The Journal of Immunology
Ubiquitination of CD86 Is a Key Mechanism in Regulating
Antigen Presentation by Dendritic Cells
Gu ¨nther Baravalle,* Hyesuk Park,* Megan McSweeney,* Mari Ohmura-Hoshino,†
Yohei Matsuki,†Satoshi Ishido,†and Jeoung-Sook Shin*
Dendritic cells (DCs) require costimulatory molecules such as CD86 to efficiently activate T cells for the induction of adaptive
immunity. DCs maintain minimal levels of CD86 expression at rest, but upregulate levels upon LPS stimulation. LPS-
stimulated DCs produce the immune suppressive cytokine IL-10 that acts in an autocrine manner to regulate CD86 levels. Inter-
thatCD86 is ubiquitinated inDCsviaMARCH1 E3ubiquitinligaseandthatthis ubiquitinationplaysa keyroleinCD86 regulation.
Ubiquitination at lysine 267 played the most critical role for this regulation. CD86 is ubiquitinated in MARCH1-deficient DCs to
is continuously ubiquitinated in DCs following activation by LPS, and this was due to the autocrine IL-10 inhibition of MARCH1
downregulation. Accordingly, DCs lacking MARCH1 and DCs expressing ubiquitination-resistant mutant CD86 both failed to reg-
ulate CD86 in response to autocrine IL-10. DCs expressing ubiquitination-resistant mutant CD86 failed to control their T cell-
activating abilities at rest as well as in response to autocrine IL-10. These studies suggest that ubiquitination serves as an important
mechanism by which DCs control CD86 expression and regulate their Ag-presenting functions.
the costimulatory signals that significantly lower the activation
threshold and allow naive T cells to be readily activated (1, 2). DCs
dynamically regulate CD86 levels depending on their need to har-
ness or reduce their T cell-activating ability. DCs upregulate CD86
expression following their contact with microbial Ags (3), which
presumably facilitates activation of microbe-specific T cells capa-
ble of combating infection. DCs downregulate CD86 expression
following exposure to IL-10, an immune-suppressive cytokine
produced by various immune cells (4–6). Importantly, DCs them-
selvesproduce IL-10following stimulationby LPS(4, 5). ThisDC-
produced IL-10 inhibits LPS-induced CD86 increase in DCs by
acting in an autocrine manner, thus preventing excessive CD86
Themolecular mechanismunderlyingCD86regulationhas begun
to emerge. A recent study demonstrated that DCs of MARCH1-
deficient mice expressed significantly elevated CD86 levels, sug-
gesting a role for MARCH1 in CD86 regulation (7). MARCH1 is a
The Journal of Immunology, 2011,
D86 is a costimulatory molecule that is mainly expressed
membrane-anchored E3 ubiquitin ligase that has previously been
shown to ubiquitinate MHC class II mediating its lysosomal deg-
radation (8, 9). A more recent study has shown that activated human
monocytes dramatically upregulate MARCH1 expression upon in-
cubation with recombinant IL-10 (10). This MARCH1 upregulation
inversely correlated with CD86 expression (10). Most recently, Tze
et al. (11) have demonstrated that mouse bone marrow-derived DCs
(BMDCs) also upregulated MARCH1 when they were incubated
with IL-10. Furthermore, pretreatment of BMDCs with IL-10 ef-
fectively suppressed LPS-induced CD86 increase only when DCs
expressed MARCH1 (11). This study suggested that MARCH1
plays an important role in IL-10–mediated regulation of CD86 in
DCs, but the specific mechanism by which MARCH1 regulates
CD86 was not clearly defined. None of the studies thus far have
demonstrated whether CD86 is ubiquitinated and, if so, whether this
ubiquitination is achieved by MARCH1 in DCs.
MARCH1’s only physiologic substrate identified thus far is
MHC class II (8, 12). In this study, we demonstrate that CD86 is
ubiquitinated in DCs and that this ubiquitination is also mediated
by MARCH1. We further demonstrate that CD86 ubiquitination
plays an important role in DCs to regulate their CD86 expression
and their T cell-activating abilities both at rest and in response to
Materials and Methods
Mice and cells
BMDCs from C57/BL6 mice (The Jackson Laboratory), CD862/2mice
(The Jackson Laboratory), and MARCH12/2mice (8) were grown as
previously described (13). B blasts from CD862/2mice were generated by
treating splenic B cells with 25 mg/ml LPS overnight. CD8+T cells from
OT-I mice (The Jackson Laboratory) were isolated from lymph nodes
using a T cell isolation kit (StemCell Technologies).
Plasmids and retroviral transduction
The cDNA of mouseCD86and MARCH1 was cloned into the LZRS-pBMN
plasmid using EcoRI and XhoI sites. Internal ribosome entry site-enhanced
GFP (EGFP) was previously inserted using NotI sites of the LZRS vector
*Department of Microbiology and Immunology, Sandler Asthma Basic Research
Center, University of California, San Francisco, San Francisco, CA 94143; and
†Laboratory for Infectious Immunity, RIKEN Research Center for Allergy and Im-
munology, Yokohama, Kanagawa 230-0045, Japan
Received for publication June 3, 2011. Accepted for publication July 18, 2011.
This work was supported by the Sandler Asthma Basic Research Center, a National
Institutes of Health training grant (to M.M.), and the Cancer Research Institute (to
Address correspondence and reprint requests to Dr. Jeoung-Sook Shin, Department of
Microbiology and Immunology, Sandler Asthma Basic Research Center, University
of California, San Francisco, 513 Parnassus Avenue, HSE-201P, San Francisco, CA
94143. E-mail address: firstname.lastname@example.org
The online version of this article contains supplemental material.
Abbreviations used in this article: BMDC, bone marrow-derived dendritic cell; DC,
dendritic cell; EGFP, enhanced GFP; K, lysine; R, arginine.
(13). All CD86 mutant constructs and the mutant MARCH1 (I66A, W97A)
construct were generated by site-directed mutagenesis. The CD86-EGFP
construct was generated by PCR, eliminating the stop codon of CD86,
and cloning into LZRS-EGFP (where EGFP was inserted between the
BamHI and EcoRI sites) using a BamHI site. mCherry-T2A–MARCH
constructs were generated by linking the cDNAs of mCherry and MARCH
with the self-cleaving 2A peptide (14). Generation of retrovirus and sub-
sequent DC transduction was performed as previously described (13).
Flow cytometry and cell sorting
Cells were incubated with indicated fluorophore-conjugated Abs and
analyzed on an FACS LSRII system (BD Biosciences). All transduced
BMDCs were presorted on an FACSAria cell sorter (BD Biosciences). Only
GFPlowDCs were used in the experiments because the levels of CD86
expressed on GFPlowDCs approximated levels of endogenous CD86 ex-
pression. BMDCs and splenic DCs used to immunoprecipitate endogenous
CD86 were prepurified with a murine DC isolation kit (Miltenyi Biotec).
Immunoprecipitation and Western blot
BMDCs or splenic DCs were solubilized in Triton-lysis buffer as de-
scribed (13). The soluble cell lysates were mixed with CD86 Ab (GL-1)
and incubated for 2 h at 4˚C. Protein G-Sepharose beads were added,
and samples were incubated for an additional 2 h at 4˚C. The beads were
washed three times with Triton-lysis buffer prior to Peptide-N–Glyco-
sidase F treatment. Thereafter, samples were run for SDS-PAGE, blotted
to polyvinylidene difluoride membranes, and developed with the indi-
Plasma membrane proteins of CD11c bead-purified DCs pretreated with
LPS in the presence or absence of anti–IL-10R Ab were biotinylated by
incubating cells with the membrane-impermeable biotinylation reagent
sulfo-NHS–biotin according to the manufacturer’s protocol (Pierce).
In vitro T cell activation assay
BMDCs were incubated with increasing amounts of SIINFEKL peptide and
CD8+OT-I T cells at a 1:10 DC/T cell ratio. After overnight culture, IL-2
concentrations in the supernatant were determined by ELISA.
Quantitative RT-PCR analysis
Total RNA was isolated by TRIzol (Invitrogen) extraction, and first-strand
(dT) as primers. Quantitative PCR was performed using SYBR Green
reagents (5 prime) on an Eppendorf realplex system. Samples were nor-
malized for b-actin expression.
CD86 is ubiquitinated in DCs
To determine if CD86 is ubiquitinated in DCs, we immunopreci-
pitated CD86 from BMDCs and performed Western blot analysis.
The anti-CD86 Ab reacted with proteins ranging from 70–80 kDa,
consistent with glycosylated forms of CD86 (Fig. 1A, right panel).
The anti-ubiquitin Ab detected proteins ranging from 100–150 kDa
(Fig. 1A, left panel), consistent with glycosylated CD86 that has
multiple ubiquitin molecules attached. Treatment of the immuno-
precipitates with a deglycosylating enzyme, PNGase F, resulted
in an expected downward shift of CD86 to a single band at
∼32 kDa (Fig. 1A, right panel). Immunoblotting for ubiquitinated
proteins demonstrated a similar downward shift of the ubiquiti-
nated species, which formed a ladder at increments of ∼8 kDa,
consistent with the addition of ubiquitin monomers (Fig. 1A, left
panel). The fusion of EGFP to CD86 increased the m.w. of CD86,
as shown by the upward shift of the anti-CD86 Ab-reactive protein
band (Fig. 1B, right panel). Accordingly, the ubiquitin ladder also
shifted upwards, indicating that the ubiquitinated species is indeed
CD86 (Fig. 1B, left panel). Lastly, we also verified ubiquitination
of endogenous CD86 in splenic DCs (Fig. 1C). Taken together,
these data demonstrate CD86 ubiquitination in DCs.
CD86 ubiquitination limits CD86 expression in DCs
We then examined the specific role of CD86 ubiquitination in
DCs. To create an ubiquitination-resistant CD86, we generated a
retroviral construct in which all five cytoplasmic lysines (K) of
CD86 were replaced by arginines (R). We also generated a retrovi-
rus encoding wild-type CD86 for a comparison. Both retroviral
constructs additionally encoded GFP bicistronically. Each retrovi-
rus was transduced into DCs cultured from CD86-deficient mice.
tinated species appeared with a ladder-like pattern (Fig. 2A), similar
to endogenous CD86 (Fig. 1A, 1C). However, ubiquitinated species
were absent in the immunoprecipitate of CD86 (K . R) mutant
(Fig. 2A), confirming that CD86 (K . R) mutant is resistant
to ubiquitination. Concurrently, we found that total CD86 (Fig. 2A)
and surface CD86 (Fig. 2B) levels were both elevated in DCs ex-
pressing CD86 (K . R) mutant. However, GFP and CD80 expres-
sion levels were similar in DCs expressing wild-type and mutant
CD86, indicating comparable efficiencies of transduction and lev-
els of maturation in these cells (Fig. 2B). These findings indicate
that DCs regulate CD86 expression via ubiquitination.
Ubiquitination at lysine 267 plays the primary role in
CD86 has five cytoplasmic lysines, which can all be potential
targets of ubiquitination (Fig. 3A). We examined whether all five
of the cytoplasmic lysines are equally important for the ubiquiti-
nation and regulation of CD86 or whether a specific lysine is
more critical than others. We made retroviral constructs encoding
a series of CD86 mutants, in which the cytoplasmic lysines (K)
were sequentially replaced with arginines (R). As shown in Fig.
3B, lysine 267 was found to be most critical and also sufficient for
the regulation of CD86 surface expression (see KRKKK and
RKRRR). To determine whether lysine 267 is the specific target of
ubiquitination, we expressed wild-type CD86, CD86 (RRRRR),
CD86 (RKRRR), and CD86 (KRKKK) in CD86-deficient DCs
and analyzed the ubiquitination of CD86. As shown in Fig. 3C,
the CD86 (RKRRR) mutant was ubiquitinated to the same degree
as wild-type CD86. Its total CD86 levels were comparably low.
was immunoprecipitated (IP) from mouse BMDCs,
treated with or without PNGase F, and immunoblotted
(IB). HC indicates the H chain of anti-CD86 Ab. B,
CD862/2BMDCs were retrovirally transduced to ex-
press wild-type CD86 or CD86-EGFP fusion proteins.
CD86 was immunoprecipitated (IP), treated with
PNGase F, and immunoblotted (IB). C, Splenic DCs
were isolated using CD11c magnetic beads. CD86
immunoprecipitates were treated and immunoblotted
as described. Asterisk indicates a ladder of proteins
reacting with anti-Ub Ab.
CD86 is ubiquitinated in DCs. A, CD86
The Journal of Immunology2967
Interestingly, CD86 (KRKKK) was also ubiquitinated to a similar
degree as wild-type CD86, However, its total levels were much
higher compared with wild-type CD86, but still lower than CD86
(RRRRR) (Fig. 3C). These findings indicate that multiple lysines
can be ubiquitinated and involved in CD86 regulation in DCs.
However, lysine 267 ubiquitination exerts the most potent and
sufficient effect for CD86 regulation.
CD86 is ubiquitinated by MARCH1
Having found that CD86 is ubiquitinated in DCs, we examined
whether the ubiquitination is mediated by MARCH1. As shown
previously (7), BMDCs cultured from MARCH1-deficient mice
expressed much higher levels of CD86 than BMDCs cultured
from wild-type mice (Fig. 4A). Notably, other hematopoietic cells
also expressed elevated levels of CD86 in MARCH1-deficient
mice (Supplemental Fig. 1). The expression of other costimula-
tory molecules such as CD80, CD83, and CD40 and adhesion
molecules such as ICAM-1 remained identical (data not shown).
To determine the requirement of MARCH1 in CD86 ubiquitina-
tion, we performed the same immunoprecipitation experiment as
described earlier using MARCH1-deficient BMDCs. We found
a marked reduction in CD86 ubiquitination in MARCH1-deficient
DCs (Fig. 4B), whereas overall ubiquitination was not affected
(data not shown). Reduced CD86 ubiquitination was accompanied
by elevated total CD86 (Fig. 4B). Importantly, CD86 increases due
to MARCH1 deficiency could be rescued by the retroviral ex-
pression of wild-type MARCH1, but not by the expression of
E3 ligase mutant MARCH1 (Fig. 4C). This finding shows that
MARCH1 regulates CD86 expression via its ubiquitin ligase ac-
tivity, but not via any other accessory function. Taken together,
these data demonstrate the important role of MARCH1 in medi-
ating CD86 ubiquitination in DCs.
Efficient CD86 regulation by MARCH1 requires lysine 267
A small but appreciable amount of ubiquitinated CD86 was de-
tected in MARCH1-deficient DCs, implying that other ubiquitin
ligases could also mediate CD86 ubiquitination in DCs (Fig. 4B).
Previous studies have shown that forced expression of MARCH1,
MARCH2, and MARCH8 resulted in a marked reduction of sur-
face CD86 in transfected HeLa cells and other cell lines (15, 16).
This finding implies that MARCH2 and MARCH8 could also
ubiquitinate CD86, thus participating in its regulation. To better
state DCs. A, BMDCs cultured from CD862/2mice were retrovirally
transduced to express wild-type CD86 (WT) and CD86 mutant in which
all of the five cytoplasmic lysines of CD86 were replaced with arginines
(K . R). Each retroviral construct included internal ribosome entry site-
EGFP cDNA downstream of the CD86 sequence. CD86 immunoprecipi-
tates (IP) were treated with PNGase F and then immunoblotted (IB). Cell
lysates were also immunoblotted. B, Flow cytometry analysis of DCs ex-
pressing WT CD86 and CD86 (K . R) mutant.
CD86 ubiquitination regulates CD86 expression in steady-
CD86 in DCs. A, Amino acid sequence of the cytoplasmic domain of mouse
CD86. Lysine residues (K) are numbered and highlighted in bold. B, Flow
cytometry analysis of CD862/2BMDCs expressing wild-type CD86 and
a series of CD86 mutants. Wild-type CD86 is designated as KKKKK to
indicate five lysines at the cytoplasmic sites numbered. Each of CD86
mutants is designated as a combination of K and R to indicate the specific
site(s) where lysine is replaced with arginine (R). The relative surface ex-
pression of CD86 in the cells expressing CD86 wild-type and mutants is
shown as mean fluorescence intensity (MFI). Data are representative of three
independent experiments. C, Western blot analysis of CD86 immunopreci-
pitates and lysates prepared from DCs expressing wild-type CD86 (WT) or
CD86 (K . R), CD86 (RKRRR), and CD86 (KRKKK) mutants.
Ubiquitinationvia lysine 267 is sufficient for the regulation of
A, Flow cytometry analysis of BMDCs generated from wild-type (WT)
mice (solid gray) or MARCH12/2mice (full line). B, BMDCs cultured
from WT or MARCH12/2mice (M1KO) were processed for immuno-
precipitation of CD86 followed by Western blot analysis. C, Flow
cytometry analysis of surface CD86 on MARCH12/2BMDCs transduced
or untransduced with retrovirus encoding wild-type MARCH1 or mutant
MARCH1 (I66A, W97A).
MARCH1-mediated ubiquitination regulates CD86 in DCs.
2968MARCH1-MEDIATED UBIQUITINATION OF CD86 IN DENDRITIC CELLS
characterize the role of these MARCH proteins, we examined
their specificity for lysines of CD86. Specifically, we wished to
determine whether MARCH1, MARCH2, or MARCH8 can ade-
quately regulate CD86 via lysine 267.
As a model, we employed mouse B blasts, in which CD86 sur-
face expression was not affected by the mutation of CD86 lysines
(Fig. 5A) unless MARCH1, MARCH2, or MARCH8 were coex-
pressed (Fig. 5B–D). We transduced CD86-deficient B blasts using
two retroviral constructs. One virus encoded for GFP and CD86,
either wild-type or lysine mutant, and the other virus encoded for
mCherry, a red fluorescent protein, and a MARCH ubiquitin li-
gase. Following transduction, we examined CD86 surface levels in
mCherry+GFP+cells. Upon MARCH1 coexpression, we found that
CD86 levels were markedly reduced in cells expressing wild-type
CD86 (KKKKK), but not in cells expressing cytoplasmic lysine-
free CD86 (RRRRR) (Fig. 5B), similarly to what we observed in
DCs. Subsequent analysis of other CD86 mutant-expressing cells
revealed that lysine 267 was critical for MARCH1 to fully down-
regulate surface CD86 to wild-type CD86 levels (see RRKKK and
KRKKK). Lysine 267 was also sufficient for MARCH1 to reduce
CD86 to its full degree (see RKRRR). Interestingly, however, ly-
sine 267 was neither necessary nor sufficient for MARCH2 to re-
mediated regulation was critically dependent on lysine 280 (see
KKRRR and RKRRR). MARCH8, similar to MARCH1, could
sufficiently regulate CD86 via lysine 267 alone (Fig. 5D, compare
KKKKK with RKRRR). However, MARCH8 effectively reduced
the surface expression of all other lysine mutants except for lysine-
free CD86. But notably, even lysine-free CD86 was appreciably
reduced by MARCH8 (see RRRRR) (Fig. 5D). Taken together,
the lysine 267 sufficiency observed in DCs was recapitulated in B
blasts expressing MARCH1 or MARCH8. However, the lysine
specificity observed in DCs was most similarly reproduced in B
blasts expressing MARCH1, implicating a major role of MARCH1
in CD86 regulation in DCs.
CD86 ubiquitination in DCs is facilitated by autocrine IL-10
A recent study has implicated an important role of MARCH1
in IL-10–mediated suppression of CD86 in DCs (11). However,
whether this suppression is mediated by CD86 ubiquitination is
not clear. DCs not only react to exogenous IL-10. Upon activa-
tion, DCs endogenously produce IL-10, which can then act in an
autocrine manner to prevent excessive CD86 expression (4, 5).
However, the underlying mechanism has not been elucidated.
Therefore, we investigated whether the regulation of CD86 by
autocrine IL-10 in DCs involves CD86 ubiquitination and, if so,
whether this ubiquitination is mediated by MARCH1.
Consistent with previous reports (4, 5), blockade of autocrine
IL-10 signaling by adding an anti–IL-10R Ab markedly enhanced
the LPS-induced increase of surface CD86 in DCs (Fig. 6A). No-
tably, this enhanced surface CD86 expression did not involve any
increase in CD86 transcription (Fig. 6B), but involved a decrease in
CD86 ubiquitination and an increase in total CD86 (Fig. 6C, 6D).
Because ubiquitination mediates endocytosis and subsequent deg-
reduced CD86 ubiquitination following anti–IL-10R Ab treatment
involves a reduced turnover of surface CD86. To test this, we in-
cubated DCs with membrane-impermeable biotinylating agents at
4˚C to label cell-surface proteins. Half of the cells were lysed im-
mediately, and the other half was further incubated at 37˚C for 16 h
prior to lysis. Each cell lysate was immunoprecipitated using anti-
CD86 Ab, and the precipitates were blotted with HRP-conjugated
streptavidin to determine the levels of biotinylated CD86. We
found ∼80% reduction of biotinylated CD86 within 16 h in
LPS-stimulated DCs (Fig. 6E, 6F), indicating that 80% of CD86
molecules that were on the surface have been degraded during
this 16 h chase. However, the degradation was markedly delayed
when DCs had been treated with anti–IL-10R Ab. Taken together,
IL-10 produced by LPS-stimulated DCs exerts an inhibitory ef-
fect on CD86 expression by promoting CD86 ubiquitination and
turnover and not by reducing CD86 transcription.
Regulation of CD86 in DCs via autocrine IL-10 is dependent
To determine the role of MARCH1 in autocrine IL-10–mediated
regulation of CD86, we first examined the expression of MARCH1
in DCs in response to autocrine IL-10. We found that blockade of
IL-10 signaling resulted in a significant reduction in MARCH1
expression inDCs(Fig.7A).Thisreductioncorrelated wellwiththe
reduction in CD86 ubiquitination (Fig. 6C, 6D). Next, we deter-
mined whether MARCH1 is required for autocrine IL-10–mediated
CD86 regulation. Blockade of IL-10 signaling markedly enhanced
the LPS-inducedCD86 increase in MARCH1-expressing wild-type
DCs, but not in MARCH1-deficient DCs (Fig. 7B). MARCH1-
deficient DCs expressed CD86 at consistently high levels regard-
less of IL-10 signaling (Fig. 7B). To verify that MARCH1-deficient
DCs are not defective in IL-10signaling per se, we examinedCD80
expression, which is known to be downregulated by IL-10 signal-
lation of CD86. A, Splenic B blasts generated from CD862/2mice were
transduced with two groups of retroviruses. One retroviral group encodes
MARCH1, -2, or -8 (B–D) along with mCherry. The other group encodes
wild-type CD86 or various CD86 mutants along with GFP. Surface levels
of CD86 expression of mCherry+GFP+cells are shown as mean fluores-
cence intensities (MFI). Data are representative of two independent ex-
Lysine 267 is sufficient for the MARCH1-mediated regu-
The Journal of Immunology2969
ing. LPS markedly increased CD80 expression, and the increase
was enhanced by the blockade of IL-10 signaling both in wild-
type and MARCH1-deficient DCs (Fig. 7B). These findings indi-
cate that MARCH1 is specifically required for autocrine IL-10 reg-
ulation of CD86 in DCs. Lastly, we determined whether MARCH1
mediates CD86 ubiquitination in DCs in response to autocrine IL-
MARCH1-deficient DCs. As shown in Fig. 7C, polyubiquitinated
CD86 mostly disappeared in MARCH1-deficient DCs compared
with wild-type DCs (Fig. 6C). However, oligoubiquitinated CD86,
which had approximately two to three ubiquitin monomers at-
tached, was still detected, indicating that those ubiquitinated CD86
species were formed by other ubiquitin ligase(s). This finding
suggests that MARCH1 mainly mediates polyubiquitination of
CD86 and that this polyubiquitination is involved in regulating
CD86 levels in DCs in response to autocrine-IL-10.
Autocrine IL-10–mediated CD86 regulation in DCs requires
To further prove that CD86 ubiquitination is required for CD86
regulation by autocrine IL-10, we retrovirally expressed ubiquiti-
nation-resistant CD86 in DCs and examined how these cells con-
trol CD86 levels in response to autocrine IL-10. As shown in Fig.
8A, DCs expressing wild-type CD86 markedly enhanced CD86
expression when IL-10 signaling was blocked. However, DCs
expressing CD86 (K . R) mutant constantly expressed high CD86
levels regardless of IL-10 signaling (Fig. 8B). Furthermore, this
defective CD86 regulation was completely rescued by the pres-
ence of lysine 267 (Fig. 8C). We also examined the effects of
autocrine IL-10 on total CD86. Expression of wild-type CD86 and
CD86 (RKRRR) mutant increased upon blockade of IL-10 sig-
naling (Fig. 8D, 8F), whereas expression of CD86 (K . R) mutant
did not (Fig. 8E). These findings demonstrate that autocrine IL-
10–mediated regulation of CD86 expression is dependent on its
CD86 ubiquitination plays a significant role in controlling
DC-mediated Ag presentation
Thus far, our findings indicate that CD86 ubiquitination serves as
an important mechanism by which DCs express minimal levels of
CD86 at rest and by which DCs express increased but controlled
levelsofCD86 following LPSstimulation. Therefore, wewished to
see whether CD86 ubiquitination also plays an important role in
controlling T cell-activating abilities of DCs. Through our studies,
we have developed two ways to interfere with CD86 ubiquitination
in DCs. One is to ablate MARCH1 expression, and the other is to
replacewild-type CD86 with ubiquitination-resistant mutant CD86
(K . R). Because the former may abolish the ubiquitination of so
far unknown MARCH1 substrates in addition to abolishing CD86
ubiquitination, we chose the latter to specifically interfere with
CD86 ubiquitination and determine its consequences on DC-
mediated Ag presentation.
First,we determinedwhether resting DCs thatfail toubiquitinate
CD86 canactivate T cellsmore efficiently. We employed an invitro
with peptide-specific naive CD8+T cells. T cell activation was then
determined by measuring the production of IL-2 by ELISA. We
found that DCs expressing CD86 (K . R) mutant activated T cells
to a much greater degree than DCs expressing wild-type CD86
tion and turnover in DCs. A, Flow cytometry of wild-type BMDCs fol-
lowing indicated treatment. B, Transcripts of CD86 in BMDCs following
indicated treatments. Values are relative to actin mRNA, and data are
depicted as fold difference of that observed in untreated BMDCs. Data
represent the mean 6 SEM of three independent experiments. C, Western
blot analysis of CD86 immunoprecipitates and cell lysates generated from
wild-type mice. D, The relative amounts of ubiquitinated CD86 are
expressed as fold difference of that observed in LPS-treated BMDCs. Data
represent individual values of three independent experiments and mean
values. E, BMDCs were first treated as indicated and then surface bio-
tinylated followed by a chase for 16 h at 37˚C. CD86 was immunopreci-
pitated and analyzed as described earlier. F, The relative amounts of
biotinylated CD86 are expressed as a percentage of the total amount of
biotinylated CD86 at t = 0. Data are representative of two independent
LPS-induced autocrine IL-10 facilitates CD86 ubiquitina-
MARCH1 expression. A, Transcripts of MARCH1 in BMDCs following
indicated treatments. Values are relative to actin mRNA, and data are
depicted as fold difference of that observed in untreated BMDCs. Data
represent the mean 6 SEM of three independent experiments. B, Flow
cytometry of wild-type and MARCH12/2BMDCs following indicated
treatment. C, Western blot analysis of CD86 immunoprecipitates and cell
lysates generated from MARCH1-deficient mice. *p , 0.05.
Regulation of CD86 by autocrine IL-10 depends on
2970MARCH1-MEDIATED UBIQUITINATION OF CD86 IN DENDRITIC CELLS
(Fig. 9A). Notably, MHC class I expression was identical in DCs
expressing wild-type or mutant CD86 (Supplemental Fig. 2). This
DCs to regulate their T cell-activating abilities in the resting state.
Next, we determinedwhether activated DCs that fail to ubiquitinate
autocrine IL-10. As shown in Fig. 9B, blockade of IL-10 signaling
by anti–IL-10R Ab markedly increased the ability of DCs to acti-
DCs expressed CD86 (K . R) mutant. In addition, CD86 (K . R)-
expressing DCs activated T cells better than wild-type CD86-
expressing DCs regardless of IL-10 signaling. These findings in-
dicate that CD86 ubiquitination in regulating T cell-activating
abilities of DCs in response to autocrine IL-10.
In conclusion, our study demonstrates that DCs ubiquitinate CD86
and that its ubiquitination plays an important role in regulating
CD86 in DCs both at rest and upon activation in response to
autocrine IL-10. Our study also demonstrates that CD86 ubiq-
uitination plays a significant role in DCs to regulate their T cell-
Ubiquitination has been implicated in CD86 regulation for a de-
cade. Early studies demonstrated that CD86 is downregulated in
human B cell lines expressing MIR2, a viral RING E3 ubiquitin
ligase (19, 20). A subsequent study revealed that MIR2 ubiq-
uitinates CD86 and induces CD86 endocytosis and subsequent
degradation (21). Given the fact that viruses often usurp host
physiology for their advantages, it was postulated that CD86 is
ubiquitinated and thus regulated by mammalian E3 ubiquitin liga-
ses. Soon afterward, the MARCH protein family was identified in
mammalian cells (9, 15). Among nine family members examined,
forced expression of MARCH1, -2, and -8 induced a significant
reduction of surface CD86 in several cell lines (16). Furthermore,
ubiquitinated CD86 was detected in MARCH8-transfected cell
lines (15). More recently, MARCH1-deficient mice were developed
(8). These mice were found to express significantly elevated CD86
levels in DCs, suggesting a role for MARCH1 in regulating CD86
in DCs (7). However, evidence that CD86 is ubiquitinated in a
physiological setting and that this ubiquitination is mediated by
MARCH1 has been missing until now.
Our study demonstrates for the first time, to our knowledge,
that CD86 is ubiquitinated in DCs and that this ubiquitination
is mediated by MARCH1. MARCH1 appears to polyubiquitinate
CD86, which then is required for the regulation of CD86 levels
in DCs. The specific mechanism by which this ubiquitination
controls CD86 expression is not clear. Ubiquitination mediates
proteasomal degradation of proteins misfolded in the endoplas-
mic reticulum (22). Ubiquitination mediates endocytosis and sub-
sequent lysosomal or proteasomal degradation of many trans-
membrane proteins (17, 18, 23). The specific involvement of
endoplasmic reticulum, proteasomes, and lysosomes in ubiquitin-
mediated CD86 regulation remains to be determined.
Our data indicate that additional E3 ligase(s) can ubiquitinate
CD86 in DCs. Appreciable amounts of ubiquitinated CD86 were
detected in MARCH1-deficient DCs, in which the major ubiq-
uitinated species were oligoubiquitinated. The functional role of
this oligoubiquitination is not clear. It may also contribute to
CD86 regulation but to a lesser degree than MARCH1-mediated
polyubiquitination. Oligoubiquitinated cargo is much less recep-
uitinated cargo (24, 25). MARCH2 or MARCH8 might be in-
volved in this MARCH1-independent oligoubiquitination. MARCH2
seems to specifically target lysine 280, whereas MARCH8 does
not require a specific lysine to reduce CD86 levels. Neverthe-
less, neither MARCH2 nor MARCH8 seem to play a critical role
in regulating CD86 in DCs in the resting state. DCs of MARCH2-
and MARCH8-deficient mice expressed CD86 at comparable
levels to wild-type mouse DCs (G. Baravalle, S. Ishido, and
J.-S. Shin, unpublished observations) (26). The contribution of
MARCH2 or MARCH8 for CD86 regulation may only become
significant in the absence of MARCH1 or upon MARCH1 down-
CD86 ubiquitination. Flow cytometry of CD862/2BMDCs transduced
with retrovirus encoding wild-type CD86 (A), CD86 (K . R) (B), or CD86
(RKRRR) (C) and treated as indicated. Western blots of CD862/2BMDCs
expressing wild-type CD86 (D), CD86 (K . R) (E), or CD86 (RKRRR)
(F) treated as indicated.
Control of CD86 expression by autocrine IL-10 requires
DC Ag-presenting functions. In vitro T cell activation assay using im-
mature BMDCs (A) or BMDCs treated as indicated (B). BMDCs cultured
from CD862/2mice were retrovirally transduced to express wild-type
CD86 or CD86 (K . R) mutant. DCs were loaded with increasing con-
centrations of the OVA-specific peptide, SIINFEKL, and cocultured with
naive CD8+T cells isolated from OT-I transgenic mice. Sixteen to 18 h
later, IL-2 in the supernatant was determined by ELISA. Data, expressed as
the mean 6 SEM, are representative of five independent experiments.
CD86 ubiquitination plays a significant role in controlling
The Journal of Immunology2971
We confirmed that MARCH1-deficient DCs expressed signifi-
cantly elevated CD86 levels at rest. This finding indicates an
important role for MARCH1-mediated ubiquitination in DCs to
actively suppress CD86 expression. However, the functional role
for this active CD86 regulation in resting DCs is not clear. DCs
continuously present self-Ags under steady-state conditions (27).
These self-Ags could be recognized by T cells, which then might
lead to the activation of self-reactive T cells. DCs may prevent this
unnecessary and potentially harmful T cell activation by actively
suppressing CD86 expression. Notably, our in vitro Ag pre-
sentation assay indicates that resting DCs can activate Ag-specific
T cells more efficiently when CD86 ubiquitination was limited.
We also found thatDCs continue to ubiquitinate CD86 following
activation by LPS and that this ubiquitination was in part mediated
by autocrine IL-10. LPS-stimulated DCs produce IL-10, which
acts in an autocrine manner, inhibiting the excessive increase of
CD86 in DCs (4). However, the mechanism underlying this
CD86 regulation by autocrine IL-10 has not been defined. Our
study indicates that autocrine IL-10 counteracts the LPS-induced
downregulation of MARCH1 in DCs. This allows DCs to con-
tinuously ubiquitinate CD86 upon activation and helps to regulate
CD86 expression. The mechanism by which IL-10 counteracts the
effects of LPS is not clear. One potential mechanism is that IL-10
may inhibit LPS signaling, which downregulates MARCH1 ex-
pression in DCs (12, 28). Alternatively but not exclusively, IL-10
may upregulate MARCH1 expression independently of LPS sig-
naling. By any means, MARCH1-mediated ubiquitination plays
an essential role for DCs to regulate CD86 in their response to
autocrine IL-10. This finding is consistent with the recent report
that MARCH1 is important for DCs to suppress CD86 expression
in response to recombinant IL-10 (11).
IL-10 produced by LPS-stimulated DCs also exerts a regulatory
effecton Ag-presentingfunctions of DCs, whichis likely toinclude
the regulation of inflammatory cytokine production and the regu-
lation of Ag-presenting molecules and costimulatory molecules
(29, 30). Our study indicates that this regulatory effect of autocrine
IL-10 in MHC class I-mediated DC Ag presentation is largely de-
pendent on CD86 ubiquitination. Especially when DCs presented
high doses of Ags, the regulation of T cell activation by DCs was
critically dependent on CD86 ubiquitination (Supplemental Fig. 3).
role in IL-10–mediated immune regulation, perhaps in the regula-
tion of CD8 T cell immunity. Previous studies have demonstrated
that lymphocytic choriomeningitis virus chronic mouse infection
model involves diminished CTL responses and hyperproduction
of IL-10 by DCs (31, 32). CD86 ubiquitination in DCs may be
a significant contributing factor to this infection.
In summary, our study indicates that CD86 is ubiquitinated
in DCs and that this ubiquitination plays an important role
not only in regulating CD86 expression but also by regulating Ag-
presenting functions of DCs. Our study also demonstrates that
CD86 ubiquitination is one of the important mechanisms under-
lying IL-10–mediated DC regulation. We speculate that CD86
ubiquitination may play additional important functional roles in
the steady state as well as during immune responses, which remain
to be explored.
We thank Bithi Chatterjee for incisive and helpful comments on this article
and the Sandler family for support of this research.
The authors have no financial conflicts of interest.
1. Lenschow, D. J., T. L. Walunas, and J. A. Bluestone. 1996. CD28/B7 system of
T cell costimulation. Annu. Rev. Immunol. 14: 233–258.
2. Greenwald, R. J., G. J. Freeman, and A. H. Sharpe. 2005. The B7 family
revisited. Annu. Rev. Immunol. 23: 515–548.
3. Mellman, I., and R. M. Steinman. 2001. Dendritic cells: specialized and regu-
lated antigen processing machines. Cell 106: 255–258.
4. Haase, C., T. N. Jørgensen, and B. K. Michelsen. 2002. Both exogenous and
endogenous interleukin-10 affects the maturation of bone-marrow-derived den-
dritic cells in vitro and strongly influences T-cell priming in vivo. Immunology
5. Corinti, S., C. Albanesi, A. la Sala, S. Pastore, and G. Girolomoni. 2001. Reg-
ulatory activity of autocrine IL-10 on dendritic cell functions. J. Immunol. 166:
6. De Smedt, T., M. Van Mechelen, G. De Becker, J. Urbain, O. Leo, and M. Moser.
1997. Effect of interleukin-10 on dendritic cell maturation and function.
Eur. J. Immunol. 27: 1229–1235.
7. Ohmura-Hoshino, M., Y. Matsuki, M. Mito-Yoshida, E. Goto, M. Aoki-
Kawasumi, M. Nakayama, O. Ohara, and S. Ishido. 2009. Cutting edge:
requirement of MARCH-I-mediated MHC II ubiquitination for the maintenance
of conventional dendritic cells. J. Immunol. 183: 6893–6897.
8. Matsuki, Y., M. Ohmura-Hoshino, E. Goto, M. Aoki, M. Mito-Yoshida,
M. Uematsu, T. Hasegawa, H. Koseki, O. Ohara, M. Nakayama, et al. 2007.
Novel regulation of MHC class II function in B cells. EMBO J. 26: 846–854.
9. Ohmura-Hoshino, M., E. Goto, Y. Matsuki, M. Aoki, M. Mito, M. Uematsu,
H. Hotta, and S. Ishido. 2006. A novel family of membrane-bound E3 ubiquitin
ligases. J. Biochem. 140: 147v154.
10. Thibodeau, J., M. C. Bourgeois-Daigneault, G. Huppe ´, J. Tremblay, A. Aumont,
M. Houde, E. Bartee, A. Brunet, M. E. Gauvreau, A. de Gassart, et al. 2008.
Interleukin-10-induced MARCH1 mediates intracellular sequestration of MHC
class II in monocytes. Eur. J. Immunol. 38: 1225–1230.
11. Tze, L. E., K. Horikawa, H. Domaschenz, D. R. Howard, C. M. Roots, R. J. Rigby,
D. A. Way, M. Ohmura-Hoshino, S. Ishido, C. E. Andoniou, et al. 2011. CD83
increases MHC II and CD86 on dendritic cells by opposing IL-10-driven
MARCH1-mediated ubiquitination and degradation. J. Exp. Med. 208: 149–165.
12. De Gassart, A., V. Camosseto, J. Thibodeau, M. Ceppi, N. Catalan, P. Pierre, and
E. Gatti. 2008. MHC class II stabilization at the surface of human dendritic cells
is the result of maturation-dependent MARCH I down-regulation. Proc. Natl.
Acad. Sci. USA 105: 3491–3496.
13. Shin, J. S., M. Ebersold, M. Pypaert, L. Delamarre, A. Hartley, and I. Mellman.
2006. Surface expression of MHC class II in dendritic cells is controlled by
regulated ubiquitination. Nature 444: 115–118.
14. Szymczak, A. L., C. J. Workman, Y. Wang, K. M. Vignali, S. Dilioglou,
E. F. Vanin, and D. A. Vignali. 2004. Correction of multi-gene deficiency in vivo
using a single ‘self-cleaving’ 2A peptide-based retroviral vector. Nat. Bio-
technol. 22: 589–594.
15. Goto, E., S. Ishido, Y. Sato, S. Ohgimoto, K. Ohgimoto, M. Nagano-Fujii, and
H. Hotta. 2003. c-MIR, a human E3 ubiquitin ligase, is a functional homolog of
herpesvirus proteins MIR1 and MIR2 and has similar activity. J. Biol. Chem.
16. Bartee, E., M. Mansouri, B. T. Hovey Nerenberg, K. Gouveia, and K. Fru ¨h. 2004.
Downregulation of major histocompatibility complex class I by human
ubiquitin ligases related to viral immune evasion proteins. J. Virol. 78:
17. Raiborg, C., T. E. Rusten, and H. Stenmark. 2003. Protein sorting into multi-
vesicular endosomes. Curr. Opin. Cell Biol. 15: 446–455.
18. Hicke, L., and R. Dunn. 2003. Regulation of membrane protein transport by
ubiquitin and ubiquitin-binding proteins. Annu. Rev. Cell Dev. Biol. 19: 141–172.
19. Ishido, S., J. K. Choi, B. S. Lee, C. Wang, M. DeMaria, R. P. Johnson,
G. B. Cohen, and J. U. Jung. 2000. Inhibition of natural killer cell-mediated
cytotoxicity by Kaposi’s sarcoma-associated herpesvirus K5 protein. Immunity
20. Coscoy, L., and D. Ganem. 2001. A viral protein that selectively downregulates
ICAM-1 and B7-2 and modulates T cell costimulation. J. Clin. Invest. 107:
21. Coscoy, L., D. J. Sanchez, and D. Ganem. 2001. A novel class of herpesvirus-
encoded membrane-bound E3 ubiquitin ligases regulates endocytosis of proteins
involved in immune recognition. J. Cell Biol. 155: 1265–1273.
22. Vembar, S. S., and J. L. Brodsky. 2008. One step at a time: endoplasmic
reticulum-associated degradation. Nat. Rev. Mol. Cell Biol. 9: 944–957.
23. Hershko, A., and A. Ciechanover. 1998. The ubiquitin system. Annu. Rev.
Biochem. 67: 425–479.
24. Thrower, J. S., L. Hoffman, M. Rechsteiner, and C. M. Pickart. 2000. Recog-
nition of the polyubiquitin proteolytic signal. EMBO J. 19: 94–102.
25. Barriere, H., C. Nemes, K. Du, and G. L. Lukacs. 2007. Plasticity of poly-
ubiquitin recognition as lysosomal targeting signals by the endosomal sorting
machinery. Mol. Biol. Cell 18: 3952–3965.
26. Ishido, S., Y. Matsuki, E. Goto, M. Kajikawa, and M. Ohmura-Hoshino. 2010.
MARCH-I: a new regulator of dendritic cell function. Mol. Cells 29: 229–232.
27. Wilson, N. S., D. El-Sukkari, and J. A. Villadangos. 2004. Dendritic cells con-
stitutively present self antigens in their immature state in vivo and regulate
antigen presentation by controlling the rates of MHC class II synthesis and
endocytosis. Blood 103: 2187–2195.
28. Young, L. J., N. S. Wilson, P. Schnorrer, A. Proietto, T. ten Broeke, Y. Matsuki,
A. M. Mount, G. T. Belz, M. O’Keeffe, M. Ohmura-Hoshino, et al. 2008. Dif-
ferential MHC class II synthesis and ubiquitination confers distinct antigen-
2972 MARCH1-MEDIATED UBIQUITINATION OF CD86 IN DENDRITIC CELLS
presenting properties on conventional and plasmacytoid dendritic cells. Download full-text
Nat. Immunol. 9: 1244–1252.
29. Mosser, D. M., and X. Zhang. 2008. Interleukin-10: new perspectives on an old
cytokine. Immunol. Rev. 226: 205–218.
30. Chang, J., S. L. Kunkel, and C. H. Chang. 2009. Negative regulation of MyD88-
dependent signaling by IL-10 in dendritic cells. Proc. Natl. Acad. Sci. USA 106:
31. Ejrnaes, M., C. M. Filippi, M. M. Martinic, E. M. Ling, L. M. Togher,
S. Crotty, and M. G. von Herrath. 2006. Resolution of a chronic viral in-
fection after interleukin-10 receptor blockade. J. Exp. Med. 203: 2461–
32. Brooks, D. G., M. J. Trifilo, K. H. Edelmann, L. Teyton, D. B. McGavern, and
M. B. Oldstone. 2006. Interleukin-10 determines viral clearance or persistence
in vivo. Nat. Med. 12: 1301–1309.
The Journal of Immunology2973