine model of experimental
vious study has shown that monocytes cultured with fibroblasts autoimmune encephalitis, TNF-induced semimature DCs lead to
and GM-CSF/IL-4 yield predominantly M␾ in response to fibro- Ag-specific protection upon repeated injection (31). In this murine
blast-derived IL-6 (5). The current study shows that adding TNF at model, whose relevance to our current study may be limited, TNF-
the onset of monocyte/fibroblast cocultures skews monocyte dif- induced semimature DCs led to the generation of Ag-specific IL-
ferentiation toward DCs. TNF overcomes IL-6-driven M␾ differ- 10-producing T cells in vivo (31). Yet, studies in TNF-deficient
entiation by down-regulating the expression and internalization of mice exposed to recombinant adenovirus reveal impaired virus-
the M-CSFR on activated monocytes. Although the inhibitory ef- specific T cell proliferation and impaired maturation of DCs in
fect of TNF on the M-CSFR expression has been described earlier draining lymph nodes as compared with wild-type mice (32). In
in murine M␾ (17), our data show the consequence of this process, vitro, TNF was required to mature DCs efficiently during virus-
i.e., skewing of human monocyte differentiation into DCs at the mediated stimulation. Furthermore, adoptive transfer of primed,
expense of M␾. mature DCs results in restored T cell responses and specific hu-
Skewing of monocyte differentiation toward DCs is specific to moral responses (32). Thus, by virtue of modulation of DC mat-
TNF and could not be obtained by replacing TNF with IL-1. In- uration TNF may play a crucial role in the induction of tolerance
terestingly, both TNF and IL-1 are essential for migration of DCs or immunity.
from tissue to secondary lymphoid organs as demonstrated in vitro The demonstration that TNF promotes DC differentiation in the
in human skin cultures (19, 20) as well as in vivo in mice upon presence of stromal cells may contribute to our understanding of
topical (19) and/or systemic administration of TNF and/or IL-1 pathophysiological steps occurring in rheumatoid arthritis (RA),
(21). Yet, our results show the distinct role of each cytokine in DC one of the most common autoimmune diseases in humans. Indeed,
differentiation from blood precursors. The lack of DCs in mono- TNF plays a key role in the pathogenesis of RA and the develop-
cyte/fibroblast cocultures with IL-1 might be explained by contin- ment of anti-TNF therapy represents a success of immunology (12,
uous M-CSF internalization through sustained or up-regulated ex- 13, 33). TNF, initially thought to originate from in situ M␾, can be
pression of the M-CSFR as shown previously in human monocytes produced locally (likely at high levels) by mast cells (34, 35). Mast
(22) as well as in mice (18). Alternatively, endogenous M-CSF cells are also able to release high amounts of IL-4 (36). Given the
and/or exogenous GM-CSF and IL-4 can result in the increased fact that GM-CSF can be released by a wide variety of cells, in-
expression of the IL-1R antagonist (23, 24). The IL-1R antagonist cluding mast cells and synovial fibroblasts (37, 38), all cytokines
is likely to counteract the biological effects of IL-1 (11, 15) in- used in our model may be present at the joint and result in the
cluding its potential activity in skewing monocyte differentiation skewing of newly arriving monocytes into DCs. Indeed, the infil-
toward DCs. tration of RA synovium with DCs has been demonstrated (39, 40).
The Journal of Immunology 2269

Furthermore, consistent with the concept that RA synovium rep- 20. Stoitzner, P., M. Zanella, U. Ortner, M. Lukas, A. Tagwerker, K. Janke,
M. B. Lutz, G. Schuler, B. Echtenacher, B. Ryffel, et al. 1999. Migration of
resents an ectopic lymphoid tissue, a distinct localization of both Langerhans cells and dermal dendritic cells in skin organ cultures: augmentation
immature and mature DCs can be observed (40). In lieu of our by TNF-␣ and IL-1␤. J. Leukocyte Biol. 66:462.
results, by promoting DC differentiation, TNF may facilitate au- 21. Roake, J. A., A. S. Rao, P. J. Morris, C. P. Larsen, D. F. Hankins, and
J. M. Austyn. 1995. Dendritic cell loss from nonlymphoid tissues after systemic
toantigen presentation and perpetuation of the disease. administration of lipopolysaccharide, tumor necrosis factor, and interleukin 1.
Thus, TNF appears as a major factor promoting DC differenti- J. Exp. Med. 181:2237.
ation in tissues. By skewing monocyte differentiation toward DCs 22. Gruber, M. F., C. C. Williams, and T. L. Gerrard. 1994. Macrophage-colony-
stimulating factor expression by anti-CD45 stimulated human monocytes is tran-
and by inducing a partial DC activation/maturation sufficient for scriptionally up-regulated by IL-1␤ and inhibited by IL-4 and IL-10. J. Immunol.
the cells to migrate, TNF may facilitate the induction of adaptive 152:1354.
23. Matsushime, H., M. F. Roussel, K. Matsushima, A. Hishinuma, and C. J. Sherr.
immunity and host protection. IL-6, by skewing monocytes newly 1991. Cloning and expression of murine interleukin-1 receptor antagonist in mac-
arriving at the site of tissue damage toward M␾, may favor innate rophages stimulated by colony-stimulating factor 1. Blood 78:616.
defense mechanisms and Ag degradation. Our results thus syner- 24. Jenkins, J. K., and W. P. Arend. 1993. Interleukin 1 receptor antagonist produc-
tion in human monocytes is induced by IL-1␣, IL-3, IL-4 and GM-CSF. Cytokine
gize with earlier studies demonstrating the crucial role of TNF in 5:407.
skewing hemopoietic progenitor differentiation toward DCs (41, 25. Mohamadzadeh, M., F. Berard, G. Essert, C. Chalouni, B. Pulendran, J. Davoust,
42) and show that TNF can act not only on hemopoietic progen- G. Bridges, A. K. Palucka, and J. Banchereau. 2001. Interleukin 15 skews mono-
cyte differentiation into dendritic cells with features of Langerhans cells. J. Exp.
itors but also on blood precursors. Med. 194:1013.
26. Geissmann, F., M. C. Dieu-Nosjean, C. Dezutter, J. Valladeau, S. Kayal,
Acknowledgments M. Leborgne, N. Brousse, S. Saeland, and J. Davoust. 2002. Accumulation of

Downloaded from http://journals.aai.org/jimmunol/article-pdf/171/5/2262/1168994/2262.pdf by guest on 29 April 2024

immature Langerhans cells in human lymph nodes draining chronically inflamed
We thank Elizabeth T. Kraus and Sebastien Coquery for help with flow skin. J. Exp. Med. 196:417.
cytometry, Sandra Clayton for help with confocal microscopy, and Victoria 27. Rappl, G., A. Kapsokefalou, C. Heuser, M. Rossler, S. Ugurel, W. Tilgen,
Cantrell for help with real-time PCR. U. Reinhold, and H. Abken. 2001. Dermal fibroblasts sustain proliferation of
activated T cells via membrane-bound interleukin-15 upon long-term stimulation
with tumor necrosis factor-␣. J. Invest. Dermatol. 116:102.
References 28. Dubois, S., J. Mariner, T. A. Waldmann, and Y. Tagaya. 2002. IL-15R␣ recycles
1. Randolph, G. J., S. Beaulieu, S. Lebecque, R. M. Steinman, and W. A. Muller. and presents IL-15 in trans to neighboring cells. Immunity 17:537.
1998. Differentiation of monocytes into dendritic cells in a model of transendo- 29. Koch, F., C. Heufler, E. Kampgen, D. Schneeweiss, G. Bock, and G. Schuler.
thelial trafficking. Science 282:480. 1990. Tumor necrosis factor ␣ maintains the viability of murine epidermal Lang-
2. Randolph, G. J., K. Inaba, D. F. Robbiani, R. M. Steinman, and W. A. Muller. erhans cells in culture, but in contrast to granulocyte/macrophage colony-stimu-
1999. Differentiation of phagocytic monocytes into lymph node dendritic cells in lating factor, without inducing their functional maturation. J. Exp. Med. 171:159.
vivo. Immunity 11:753. 30. Geissmann, F., P. Revy, A. Regnault, Y. Lepelletier, M. Dy, N. Brousse,
3. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of S. Amigorena, O. Hermine, and A. Durandy. 1999. TGF-␤1 prevents the non-
immunity. Nature 392:245. cognate maturation of human dendritic Langerhans cells. J. Immunol. 162:4567.
4. Banchereau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y. Liu, 31. Menges, M., S. Rossner, C. Voigtlander, H. Schindler, N. A. Kukutsch,
B. Pulendran, and K. Palucka. 2000. Immunobiology of dendritic cells. Annu. C. Bogdan, K. Erb, G. Schuler, and M. B. Lutz. 2002. Repetitive injections of
Rev. Immunol. 18:767. dendritic cells matured with tumor necrosis factor ␣ induce antigen-specific pro-
5. Chomarat, P., J. Banchereau, J. Davoust, and A. K. Palucka. 2000. IL-6 switches tection of mice from autoimmunity. J. Exp. Med. 195:15.
the differentiation of monocytes from dendritic cells to macrophages. Nat. Im- 32. Trevejo, J. M., M. W. Marino, N. Philpott, R. Josien, E. C. Richards, K. B. Elkon,
munol. 1:510. and E. Falck-Pedersen. 2001. TNF-␣-dependent maturation of local dendritic
6. Steinman, R. M. 1991. The dendritic cell system and its role in immunogenicity. cells is critical for activating the adaptive immune response to virus infection.
Annu. Rev. Immunol. 9:271. Proc. Natl. Acad. Sci. USA 98:12162.
7. Van Furth, R. 1988. Phagocytic cells: development and distribution of mononu- 33. Elliott, M. J., R. N. Maini, M. Feldmann, A. Long-Fox, P. Charles, H. Bijl, and
clear phagocytes in normal steady state and inflammation. In Inflammation: Basic J. N. Woody. 1994. Repeated therapy with monoclonal antibody to tumour ne-
Principles and Clinical Correlates. J. I. Gallin, I. M. Goldstein, and crosis factor ␣ (cA2) in patients with rheumatoid arthritis. Lancet 344:1125.
R. Snyderman, eds., Lippincott Williams & Wilkins, Philadelphia, p. 281. 34. Benoist, C., and D. Mathis. 2002. Mast cells in autoimmune disease. Nature
8. Rossi, D., and A. Zlotnik. 2000. The biology of chemokines and their receptors. 420:875.
Annu. Rev. Immunol. 18:217. 35. Woolley, D. E. 2003. The mast cell in inflammatory arthritis. N. Engl. J. Med.
9. Taga, T., and T. Kishimoto. 1997. Gp130 and the interleukin-6 family of cyto- 348:1709.
kines. Annu. Rev. Immunol. 15:797. 36. Brown, M. A., J. H. Pierce, C. J. Watson, J. Falco, J. N. Ihle, and W. E. Paul.
10. Matsuda, T., and T. Hirano. 2001. IL-6. In Cytokine Reference. 1987. B cell stimulatory factor-1/interleukin-4 mRNA is expressed by normal and
J. Oppenheim and M. Feldmann, eds. Academic Press, New York, p. 537. transformed mast cells. Cell 50:809.
11. Arend, W. P., M. Malyak, C. J. Guthridge, and C. Gabay. 1998. Interleukin-1 37. Dechanet, J., J. L. Taupin, P. Chomarat, M. C. Rissoan, J. F. Moreau,
receptor antagonist: role in biology. Annu. Rev. Immunol. 16:27. J. Banchereau, and P. Miossec. 1994. Interleukin-4 but not interleukin-10 inhibits
12. Tracey, K. J., and A. Cerami. 1994. Tumor necrosis factor: a pleiotropic cytokine the production of leukemia inhibitory factor by rheumatoid synovium and syno-
and therapeutic target. Annu. Rev. Med. 45:491. viocytes. Eur. J. Immunol. 24:3222.
13. Feldmann, M., and R. N. Maini. 2001. Anti-TNF ␣ therapy of rheumatoid ar- 38. Fossiez, F., O. Djossou, P. Chomarat, L. Flores-Romo, S. Ait-Yahia, C. Maat,
thritis: what have we learned? Annu. Rev. Immunol. 19:163. J. J. Pin, P. Garrone, E. Garcia, S. Saeland, et al. 1996. T cell interleukin-17
14. Feldmann, M., F. M. Brennan, and R. N. Maini. 1996. Role of cytokines in induces stromal cells to produce proinflammatory and hematopoietic cytokines.
rheumatoid arthritis. Annu. Rev. Immunol. 14:397. J. Exp. Med. 183:2593.
15. Dinarello, C. A. 2000. The role of the interleukin-1-receptor antagonist in block- 39. Woolley, D. E., E. D. Harris, Jr., C. L. Mainardi, and C. E. Brinckerhoff. 1978.
ing inflammation mediated by interleukin-1. N. Engl. J. Med. 343:732. Collagenase immunolocalization in cultures of rheumatoid synovial cells. Science
16. Palucka, K. A., Taquet, N., Sanchez-Chapui, F., and J. C. Gluckman. 1998. Den- 200:773.
dritic cells as the terminal stage of monocyte differentiation. J. Immunol. 40. Page, G., S. Lebecque, and P. Miossec. 2002. Anatomic localization of immature
160:4587. and mature dendritic cells in an ectopic lymphoid organ: correlation with selec-
17. Shieh, J. H., R. H. Peterson, D. J. Warren, and M. A. Moore. 1989. Modulation tive chemokine expression in rheumatoid synovium. J. Immunol. 168:5333.
of colony-stimulating factor-1 receptors on macrophages by tumor necrosis fac- 41. Caux, C., C. Dezutter-Dambuyant, D. Schmitt, and J. Banchereau. 1992. GM-
tor. J. Immunol. 143:2534. CSF and TNF-␣ cooperate in the generation of dendritic Langerhans cells. Nature
18. Shieh, J. H., R. H. Peterson, and M. A. Moore. 1991. IL-1 modulation of cytokine 360:258.
receptors on bone marrow cells: in vitro and in vivo studies. J. Immunol. 42. Caux, C., B. Vanbervliet, C. Massacrier, C. Dezutter-Dambuyant,
147:1273. B. de Saint-Vis, C. Jacquet, K. Yoneda, S. Imamura, D. Schmitt, and
19. Enk, A. H., V. L. Angeloni, M. C. Udey, and S. I. Katz. 1993. An essential role J. Banchereau. 1996. CD34⫹ hematopoietic progenitors from human cord blood
for Langerhans cell-derived IL-1␤ in the initiation of primary immune responses differentiate along two independent dendritic cell pathways in response to GM-
in skin. J. Immunol. 150:3698. CSF⫹TNF ␣. J. Exp. Med. 184:695.