Microsomal triglyceride transfer protein lipidation and control of CD1d on antigen-presenting cells.
ABSTRACT Microsomal triglyceride transfer protein (MTP), an endoplasmic reticulum (ER) chaperone that loads lipids onto apolipoprotein B, also regulates CD1d presentation of glycolipid antigens in the liver and intestine. We show MTP RNA and protein in antigen-presenting cells (APCs) by reverse transcription-polymerase chain reaction and by immunoblotting of mouse liver mononuclear cells and mouse and human B cell lines. Functional MTP, demonstrated by specific triglyceride transfer activity, is present in both mouse splenocytes and a CD1d-positive mouse NKT hybridoma. In a novel in vitro transfer assay, purified MTP directly transfers phospholipids, but not triglycerides, to recombinant CD1d. Chemical inhibition of MTP lipid transfer does not affect major histocompatibility complex class II presentation of ovalbumin, but considerably reduces CD1d-mediated presentation of alpha-galactosylceramide (alpha-galcer) and endogenous antigens in mouse splenic and bone marrow-derived dendritic cells (DCs), as well as in human APC lines and monocyte-derived DCs. Silencing MTP expression in the human monocyte line U937 affects CD1d function, as shown by diminished presentation of alpha-galcer. We propose that MTP acts upstream of the saposins and functions as an ER chaperone by loading endogenous lipids onto nascent CD1d. Furthermore, our studies suggest that a small molecule inhibitor could be used to modulate the activity of NKT cells.
- [show abstract] [hide abstract]
ABSTRACT: Major histocompatibility complex (MHC) class I and class II molecules bind immunogenic peptides and present them to lymphocytes bearing the alpha beta T-cell antigen receptor (TCR). An analogous antigen-presenting function also has been proposed for the non-MHC-encoded CD1 molecules, a family of non-polymorphic, beta 2-microglobulin-associated glycoproteins expressed on most professional antigen-presenting cells. In support of this hypothesis, CD1 molecules are recognized by selected CD4-CD8- alpha beta or gamma delta TCR+ T-cell clones, and we have recently shown that CD1 molecules restrict the recognition of foreign microbial antigens by alpha beta TCR+ T cells. But the substantial structural divergence of CD1 from MHC class I and class II molecules, raises the possibility that the antigens presented by the CD1 system may differ fundamentally from those presented by MHC-encoded molecules. Here we report that a purified CD1b-restricted antigen of Mycobacterium tuberculosis presented to alpha beta TCR+ T cells is mycolic acid, a family of alpha-branched, beta-hydroxy, long-chain fatty acids found in mycobacteria. This example of non-protein microbial antigen recognition suggests that alpha beta TCR+ T cells recognize a broader range of antigens than previously appreciated and that at least one member of the CD1 family has evolved the ability to present lipid antigens.Nature 01/1994; 372:691-4. · 38.60 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: CD1d-restricted natural killer T (NKT) cells are innate-like lymphocytes that express a conserved T-cell receptor and contribute to host defence against various microbial pathogens. However, their target lipid antigens have remained elusive. Here we report evidence for microbial, antigen-specific activation of NKT cells against Gram-negative, lipopolysaccharide (LPS)-negative alpha-Proteobacteria such as Ehrlichia muris and Sphingomonas capsulata. We have identified glycosylceramides from the cell wall of Sphingomonas that serve as direct targets for mouse and human NKT cells, controlling both septic shock reaction and bacterial clearance in infected mice. In contrast, Gram-negative, LPS-positive Salmonella typhimurium activates NKT cells through the recognition of an endogenous lysosomal glycosphingolipid, iGb3, presented by LPS-activated dendritic cells. These findings identify two novel antigenic targets of NKT cells in antimicrobial defence, and show that glycosylceramides are an alternative to LPS for innate recognition of the Gram-negative, LPS-negative bacterial cell wall.Nature 04/2005; 434(7032):525-9. · 38.60 Impact Factor
- Science. 01/2004;
The Journal of Experimental Medicine
JEM © The Rockefeller University Press
Vol. 202, No. 4, August 15, 2005 529–539
Microsomal triglyceride transfer protein
lipidation and control of CD1d on
Stephanie K. Dougan,
and Richard S. Blumberg
M. Mahmood Hussain,
Gastroenterology Division, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School,
Boston, MA 02115
Program in Immunology, Division of Medical Sciences and
Harvard Medical School, Boston, MA 02215
Department of Anatomy and Cell Biology, State University of New York Downstate Medical Center, Brooklyn, NY 11203
Division of Developmental Immunology, La Jolla Institute for Allergy and Immunology, San Diego, CA 92121
Division of Hematology, Beth Israel Deaconess Medical Center,
Microsomal triglyceride transfer protein (MTP), an endoplasmic reticulum (ER) chaperone
that loads lipids onto apolipoprotein B, also regulates CD1d presentation of glycolipid
antigens in the liver and intestine. We show MTP RNA and protein in antigen-presenting
cells (APCs) by reverse transcription–polymerase chain reaction and by immunoblotting of
mouse liver mononuclear cells and mouse and human B cell lines. Functional MTP,
demonstrated by specific triglyceride transfer activity, is present in both mouse splenocytes
and a CD1d-positive mouse NKT hybridoma. In a novel in vitro transfer assay, purified MTP
directly transfers phospholipids, but not triglycerides, to recombinant CD1d. Chemical
inhibition of MTP lipid transfer does not affect major histocompatibility complex class II
presentation of ovalbumin, but considerably reduces CD1d-mediated presentation of
-galcer) and endogenous antigens in mouse splenic and bone
marrow–derived dendritic cells (DCs), as well as in human APC lines and monocyte-derived
DCs. Silencing MTP expression in the human monocyte line U937 affects CD1d function, as
shown by diminished presentation of
-galcer. We propose that MTP acts upstream of the
saposins and functions as an ER chaperone by loading endogenous lipids onto nascent CD1d.
Furthermore, our studies suggest that a small molecule inhibitor could be used to modulate
the activity of NKT cells.
Structurally homologous to MHC class I, the
CD1 family of glycoproteins has evolved to
present lipid antigens (1). The human group I
CD1a, b, and c proteins present mycobacterial
lipids and lipopeptides, but can also present
host lipids to autoreactive CD1-restricted T
cells (2–6). The intracellular localization of
CD1 proteins is controlled by dileucine and
tyrosine sorting motifs in their cytoplasmic
tails (7, 8). The type of lipid each CD1 family
member presents reflects both the shape of the
CD1 hydrophobic antigen-binding pocket and
the endosomal compartments through which
the CD1 proteins traffic (9). Group II CD1d,
which is found in humans and is the only CD1
protein in rodents, presents glycolipid antigens
to NKT cells, which are defined as cells express-
ing NK surface markers and CD1d-restricted T
cell receptors (1). Marine sponge–derived
model antigen for NKT cells (10), and phos-
phatidyl inositol mannoside from mycobacteria,
antigens, and sphingolipids
have been shown to activate
subsets of NKT cells in a CD1d-restricted
manner (11–14). CD1d in vivo also presents
endogenous glycolipid antigens (15, 16). Several
host lipids have been proposed to associate
with CD1d and activate subsets of NKT cells,
including phosphatidyl inositol, phosphatidyl
ethanolamine (PE), and isoglobotrihexosylcera-
-galcer) is an exogenous
Richard S. Blumberg:
microglobulin; BMDC, BM-
derived DC; ER, endoplasmic
reticulum; IEC, intestinal epithe-
lial cell; LMNC, liver mononu-
clear cell; MFI, mean fluorescence
intensity; moDC, monocyte-
derived DC; MTP, microsomal
triglyceride transfer protein;
azol-4-yl; PDI, protein disulfide
isomerase; PE, phosphatidyl
ethanolamine; pIpC, poly-
si, small interfering.
The online version of this article contains supplemental material.
MTP REGULATES CD1
| Dougan et al.
mide, which is required for NKT cell development (17–20).
Activation of autoreactive NKT cells by CD1d-presenting
host lipids can be beneficial during bacterial and viral infec-
tions, some antitumor responses, and regulation of autoim-
mune diseases such as diabetes (21–24). However, improper
activation of NKT cells can lead to inflammatory bowel dis-
ease, asthma, and atherosclerosis (25–27).
The endoplasmic reticulum (ER) chaperones calnexin,
calreticulin, and ERp57 associate with nascent CD1d and as-
sist in folding and disulfide bond formation (28). Unlike
MHC class I, which associates with
early in biogenesis, CD1d acquires
the ER, and
2m is not essential for CD1d cell surface ex-
pression (28, 29). Phospholipids bind to the hydrophobic
pocket of nascent CD1d and likely enable proper folding in
a manner analogous to that of peptide in MHC class I assem-
bly (15). NKT cell clones that recognize phosphatidyl inosi-
tol and PE have been characterized but likely represent a mi-
nority of NKT cells in vivo (15, 19). The ER phospholipids
bound to CD1d may be replaced when CD1d recycles into
endosomal compartments, and recent work on lysosomal
lipid transfer proteins has shown that a family of lipid transfer
2m just before exiting
proteins, including the saposins and GM2 activator, is capa-
ble of exchanging or editing the lipid cargo of CD1d (30,
31). In mice, saposin B can load isoglobotrihexosylceramide
onto CD1d, which then activates invariant V
(20). The relative importance of ER lipids versus endoso-
mal-derived lipids is unclear. Tail-deleted forms of CD1d
that fail to traffic to endosomes activate V
NKT cells but cannot present antigen to invariant NKT cells
(32). Furthermore, mice that express only the tail-deleted
form of CD1d support thymic development of diverse but
not invariant NKT cells, indicating a distinction in the host
lipids recognized by these two NKT cell populations (32).
Microsomal triglyceride transfer protein (MTP) is pre-
dominately found in the ER of hepatocytes and intestinal epi-
thelial cells (IECs), where it loads triglycerides, cholesterol es-
ters, and phospholipids onto apolipoprotein B (apoB) (33–35).
In the absence of MTP-mediated lipid transfer, apoB is de-
graded and very low density lipoproteins or chylomicrons are
not secreted from the liver or intestines, respectively (36–39).
Humans with mutations in the gene encoding MTP develop
abetalipoproteinemia, a disorder characterized by low serum
lipoproteins and severe lipophilic vitamin deficiencies (40).
14 NKT cells
1 ?g RNA from the indicated tissues. MTP transcripts were amplified by
RT-PCR, and the volume of sample loaded was normalized to ?-actin
transcript levels. MTP? DCs were obtained from the liver of a pIpC-injected
MTPmx1 mouse. (B) Protein was harvested from LMNCs isolated from
MTPMx1 or C57BL/6 mice either injected or not injected with pIpC and B
cell lines C1R and 721 transfected with human and mouse CD1d, respec-
MTP is present in APCs. (A) cDNA was synthesized from
tively. 25 ?g of cell lysate was loaded per lane, separated by reducing
SDS-PAGE, and immunoblotted with polyclonal antisera to MTP:PDI.
(C) Triglyceride transfer activity in 100 ?g splenocyte lysate (?), 50 ?g
splenocyte lysate (?), 100 ?g splenocyte lysate with BMS197636 (?), and
50 ?g splenocyte lysate with BMS197636 (?) is shown. Assays were in
triplicate with SDs less than 10% of the percent transfer at 3 h. Each well
contained 100 picomoles of NBD-labeled triglyceride.
JEM VOL. 202, August 15, 2005
Recent work from our laboratory has shown the impor-
tance of MTP in CD1d antigen presentation by hepatocytes
and IECs (41). We observed that MTP associates with CD1d
in hepatocytes and that CD1d surface expression on hepato-
cytes was reduced in the absence of MTP. MTP silencing or
gene deletion in hepatocytes or IECs resulted in diminished
antigen presentation by CD1d in hepatocytes and IECs and
protection from NKT cell–mediated hepatitis and colitis.
However, whether MTP affects CD1d presentation in a
broad range of tissues and cell types, and not in APCs, or
whether MTP can mediate direct lipid transfer to CD1d was
unclear. CD1d is expressed on DCs, monocytes, macro-
phages, cortical thymocytes, and most peripheral lymphocytes
(42). We now show that MTP expression is also present in
mouse and human professional APCs, that MTP lipid trans-
fer activity is present in mouse splenocytes, and that MTP
can directly transfer phospholipid to recombinant CD1d in
vitro. We further show, by gene silencing and chemical in-
hibition, that the lack of MTP in such professional APCs re-
sults in a selective defect in CD1d presentation of exogenous
and endogenous CD1d-restricted antigens. Importantly,
MHC class II–restricted presentation of ovalbumin is unaf-
fected by the loss of MTP, indicating that MTP plays a spe-
cific role in CD1d presentation. We therefore conclude that
MTP is a broadly expressed ER chaperone that can lipidate
CD1d and is involved in CD1d-restricted antigen presenta-
tion in professional APCs.
MTP is present and functional in APCs
Because MTP plays a role in CD1d antigen presentation by
hepatocytes and IECs, we hypothesized that MTP may serve
as a chaperone for CD1d in all CD1d-positive cell types.
MTP expression has never been detected in professional
APCs, and most research in the field has focused on the func-
tions of MTP in the liver and intestine (33). We therefore
transcript expression in a variety of potential
APC types (Fig. 1 A; not depicted).
mouse primary tissues (liver, heart, lung, ovaries, peritoneal
exudate cells, small intestine, colon, spleen, thymus, lymph
node, B cells, liver DCs, and BM-derived DCs [BMDCs]);
mouse thymus (RMAS), APC (RAW), and NKT cell
(DN32) lines; and in human B, T, and monocyte cell lines
(C1Rd, Jurkat, and U937).
lated from MTP gene–deleted mice, but was present in wild-
type mice at low abundance in every other immunologic and
nonimmunologic tissue examined.
testine was notably increased relative to APCs, as would be
expected from a tissue actively secreting lipoproteins.
To verify the presence of MTP protein in APCs, we iso-
lated liver mononuclear cells (LMNCs) from MTPMx1
mice that either had or had not been injected with poly-
inositic:poly-cytodylic acid (pIpC) to induce MTP gene de-
letion. The LMNCs were then lysed, and proteins were
separated by SDS-PAGE. Immunoblotting was performed
was detected in
was not present in DCs iso-
expression in the in-
using a polyclonal antisera raised against recombinant human
MTP:protein disulfide isomerase (PDI), which has been
demonstrated to recognize human and mouse MTP large
subunits and PDI. As shown in Fig. 1 B, we were able to de-
tect a 97-kD protein in the nondeleted mice that was not
present in the MTP gene–deleted mice. To show that the
lack of MTP protein was not caused by pIpC injection, we
also analyzed LMNC lysates from wild-type mice treated or
not treated with pIpC. In addition, we were able to detect
evidence of MTP protein in lysates of the human B cell line
C1Rd and the mouse B cell line 721d.
To demonstrate that MTP is functional in APCs, we used
a previously described triglyceride transfer assay (43). In this
assay, MTP function is assessed by studying the transfer of flu-
orescently labeled triglycerides embedded in donor phospho-
lipid vesicles to acceptor vesicles. During the transfer, the
amount of triglyceride associated with MTP can be measured
as an increase in fluorescence over time. Splenocyte lysates
contained measurable triglyceride transfer activity, which
could be abrogated by the addition of 200 nM BMS197636,
a known chemical inhibitor of MTP lipid transfer (Fig. 1 C;
references 44, 45). Specific activity in the splenocyte lysates
was 0.04% triglyceride being transferred per microgram of ly-
sate protein per hour. Although the MTP-specific activity in
primary splenocytes was low, the activity was abolished by
the addition of BMS197636, which is consistent with MTP
being present in these cells. The splenocytes used were an
unfractionated population containing CD1d-positive and
-negative cells. When we examined a homogenous popula-
tion of CD1d-positive cells, the mouse NKT cell line DN32,
a specific activity of 0.22 was observed. Athar et al. calculated
MTP-specific activity from whole cell lysates of the hepato-
cyte cell line HepG2 to be 0.982, and, in our own assays, pu-
rified rat liver microsomes exhibited a specific activity of 4.78
(43). Thus, MTP lipid transfer activity is present in APCs but
less abundant than that observed in tissues such as the liver,
which actively secrete lipoproteins.
MTP can transfer a lipid to recombinant CD1d in vitro
MTP could directly transfer lipids to CD1d, or it could indi-
rectly influence CD1d presentation, such as through a lipo-
protein intermediate. To test for direct lipid transfer from
MTP to CD1d, we used a reductionist in vitro approach us-
ing PE as a model lipid antigen. PE has been shown to bind
to CD1d and can be recognized by several invariant NKT
cell hybridomas (15, 19). PE is also a known substrate for
MTP (34, 35).
2m-associated CD1d was coated onto a
96-well plate and incubated with liposomes containing a 6:1
ratio of unlabeled PE/7-nitro-2,1,3-benzoxadiazol-4-yl
(NBD)–labeled PE. The fluorescent NBD label was conju-
gated to the PE headgroup in order to avoid steric interfer-
ence between PE acyl tails and CD1d hydrophobic pockets.
MTP purified from rat liver homogenates was added to the
CD1d- or BSA-coated wells at a 1:10 molar ratio of MTP/
MTP REGULATES CD1
| Dougan et al.
CD1d, incubated at 37
Lipids bound to CD1d were eluted with isopropanol, trans-
ferred to a black microtiter plate, and the level of fluores-
cence was determined. As shown in Fig. 2 A, PE was trans-
ferred to CD1d in the presence of MTP, but not to BSA, as
seen by a threefold increase in fluorescence above back-
ground. To show that MTP transfer to CD1d is specific for
phospholipid, the same assay was performed using NBD-
labeled triglyceride vesicles. No transfer of fluorescence was
observed under these conditions, indicating that, as ex-
pected, phospholipid but not triglyceride binds to CD1d
and, furthermore, that MTP itself is not remaining in the
wells during the elution phase of the assay. As can be seen in
Fig. 2 B, the quantity of fluorescent PE transferred to CD1d
increased with increasing concentrations of purified MTP.
C for 2 h, and washed with PBS.
In vitro inhibition of MTP reduces CD1d-mediated antigen
presentation of mouse DCs
BMS212122 is a specific chemical inhibitor of MTP-medi-
ated lipid transfer that functions in vitro at concentrations as
low as 0.03 nM and in vivo in primates (46). CD1d presen-
tation in the mouse epithelial cell line, MODE-K, is regu-
lated by MTP (41). As shown in Fig. 3 A, incubation of
MODE-K cells with 13
M BMS212122 decreased the pre-
-galcer to DN32 cells. These experiments in-
dicate that the MTP inhibitor BMS212122 blocks CD1d
function, similar to its previously established role in inhibit-
ing apoB secretion (46).
We next examined the role of MTP in CD1d presenta-
tion by primary mouse APCs. Splenocytes were isolated
from wild-type mice and assayed for their ability to present
an exogenous antigen,
-galcer, to DN32 NKT cells and
their ability to stimulate the autoreactive NKT cell line
24.8. When splenocytes were incubated with BMS212122
for 24 h before the addition of NKT cells, their ability
to present exogenous
-galcer and endogenous CD1d-
restricted antigens was reduced (Fig. 3 B). BMS212122 ex-
hibited no effect, however, on splenocyte presentation of
ovalbumin to the MHC class II–restricted T cell line, OT-
II (Fig. 3 B). DCs were then isolated from wild-type sple-
nocytes by positive selection on CD11c magnetic beads and
cultured with DMSO, BMS212122, or 9-fluorenyl carbox-
ylic acid, a negative control compound of similar structure
to BMS212122, but with no MTP inhibitory activity at
concentrations as high as 30
transfer assay (Harrity, T.W., personal communication).
The effect of MTP inhibition on presentation of an endog-
enous ligand to the 24.8 NKT cells was highly pronounced
with BMS212122, but not 9-fluorenyl carboxylic acid, in
the CD11c-positive DC population (Fig. 3 C). CD11c-pos-
itive cells were also pulsed with titrated amounts of
cer, washed, and co-cultured with DN32 iNKT cells. At
every concentration of
-galcer, the MTP inhibitor–treated
cells showed reduced ability to activate iNKT cells com-
pared with vehicle- or control compound–treated splenic
DCs (Fig. 3 D). BMDCs cultured with BMS212122 or
DMSO were also pulsed with titrated amounts of
up to 1000 ng/ml, with a similar inhibitory effect of
BMS212122 observed at all concentrations of
(unpublished data). CD11c-positive DCs cultured with
BMS212122 or vehicle were cultured with ovalbumin and
CD4 T cells isolated from OT-II transgenic mice (Fig. 3 E).
MTP inhibition showed no effect on MHC class II presen-
tation of ovalbumin implying a specific role for MTP in the
BMDCs were cultured with BMS212122 or vehicle
control after 6 d of differentiation in the presence of GM-
CSF. On days 10–13, BMDCs were collected and co-cul-
tured with NKT cells. BMS212122 had a significant effect
on both the presentation of an endogenous ligand to autore-
active 24.8 cells (P
0.01; Fig. 4 A) and on the presenta-
-galcer to DN32 cells (P
of BMDCs were CD11c-positive on day 13, and viability
was unaffected by MTP inhibition. Surface expression of
CD11b, CD11c, CD40, and CD80 was unaffected by
M in an in vitro triglyceride
0.05; Fig. 4 B).
vitro. (A) 2 ?g BSA or recombinant CD1d was coated onto microtiter wells
and incubated with either lipid vesicles alone or lipid vesicles plus 0.5 ?g
purified MTP. Lipid vesicles contained either PE:NBD or TG:NBD. Donor
PE vesicles were composed of 450 nmol unlabeled PE and 14 nmol PE:NBD
per milliliter. Results are representative of six independent experiments.
*, P ? 0.01. (B) 2 ?g of recombinant CD1d was coated onto microtiter wells
and incubated with PE:NBD vesicles and increasing concentrations of purified
MTP. *, P ? 0.005; **, P ? 0.001 compared with 0 ?g MTP. Values are ?SD.
Purified MTP transfers a lipid to recombinant CD1d in
JEM VOL. 202, August 15, 2005
BMS212122, indicating that no deleterious effects on the
maturation of BMDCs had occurred (Fig. S1, available at
In contrast, a decrease in CD1d surface expression on
BMDCs was observed over time in the presence of
BMS212122. One day after the addition of BMS212122,
CD1d expression on BMDCs was 86% of the level observed
in vehicle-treated controls, and after 4 d of BMS212122
treatment, CD1d expression was decreased to 66% of the
vehicle controls. In a total of three independent experiments
on BMDCs, extended culture with BMS212122 reduced
surface CD1d expression by an average of 68
negative control compound 9-fluorenyl carboxylic acid had
no effect on CD1d surface expression (Fig. 4 C). The MHC
II pathway was intact in BMS212122-treated cells; like
CD11c-positive splenic DCs, BMDCs exhibited efficient
presentation of ovalbumin to CD4
ovalbumin-specific T cells (Fig. 4 D). Thus, MTP inhibi-
tion results in a selective defect in CD1d-mediated antigen
, MHC class II–restricted,
MTP is critical for CD1d-restricted antigen presentation
in human cells
MTP is highly conserved among mammalian species with an
86% identity between mouse and human MTP at the amino
acid level. We therefore hypothesized that MTP inhibition in
human APCs would affect CD1d antigen presentation. To
test this, MTP expression was silenced in the human mono-
cyte cell line U937. U937 cells express MTP (Fig. 1 A) and
low levels of surface CD1d but are highly potent APCs for
CD1d-restricted NKT cells (10). U937 cells that were trans-
fected with human
-specific small interfering (si) RNA
oligomers displayed a 48% reduction in
compared with cells transfected with irrelevant oligomers
(mock) after 48 h, as shown in Fig. 5 A. The silenced and
mock-silenced U937 cells were incubated with
3 h, washed, and co-cultured with the iNKT cell line DN32.
NKT cell activation was measured by IL-2 production, and,
as shown in Fig. 5 B, the
-silenced U937 cells exhibited a
0.001) reduction in their ability to present the
model CD1d-restricted exogenous antigen. A similar reduc-
defect in CD1d antigen presentation. (A) MODE-K cells cultured in the
presence of BMS212122 (MTPi) or vehicle were incubated with ?-galcer
for 3 h, washed, and co-cultured with DN32 cells. (B) Splenocytes were
isolated from wild-type mice, incubated for 24 h with BMS212122, and
washed. The exogenous antigens ovalbumin or ?-galcer were added as
indicated. Splenocytes were then co-cultured with autoreactive 24.8 NKT
cells (left), DN32 NKT cells (center), or CD4? cells from an OT-II transgenic
mouse (right). *, P ? 0.05. Results are representative of two independent
experiments. (C) CD11c? splenocytes were incubated for 24 h with
Chemical inhibition of mouse MTP causes a selective
BMS212122, DMSO, or 9-fluorenyl carboxylic acid, washed, and co-cultured
with 50,000 24.8 NKT cells per well. *, P ? 0.005; **, P ? 0.05 compared
with DMSO values. Results are representative of five independent experi-
ments. (D) CD11c? splenocytes were incubated for 4 d with BMS212122
or DMSO, pulsed with the indicated concentrations of ?-galcer, washed,
and co-cultured with DN32 NKT cells (E/T ? 100,000:30,000). *, P ? 0.05.
(E) 30,000 CD11c? splenocytes incubated for 24 h with BMS212122 or
DMSO were washed and co-cultured with 100 ?g/ml ovalbumin and
100,000 CD4? cells from an OT-II transgenic mouse. Values are ?SD.
MTP REGULATES CD1
D IN APCS | Dougan et al.
tion in IL-2 release from DN32 cells was observed when
U937 cells were cultured with ?-galcer and BMS212122, as
compared with the vehicle control (Fig. 5 C).
C1Rd is a human B cell line transfected with human
CD1d (7). C1Rd cells were cultured with MTP inhibitor or
vehicle for 4 d, pulsed with ?-galcer, fixed with glutaralde-
hyde, and washed before co-culture with DN32 NKT cells.
Incubation of C1Rd cells with BMS212122, but not the ve-
hicle, inhibited the presentation of ?-galcer, as determined
by a reduction in IL-2 secretion (Fig. 5 D). To exclude non-
specific effects of BMS212122, C1Rd cells were incubated
with a second MTP chemical inhibitor, BMS197636, which
resulted in dose-dependent inhibition of CD1d-mediated
activation of a human peripheral blood–derived iNKT cell
line in the presence of a limiting concentration of PMA (Fig.
5 E). Thus, MTP inhibition in C1Rd cells resulted in a di-
minished ability to present CD1d-restricted antigens to
mouse and human iNKT cells.
To further determine the effect of MTP inhibition in pri-
mary human APCs, we differentiated monocyte-derived
DCs (moDC) from CD14-positive peripheral blood mono-
nuclear cells obtained from healthy volunteers. Monocytes
were selected on CD14 magnetic beads and cultured in me-
dia supplemented with autologous plasma, GM-CSF, IL-4,
and either BMS212122 or vehicle. ?95% of cells expressed
selective defect in CD1d antigen presentation. (A) BMDCs cultured for
6 d with BMS212122 were washed and co-cultured with 50,000 24.8 NKT
cells per well. *, P ? 0.01. Results are representative of three experiments.
(B) BMDCs cultured for 4 d with BMS212122 were pulsed with ?-galcer,
washed, and co-cultured with 100,000 DN32 NKT cells. *, P ? 0.05. Results
are representative of three experiments. (C) BMDCs cultured for 4 d with
BMS212122 or 9-fluorenyl carboxylic acid were stained with the FITC-
conjugated mAb 1B1 and analyzed by flow cytometry. The dashed line
represents the isotype control. Results are representative of three indepen-
dent experiments. (D) 50,000 BMDCs incubated for 3 d with BMS212122
or DMSO were washed and co-cultured with 100 ?g/ml ovalbumin and
100,000 CD4? cells from an OT-II transgenic mouse. Values are ?SD.
Chemical inhibition of MTP in mouse BMDCs causes a
tation. (A) U937 cells were treated with irrelevant or MTP-specific siRNA
oligomers. RNA was isolated 48 h after silencing, and transcript levels
of mtp and ?-actin were determined by RT-PCR. (B) Silenced and
mock-silenced U937 cells were incubated with ?-galcer for 3 h, washed,
and co-cultured with DN32 cells (E/T ? 1:1). *, P ? 0.001. Results are
representative of two independent experiments. (C) U937 cells cultured
with BMS212122 (MTPi) or vehicle for 3 d were incubated with ?-galcer
for 3 h, washed, and co-cultured with DN32 cells (E/T ? 1:1). *, P ? 0.001.
(D) C1Rd cells cultured with BMS212122 (MTPi) or vehicle for 4 d were
incubated with ?-galcer for 4 h, fixed with 0.05% glutaraldehyde for 30 s,
washed, and co-cultured with DN32 cells (E/T ? 2:1). *, P ? 0.02. Results
are representative of three independent experiments. (E) C1Rd cells cultured
with BMS197636 or vehicle were incubated with NKT cell lines derived
from human peripheral blood in the presence of 1 ng/ml PMA. NKT cells
co-cultured with C1R mock transfected cells in the same assay yielded
33.6 ? 9.8 pg/ml IFN-?. (F) Human monocyte-derived DCs were stained
with 42.1 (CD1d), anti–HLA-A,B,C (MHC class I), or isotype control antibodies
(dashed lines) after 6 d of differentiation in the presence of BMS212122
(gray lines, MTPi) or vehicle control (black lines). Results are representative
of two independent experiments. (G) Day 6 human moDCs differentiated
in the presence of BMS212122 (?, MTPi) or vehicle (?) were incubated
with ?-galcer for 3 h, washed, and co-cultured with 50,000 DN32 cells.
*, P ? 0.05. Results are representative of two independent experiments.
Values are ?SD.
MTP in human cells is critical for CD1d antigen presen-
JEM VOL. 202, August 15, 2005
high levels of DC surface markers (unpublished data). After
6 d in the presence of BMS212122, the moDCs had up-regu-
lated MHC class I to an equal extent in comparison to vehi-
cle treatment (mean fluorescence intensity [MFI]: 129 vs.
100, respectively); however, moDCs cultured in the presence
of BMS212122 expressed significantly lower levels of surface
CD1d (MFI: 4.8 vs. 6.3; P ? 0.01; Fig. 5 F). As can be seen
in Fig. 5 G, the MTP-inhibited moDCs also displayed a re-
duced capacity to present ?-galcer to DN32 iNKT cells.
We show the presence of MTP transcripts, protein, and tri-
glyceride transfer ability in APCs and demonstrate the ability
of MTP to transfer a phospholipid antigen directly to CD1d
in vitro. We propose that MTP modulates CD1d function in
APCs because chemical inhibition of MTP lipid transfer
and/or silencing of MTP expression results in diminished
CD1d-restricted presentation of exogenous and endogenous
antigens to NKT cells. Notably, presentation via MHC class
II is unaffected by the absence of MTP, indicating that the
lack of MTP specifically affects CD1d function.
We therefore hypothesize that the ER-resident MTP
serves as a chaperone during CD1d biogenesis and could be
responsible for transferring lipids to the nascent CD1d
pocket within the ER. We have provided evidence that pu-
rified MTP can transfer phosphatidylethanolamine to re-
combinant immobilized CD1d in vitro. Given that we de-
tected MTP lipid transfer activity in primary splenocytes,
that MTP can directly lipidate CD1d in vitro, and that treat-
ment of APCs with small molecule inhibitors of MTP lipid
transfer function caused considerable impairment in CD1d
function, we predict that MTP transfers lipids to CD1d in
vivo. Although the endogenous ligands associated with na-
scent CD1d are not well characterized, previous studies (17–
19) have isolated endogenous PE and glycophosphatidyl
inositol from the CD1d pocket; thus, PE could be one of
several host lipids that MTP might transfer to nascent CD1d.
The reductionist system described here can be used to test a
variety of lipids to determine potential substrates for MTP,
which should provide valuable insights into the role of MTP
in loading endogenous glycolipid ligands.
Recent work has shown that lysosomal saposins edit the
antigens presented on CD1d by replacing presumed ER-
derived lipids with lipids present in endosomes and lysosomes
(30, 31). We propose that MTP could act upstream of the
saposins as an ER lipid transfer protein and chaperone for
CD1d by loading endogenous lipids into the nascent antigen
binding pocket. The types of lipids loaded initially onto
CD1d could affect the ability of the CD1d antigen to be ed-
ited by saposins or other endosomal lipid transfer proteins.
Further studies are needed to reveal the precise lipids trans-
ferred to CD1d, how the lipid profile varies by cell type or
activation state, and the downstream effects of MTP on the
loading of endosomal and lysosomal antigens.
Although ?-galcer presentation is enhanced by entry into
the endosomal pathway where saposins facilitate loading onto
lysosomal CD1d, ?-galcer is also capable of direct binding to
cell surface CD1d (47). MTP inhibition caused a decrease in
?-galcer presentation that could be explained by two nonex-
clusive hypotheses. First, the absence of MTP chaperone
function could result in fewer molecules of CD1d egressing
from the ER, as supported by the fact that, as we previously
observed in hepatocytes (41) and in some APCs examined
here, MTP inhibition resulted in decreased surface expression
of CD1d. Although important, this diminution in MFI was a
modest 34% decrease on BMS212122-treated murine BM-
DCs and a 23% decrease on BMS212122-treated human
moDCs. The human monocyte line U937, despite being a
potent stimulator of NKT cells, expresses such low endoge-
nous levels of CD1d (48, 49) that we were unable to observe
a change in CD1d expression in the absence of MTP func-
tion. On the other hand, C1Rd cells transfected with human
CD1d expressed equally high levels of CD1d in the presence
and absence of BMS212122 (unpublished data). Because the
number of molecules of CD1d required for optimal NKT cell
activation has been previously reported to vary by cell type
but is usually considered to be quite low (48, 49), it is likely
that alterations in surface expression contribute to, but might
not completely account for, the reduction in ?-galcer presen-
tation observed. As a second consideration, MTP-loaded lip-
ids could be preferentially replaced by ?-galcer or be optimal
targets for saposin-mediated exchange. In the absence of
MTP-loaded lipids, the CD1d pocket could thus be aber-
rantly folded or refractory to ?-galcer loading. Precise quanti-
tation of the assembly, folding, and glycosylation of CD1d, as
well as the subsequent editing of CD1d by saposins, in the ab-
sence of MTP would be needed to answer this question.
MTP inhibition causes a reduction in the surface expres-
sion of CD1d consistent with MTP being an ER chaperone
for CD1d. MTP inhibition did not, however, completely
eradicate cell surface expression of CD1d on primary APCs,
possibly because of incomplete blockade of MTP or the long
half-life of CD1d, which could account for some residual
persistence (50). Furthermore, CD1d is less reliant on ER
chaperones than MHC I because CD1d can escape from the
ER without associated ?2m (28, 29). Determining, there-
fore, the relative location of MTP-mediated lipidation
within the cascade of chaperoning events in the ER associ-
ated with CD1d folding will be important to define (28, 29).
Because MTP lipidation of apoB likely occurs co-transla-
tionally as the nascent apoB protein emerges into the ER
(33), it might be predicted that MTP lipidation of CD1d is
an early event in CD1d biogenesis. Nevertheless, it will be
interesting to determine if CD1d, in the absence of MTP-
transferred lipids, is unstable in a manner analogous to MHC
class I molecules in the absence of peptide loading (51).
Although MTP is known to transfer a wide range of lip-
ids, kinetics studies have shown that triglycerides are the
preferred substrate and are transferred much faster than
phospholipids, suggesting at least two different lipid binding
sites on MTP (34, 35). Importantly, the ability of MTP to
transfer lipids depends on the number and length of the
MTP REGULATES CD1D IN APCS | Dougan et al.
lipid acyl chains and does not depend on the head group, as
MTP has been shown to transfer all types of phospholipids
tested with equal efficacy (34, 35). We have shown for the
first time that a series of small molecules that specifically in-
hibit MTP-mediated lipid transfer and lipidation of apoB
(44–46) are also capable of inhibiting MTP-mediated regu-
lation of CD1d function. However, these inhibitors were
developed to block transfer of triglycerides and are only
partially effective at blocking transfer of phospholipids, as
shown by transfer of radiolabeled lipids from vesicles in
vitro. For example, BMS200150, an MTP inhibitor with
published lipid transfer values, exhibited IC50 ? 0.6 ?M for
triglyceride transfer, yet only achieved 30% inhibition of
phosphatidylcholine transfer at the highest concentration
tested (30 ?M), presumably because of differential effects of
the drug on multiple MTP lipid transfer domains or differ-
ent affinities of MTP for the various lipid classes (44). In this
study, BMS197636 was used at a concentration (200 nM)
known to specifically inhibit MTP-mediated triglyceride
transfer with no effect on other common lipid transfer pro-
teins. At this concentration, BMS197636 inhibits 95% of
MTP-mediated triglyceride transfer, but only 60% of MTP-
mediated phospholipid transfer, which again implies that
compounds binding to different sites on MTP can differen-
tially affect classes of lipid transfer (52). In the assays re-
ported here involving the inhibition of MTP transfer of
lipid to CD1d, BMS212122 was used at a concentration of
13 ?M despite having IC50 ? 1 nM for triglyceride transfer
(46). Although the current inhibitors are not optimal for in-
hibiting phospholipid transfer, our data suggest that CD1d
function could potentially be inhibited in vivo by the use of
small molecule inhibitors targeted to the putative phospho-
lipid transfer site of MTP. In this case, diseases such as in ul-
cerative colitis, lupus, atherosclerosis, and airway hypersen-
sitivity, in which CD1d-mediated antigen presentation has
been shown to contribute to pathogenesis (1, 25-27), may
be particularly benefited.
Before this study, apoB was the only known recipient of
MTP-transferred lipids (33), yet we were unable to detect
apoB transcripts by RT-PCR in MTP-positive APCs (un-
published data). MTP protein has also been recently re-
ported in adipocytes that do not secrete lipoproteins (53).
We have also observed MTP expression and function in the
absence of lipoprotein production and suggest that MTP
may have evolved as a general ER chaperone that developed
additional importance in the liver and intestine for its distinct
functions in lipid absorption and the distribution of essential
lipid nutrients. As such, it might be surmised that DCs and
other APC types have coopted MTP for a role in CD1d pre-
sentation of lipid antigens. The possibility that MTP is a
general mediator of CD1d acquisition of lipids in APCs
rather than being restricted to hepatocytes and IECs predicts
a role for MTP in immune responses associated with infec-
tions, cancers, immune tolerance, allergy, and autoimmu-
nity, given the known roles of CD1d presentation and NKT
cell function in these contexts (1). In addition, given the
similarities between CD1d and the type 1 CD1 proteins
(CD1a–c; reference 1), the knowledge that MTP regulates
human CD1d function suggests that MTP could play a role
in lipidation and function of human type 1 CD1 molecules.
We provide direct evidence of MTP-mediated lipid trans-
fer to CD1d in vitro and show the results of chemical inhibi-
tion and gene silencing of MTP in CD1d-mediated endoge-
nous and exogenous antigen presentation by primary murine
and human APCs and APC lines to murine, as well as human,
NKT cells. Further studies are needed to characterize the lip-
ids transferred from MTP to CD1d, and understanding how
endogenous lipids are presented to nascent CD1d in vivo will
provide important insights into the regulation of NKT cells
and the mechanisms of NKT cell–mediated disease.
MATERIALS AND METHODS
Cell cultures. Cells were maintained in R10 (RPMI 1640 with 2 mM
L-glutamine; 10% FBS) unless otherwise indicated in the figures. MODE-K, a
mouse IEC line, has been previously described (54). DN32.D3 (referred to as
DN32), a mouse Va14Ja18 invariant TCR-positive T cell hybridoma, was
provided by A. Bendelac (University of Chicago, Chicago, IL; reference 55).
The autoreactive mouse V?14J?18 invariant TCR-positive hybridoma 24.8
was provided by S. Behar (Brigham and Women’s Hospital, Boston, MA; ref-
erences 15, 19). RMAS is a transporter associated with antigen processing–
deficient mouse T cell lymphoma (51); RAW is an Abelson virus–transformed
mouse macrophage cell line (American Type Culture Collection). Jurkat is a
human T lymphocyte line (10); U937 is a human histocytic lymphoma cell
line (10). C1R, a human B cell line which lacks expression of MHC class I,
was transfected with human CD1d, as previously described, to generate
C1Rd (7). 721, a mouse B cell line that lacks expression of MHC class I, was
transfected with mouse CD1d, as previously described, to generate 721d (7).
Antigen presentation assay. APCs were incubated with 100 ng/ml
?-galcer (provided by H. Ploegh, Harvard Medical School, Boston, MA) for
3 h, washed three times in PBS, and aliquoted into a 96-well plate at 106
cells/ml. DN32 cells were added at a 1:1 ratio unless otherwise indicated in
the figures. CD4? cells were isolated from spleens of OT-II transgenic mice
using positive selection on CD4 magnetic beads (Miltenyi Biotec) and incu-
bated with APCs and 100 ?g/ml ovalbumin. The 24.8 NKT cell line was
incubated with APCs in the absence of added antigen at a 2:1 E/T ratio un-
less otherwise indicated in the figures. Mouse IL-2 production was assessed
by ELISA (OptEIA; BD Biosciences) of 24- or 40-h culture supernatants.
APCs were treated with 13 ?M BMS212122 (provided by Bristol Myers
Squibb, Princeton, NJ), dissolved in DMSO, in all assays or 13 ?M 9-fluo-
renyl carboxylic acid (Acros Chemicals) dissolved in DMSO as indicated in
the figures and washed before the addition of T or NKT cells. Human
NKT cell lines generated from healthy donor leukopaks (?90% pure) were
incubated with 1 ng/ml PMA and C1Rd or C1R mock-transfected cells in
the presence of BMS197636 (provided by Bristol Myers Squibb, Princeton,
NJ) or DMSO control at the concentrations indicated in the figures.
Animals. Mice with two “floxed” mttp alleles were bred with mice trans-
genic for Mx1 promoter–driven Cre recombinase to generate MTPmx1
(previously referred to as Mttpflox/flox/Mx1-Cre) mice (56). C57BL/6 mice
were purchased from Charles River Laboratories. All animal experimenta-
tion was done in accordance with institutional guidelines and the review
board of Harvard Medical School, which granted permission for this study.
RT-PCR. Total RNA was extracted using Trizol (Invitrogen), according
to the manufacturer’s instructions, and cDNA was synthesized with Power-
script reverse transcriptase (CLONTECH Laboratories, Inc.). mttp tran-
JEM VOL. 202, August 15, 2005
scripts were amplified by PCR using 5 ?l cDNA per reaction and primers
5?-GGACTTTTTGGATTTCAAAAGTGAC-3? and 5?-GGAGAAACG-
GTCATAATTGTG-3?, which amplify both mouse (696 bp) and human
(699 bp) transcripts. PCR reactions were heated at 94?C for 3 min, fol-
lowed by 35 cycles of 94?C for 40 s, 53?C for 90 s, and 72?C for 60 s. Sam-
ples were loaded onto a 1.2% agarose gel and visualized by ethidium bro-
mide staining. The volumes of the PCR products loaded were normalized
to ?-actin transcripts amplified using 2 ?l cDNA and primers 5?-ATCTG-
GCACCACACCTTCTACATTGAGCTGCG-3? and 5?-CGTCATACT-
LMNC isolation and Western blotting. Livers were perfused with 10
ml PBS, excised, and crushed using the back of a 3-ml syringe plunger. Liver
suspensions were incubated in PBS containing 1 mg/ml collagenase IV
(Sigma-Aldrich) at 37?C with shaking for 30 min. Cell suspensions were
then passed through a 70-?M cell strainer and centrifuged at 1,500 g for 10
min. Pellets were resuspended in 40% Percoll, layered onto 70% Percoll, and
centrifuged at 1,500 g for 25 min. LMNCs were collected from the interface,
washed, and depleted of erythrocytes using hypotonic lysis buffer. Purified
LMNCs or B cell lines were lysed in radioimmunoprecipitation assay buffer,
and the protein concentration was quantified using a B cell–attracting che-
mokine assay (Pierce Chemical Co.). Separation by SDS-PAGE was done by
standard methods, followed by immunoblotting with polyclonal antisera
raised in rabbits against recombinant human MTP:PDI complexes. The anti-
serum was provided by C. Shoulders (Imperial College London, London,
UK) and recognizes both human and mouse MTP and PDI.
Triglyceride and PE transfer assays. The triglyceride transfer activity of
MTP was measured using donor phospholipid vesicles (Chylos Inc.) as de-
scribed previously (43). In brief, donor vesicles comprising 100 nmol of
NBD-labeled triglycerides per assay embedded in a phosphatidylcholine bi-
layer were added with an equal volume of phosphatidylcholine acceptor ves-
icles. NBD fluorescence is quenched when embedded in a lipid bilayer, but
it is not quenched when bound to MTP such that MTP transfer is recorded
as an increase in fluorescence over time. BMS197636 (IC50 ? 0.5 nM) was
used at 200 nM to inhibit MTP triglyceride transfer (44, 45). PE transfer was
done by overnight coating of a microtiter plate with 2 ?g per well of recom-
binant murine CD1d purified by baculovirus expression system (57). Wells
were washed with PBS, incubated with PBS containing 0.5% isopropanol
for 2 h at 37?C, and washed again with PBS. MTP purified from rat liver
(?95% pure by SDS-PAGE) and PE vesicles containing 6:1 PE/NBD-PE
were resuspended in transfer buffer (1 mM Tris-HCl, pH 7.4, 0.2 mM
EDTA, 15 mM NaCl, and 0.1% fatty acid–free BSA; Sigma-Aldrich), added
to the wells, and incubated at 37?C for 2 h. Wells were washed three times
with PBS. 100 ?l isopropanol was added to each well for 60 s and transferred
to a black microtiter plate (Thermo Labsystems). Isopropanol elutes were
from both unlabeled and NBD-labeled PE. Fluorescence was read with a
fluorescence plate reader (7620 Microplate Fluorimeter; Cambridge Tech-
nology) using 460-nm excitation and 530-nm emission wavelengths.
DC cultures. BM was extracted from the femurs of C57BL6 or
MTPmx1cre mice, washed once in PBS, and resuspended in R10 supple-
mented with culture supernatant from murine GM-CSF–transfected cells
for a final concentration of 200 U/ml GM-CSF. BMDCs were then cul-
tured in bacteriological plates for 12–14 d as described previously (58). Dif-
ferentiated cells were harvested by gentle pipetting, washed in PBS, and an-
alyzed by flow cytometry. Human monocytes were harvested from the
peripheral blood of healthy volunteers by positive selection on CD14 mag-
netic beads (Miltenyi Biotec) and cultured in R10 medium at 106 cells/ml,
supplemented with 200 U/ml recombinant hIL-4 and 300 U/ml recombi-
nant hGM-CSF (PeproTech). After 5–6 d, cells were dislodged by gentle
pipetting, analyzed by flow cytometry, and assayed for antigen presentation.
Flow cytometry. Mouse cells were stained using the following antibod-
ies: FITC-conjugated ?-CD1d (1B1; BD Biosciences), PE-conjugated
?-CD11c (HL3; BD Biosciences), FITC-conjugated ?-I-Ad (AMS-32.1;
BD Biosciences), FITC-conjugated ?-CD11b (M1/70; BD Biosciences),
FITC-conjugated ?-CD40 (HM40-3; BD Biosciences), and FITC-conju-
gated ?-CD80 (16-10A1; BD Biosciences). Human cells were stained using
PE-conjugated ?-CD83 (550634; BD Biosciences), FITC-conjugated
?-HLA-A,B,C (557348; BD Biosciences), ?-HLA DR,DP,DQ (32381A;
BD Biosciences), ?-CD1d (42.1 or 51.1.3), and ?-mouse IgG?IgM
(AMI1708; Biosource International). Staining was performed in the pres-
ence of Via-Probe (BD Biosciences) and analyzed with a flow cytometer
(FACSort; Becton Dickinson).
Silencing. U937 cells in American Type Culture Collection complete
media (RPMI 1640 with 2 mM L-glutamine adjusted to contain 1.5 g/liter
sodium bicarbonate, 4.5 g/liter glucose, 10 mM Hepes, and 1.0 mM so-
dium pyruvate [90%]; FBS [10%]) were aliquoted into a six-well plate and
simultaneously transfected with two human mtp-specific siRNA oligos at 25
nM each (target sequences: NNUUAUGACCGUUUCUCCAGG and
AAGCUCACGUACUCCACUGAA) or transfected with mock siRNA
(specific for mouse but not human MTP transcripts) at 50 nM. The gene
encoding MTP is mttp in mice and mtp in humans. Oligos were complexed
with siPORT amine (Ambion), according to the manufacturer’s instruc-
tions, and added to U937 cells for a final culture volume of 2 ml. Fresh me-
dia was added after 24 hr. Cells were harvested 72 h later, incubated with
100 ng/ml ?-galactosylceramide (?-galcer) or vehicle, washed, and co-cul-
tured with DN32 cells (E/T ? 1:1) for 24 h. Culture supernatants were
collected and assayed for IL-2 by sandwich ELISA. A portion of the 72-h si-
lenced cells were used for RNA extraction, and mtp transcript knockdown
was measured by RT-PCR.
Online supplemental material. Fig. S1 shows a flow cytometric analysis
of mouse BMDCs. No effect of BMS212122 on surface expression of CD11c,
CD11b, CD40, or CD80 was observed. Online supplemental material is avail-
able at http://www.jem.org/cgi/content/full/jem.20050183/DC1.
We thank M. Brenner, G. Dranoff, H. Ploegh, R. Gregg, J. Wetterau, and M. Dougan
for their advice.
M. Kronenberg was supported by National Institutes of Health (NIH) grant RO1
AI 40617; M. Exley was supported by NIH grant DK66917; M.M. Hussain was
supported by NIH grants DK46900 and HL64272; and R.S. Blumberg was supported
by NIH grants DK44319, DK53056, and DK51362 and by Harvard Digestive Diseases
Center grant P30 DK034854-21.
The authors have no conflicting financial interests.
Submitted: 21 January 2005
Accepted: 8 July 2005
1. Brigl, M., and M.B. Brenner. 2004. CD1: antigen presentation and T
cell function. Annu. Rev. Immunol. 22:817–890.
2. Beckman, E.M., S.A. Porcelli, C.T. Morita, S.M. Behar, S.T. Furlong,
and M.B. Brenner. 1994. Recognition of a lipid antigen by CD1-
restricted alpha beta? T cells. Nature. 372:691–694.
3. Moody, D.B., D.C. Young, T.Y. Cheng, J.P. Rosat, C. Roura-Mir,
P.B. O’Connor, D.M. Zajonc, A. Walz, M.J. Miller, S.B. Levery, et al.
2004. T cell activation by lipopeptide antigens. Science. 303:527–531.
4. Porcelli, S., C.T. Morita, and M.B. Brenner. 1992. CD1b restricts the
response of human CD4-8- T lymphocytes to a microbial antigen. Na-
5. Moody, D.B., T. Ulrichs, W. Muhlecker, D.C. Young, S.S. Gurcha,
E. Grant, J.P. Rosat, M.B. Brenner, C.E. Costello, G.S. Besra, and
S.A. Porcelli. 2000. CD1c-mediated T-cell recognition of isoprenoid
glycolipids in Mycobacterium tuberculosis infection. Nature. 404:884–888.
6. Vincent, M.S., D.S. Leslie, J.E. Gumperz, X. Xiong, E.P. Grant, and
M.B. Brenner. 2002. CD1-dependent dendritic cell instruction. Nat.
7. Sugita, M., R.M. Jackman, E. van Donselaar, S.M. Behar, R.A. Rog-
MTP REGULATES CD1D IN APCS | Dougan et al.
ers, P.J. Peters, M.B. Brenner, and S.A. Porcelli. 1996. Cytoplasmic
tail-dependent localization of CD1b antigen-presenting molecules to
MIICs. Science. 273:349–352.
8. Jayawardena-Wolf, J., K. Benlagha, Y.H. Chiu, R. Mehr, and A. Ben-
delac. 2001. CD1d endosomal trafficking is independently regulated by
an intrinsic CD1d-encoded tyrosine motif and by the invariant chain.
9. Dascher, C.C., K. Hiromatsu, X. Xiong, M. Sugita, J.E. Buhlmann,
I.L. Dodge, S.Y. Lee, C. Roura-Mir, G.F. Watts, C.J. Roy, et al.
2002. Conservation of CD1 intracellular trafficking patterns between
mammalian species. J. Immunol. 169:6951–6958.
10. Nicol, A., M. Nieda, Y. Koezuka, S. Porcelli, K. Suzuki, K. Tadokoro,
S. Durrant, and T. Juti. 2000. Human invariant V?24? natural killer
T cells activated by ?-galactoslceramide (KRN7000) have cytotoxic
anti-tumor activity through mechanisms distinct from T cells and natu-
ral killer cells. Immunology. 99:229–234.
11. Fischer, K., E. Scotet, M. Niemeyer, H. Koebernick, J. Zerrahn, S.
Maillet, R. Hurwitz, M. Kursar, M. Bonneville, S.H. Kaufmann, and
U.E. Schaible. 2004. Mycobacterial phosphatidylinositol mannoside is
a natural antigen for CD1d-restricted T cells. Proc. Natl. Acad. Sci.
12. Amprey, J.L., J.S. Im, S.J. Turco, H.W. Murray, P.A. Illarionov, G.S.
Besra, S.A. Porcelli, and G.F. Spath. 2004. A subset of liver NKT cells
is activated during Leishmania donovani infection by CD1d-bound lipo-
phosphoglycan. J. Exp. Med. 200:895–904.
13. Kinjo, Y., D. Wu, G. Kim, G.W. Xing, M.A. Poles, D.D. Ho, M.
Tsuji, K. Kawahara, C.H. Wong, and M. Kronenberg. 2005. Recog-
nition of bacterial glycosphingolipids by natural killer T cells. Nature.
14. Mattner, J., K.L. Debord, N. Ismail, R.D. Goff, C. Cantu III, D.
Zhou, P. Saint-Mezard, V. Wang, Y. Gao, N. Yin, et al. 2005. Exoge-
nous and endogenous glycolipid antigens activate NKT cells during
microbial infections. Nature. 434:525–529.
15. Gumperz, J.E., C. Roy, A. Makowska, D. Lum, M. Sugita, T. Po-
drebarac, Y. Koezuka, S.A. Porcelli, S. Cardell, M.B. Brenner, and
S.M. Behar. 2000. Murine CD1d-restricted T cell recognition of cellu-
lar lipids. Immunity. 12:211–221.
16. Brigl, M., L. Bry, S.C. Kent, J.E. Gumperz, and M.B. Brenner. 2003.
Mechanism of CD1d-restricted natural killer T cell activation during
microbial infection. Nat. Immunol. 4:1230–1237.
17. De Silva, A.D., J.J. Park, N. Matsuki, A.K. Stanic, R.R. Brutkiewicz,
M.E. Medof, and S. Joyce. 2002. Lipid protein interactions: the assem-
bly of CD1d1 with cellular phospholipids occurs in the endoplasmic
reticulum. J. Immunol. 168:723–733.
18. Park, J.J., S.J. Kang, A.D. De Silva, A.K. Stanic, G. Casorati, D.L.
Hachey, P. Cresswell, and S. Joyce. 2004. Lipid-protein interactions:
biosynthetic assembly of CD1 with lipids in the endoplasmic reticulum is
evolutionarily conserved. Proc. Natl. Acad. Sci. USA. 101:1022–1026.
19. Rauch, J., J. Gumperz, C. Robinson, M. Skold, C. Roy, D.C. Young,
M. Lafleur, D.B. Moody, M.B. Brenner, C.E. Costello, and S.M. Be-
har. 2003. Structural features of the acyl chain determine self-phospho-
lipid antigen recognition by a CD1d-restricted invariant NKT (iNKT)
cell. J. Biol. Chem. 278:47508–47515.
20. Zhou, D., J. Mattner, C. Cantu, N. Schrantz, N. Yin, Y. Gao, Y. Sa-
giv, K. Hudspeth, Y.P. Wu, T. Yamashita, et al. 2004. Lysosomal gly-
cosphingolipid recognition by NKT cells. Science. 306:1786–1789.
21. Nieuwenhuis, E.E., T. Matsumoto, M. Exley, R.A. Schleipman, J.
Glickman, D.T. Bailey, N. Corazza, S.P. Colgan, A.B. Onderdonk,
and R.S. Blumberg. 2002. CD1d-dependent macrophage-mediated
clearance of Pseudomonas aeruginosa from lung. Nat. Med. 8:588–593.
22. Exley, M.A., N.J. Bigley, O. Cheng, A. Shaulov, S.M. Tahir, Q.L.
Carter, J. Garcia, C. Wang, K. Patten, H.F. Stills, et al. 2003. Innate
immune response to encephalomyocarditis virus infection mediated by
CD1d. Immunology. 110:519–526.
23. Gillessen, S., Y.N. Naumov, E.E. Nieuwenhuis, M.A. Exley, F.S. Lee,
N. Mach, A.D. Luster, R.S. Blumberg, M. Taniguchi, S.P. Balk, et al.
2003. CD1d-restricted T cells regulate dendritic cell function and anti-
tumor immunity in a granulocyte-macrophage colony-stimulating fac-
tor-dependent fashion. Proc. Natl. Acad. Sci. USA. 100:8874–8879.
24. Wang, B., Y.B. Geng, and C.R. Wang. 2001. CD1-restricted NK T
cells protect nonobese diabetic mice from developing diabetes. J. Exp.
25. Heller, F., I.J. Fuss, E.E. Nieuwenhuis, R.S. Blumberg, and W.
Strober. 2002. Oxazolone colitis, a Th2 colitis model resembling ulcer-
ative colitis, is mediated by IL-13-producing NK-T cells. Immunity.
26. Akbari, O., P. Stock, E. Meyer, M. Kronenberg, S. Sidobre, T. Na-
kayama, M. Taniguchi, M.J. Grusby, R.H. DeKruyff, and D.T.
Umetsu. 2003. Essential role of NKT cells producing IL-4 and IL-13
in the development of allergen-induced airway hyperreactivity. Nat.
27. Tupin, E., A. Nicoletti, R. Elhage, M. Rudling, H.G. Ljunggren,
G.K. Hansson, and G.P. Berne. 2004. CD1d-dependent activation of
NKT cells aggravates atherosclerosis. J. Exp. Med. 199:417–422.
28. Kang, S.J., and P. Cresswell. 2002. Calnexin, calreticulin, and ERp57
cooperate in disulfide bond formation in human CD1d heavy chain. J.
Biol. Chem. 277:44838–44844.
29. Kim, H.S., J. Garcia, M. Exley, K.W. Johnson, S.P. Balk, and R.S.
Blumberg. 1999. Biochemical characterization of CD1d expression in
the absence of beta2-microglobulin. J. Biol. Chem. 274:9289–9295.
30. Zhou, D., C. Cantu, Y. Sagiv, N. Schrantz, A.B. Kulkarni, X. Qi, D.J.
Mahuran, C.R. Morales, G.A. Grabowski, K. Benlagha, et al. 2004.
Editing of CD1d-bound lipid antigens by endosomal lipid transfer pro-
teins. Science. 303:523–527.
31. Kang, S.J., and P. Cresswell. 2004. Saposins facilitate CD1d-restricted
presentation of an exogenous lipid antigen to T cells. Nat. Immunol.
32. Chiu, Y.H., S.H. Park, K. Benlagha, C. Forestier, J. Jayawardena-
Wolf, P.B. Savage, L. Teyton, and A. Bendelac. 2002. Multiple defects
in antigen presentation and T cell development by mice expressing cy-
toplasmic tail-truncated CD1d. Nat. Immunol. 3:55–60.
33. Hussain, M.M., J. Shi, and P. Dreizen. 2003. Microsomal triglyceride
transfer protein and its role in apoB-lipoprotein assembly. J. Lipid Res.
34. Jamil, H., J.K. Dickson Jr., C.H. Chu, M.W. Lago, J.K. Rinehart, S.A.
Biller, R.E. Gregg, and J.R. Wetterau. 1995. Microsomal triglyceride
transfer protein. Specificity of lipid binding and transport. J. Biol.
35. Atzel, A., and J.R. Wetterau. 1994. Identification of two classes of lipid
molecule binding sites on the microsomal triglyceride transfer protein.
36. Mitchell, D.M., M. Zhou, R. Pariyarath, H. Wang, J.D. Aitchison,
H.N. Ginsberg, and E.A. Fisher. 1998. Apoprotein B100 has a pro-
longed interaction with the translocon during which its lipidation and
translocation change from dependence on the microsomal triglyceride
transfer protein to independence. Proc. Natl. Acad. Sci. USA. 95:
37. Gordon, D.A., H. Jamil, D. Sharp, D. Mullaney, Z. Yao, R.E. Gregg,
and J. Wetterau. 1994. Secretion of apolipoprotein B-containing lipo-
proteins from HeLa cells is dependent on expression of the microsomal
triglyceride transfer protein and is regulated by lipid availability. Proc.
Natl. Acad. Sci. USA. 91:7628–7632.
38. Liang, J., and H.N. Ginsberg. 2001. Microsomal triglyceride transfer
protein binding and lipid transfer activities are independent of each
other, but both are required for secretion of apolipoprotein B lipopro-
teins from liver cells. J. Biol. Chem. 276:28606–28612.
39. Bakillah, A., N. Nayak, U. Saxena, R.M. Medford, and M.M. Hus-
sain. 2000. Decreased secretion of apoB follows inhibition of apoB-
MTP binding by a novel antagonist. Biochemistry. 39:4892–4899.
40. Berriot-Varoqueaux, N., L.P. Aggerbeck, M. Samson-Bouma, and
J.R. Wetterau. 2000. The role of the microsomal triglyceride transfer
protein in abetalipoproteinemia. Annu. Rev. Nutr. 20:663–697.
41. Brozovic, S., T. Nagaishi, M. Yoshida, S. Betz, A. Salas, D. Chen, A.
Kaser, J. Glickman, T. Kuo, A. Little, et al. 2004. CD1d function is
regulated by microsomal triglyceride transfer protein. Nat. Med. 10:
JEM VOL. 202, August 15, 2005
42. Exley, M., J. Garcia, S.B. Wilson, F. Spada, D. Gerdes, S.M. Tahir,
K.T. Patton, R.S. Blumberg, S. Porcelli, A. Chott, and S.P. Balk.
2000. CD1d structure and regulation on human thymocytes, peripheral
blood T cells, B cells and monocytes. Immunology. 100:37–47.
43. Athar, H., J. Iqbal, X.C. Jiang, and M.M. Hussain. 2004. A simple,
rapid, and sensitive fluorescence assay for microsomal triglyceride
transfer protein. J. Lipid Res. 45:764–772.
44. Jamil, H., D.A. Gordon, D.C. Eustice, C.M. Brooks, J.K. Dickson Jr.,
Y. Chen, B. Ricci, C.H. Chu, T.W. Harrity, C.P. Ciosek Jr., et al.
1996. An inhibitor of the microsomal triglyceride transfer protein in-
hibits apoB secretion from HepG2 cells. Proc. Natl. Acad. Sci. USA. 93:
45. Wetterau, J.R., R.E. Gregg, T.W. Harrity, C. Arbeeny, M. Cap, F.
Connolly, C.H. Chu, R.J. George, D.A. Gordon, H. Jamil, et al.
1998. An MTP inhibitor that normalizes atherogenic lipoprotein levels
in WHHL rabbits. Science. 282:751–754.
46. Robl, J.A., R. Sulsky, C.Q. Sun, L.M. Simpkins, T. Wang, J.K. Dick-
son Jr., Y. Chen, D.R. Magnin, P. Taunk, W.A. Slusarchyk, et al.
2001. A novel series of highly potent benzimidazole-based microsomal
triglyceride transfer protein inhibitors. J. Med. Chem. 44:851–856.
47. Prigozy, T.I., O. Naidenko, P. Qasba, D. Elewaut, L. Brossay, A.
Khurana, T. Natori, Y. Koezuka, A. Kulkarni, and M. Kronenberg.
2001. Glycolipid antigen processing for presentation by CD1d mole-
cules. Science. 291:664–667.
48. Spada, F.M., F. Borriello, M. Sugita, G.F.M. Watts, Y. Koezuka, and S.A.
Porcelli. 2000. Low expression level but potent antigen presenting func-
tion of CD1d on monocyte lineage cells. Eur. J. Immunol. 30:3468–3477.
49. Metelitsa, L.S., O.V. Naidenko, A. Kant, H.W. Wu, M.J. Loza, B. Pe-
russia, M. Kronenberg, and R.C. Seeger. 2001. Human NKT cells
mediate antitumor cytotoxicity directly by recognizing target cell
CD1d with bound ligand or indirectly by producing IL-2 to activate
NK cells. J. Immunol. 167:3114–3122.
50. Kim, H.S., S.P. Colgan, R. Pitman, R.M. Hershberg, and R.S. Blum-
berg. 2000. Human CD1d associates with prolyl-4-hydroxylase during
its biosynthesis. Mol. Immunol. 37:861–868.
51. Townsend, A., T. Elliott, V. Cerundolo, L. Foster, B. Barber, and A.
Tse. 1990. Assembly of MHC class I molecules analyzed in vitro. Cell.
52. Rava, P., H. Athar, C. Johnson, and M.M. Hussain. 2005. Measuring
cholesterol ester and phospholipid transfer and net deposition by mi-
crosomal triglyceride transfer protein using fluorescent lipids. J. Lipid
53. Swift, L.L., B. Kakkad, C. Boone, A. Jovanovska, W.G. Jerome, P.J.
Mohler, and D.E. Ong. 2005. Microsomal triglyceride transfer protein
expression in adipocytes: a new component in fat metabolism. FEBS
54. Van de Wal, Y., N. Corazza, M. Allez, L.F. Mayer, H. Iijima, M.
Ryan, S. Cornwall, D. Kaiserlian, R. Hershberg, Y. Koezuka, et al.
2003. Delineation of a CD1d-restricted antigen presentation pathway
associated with human and mouse intestinal epithelial cells. Gastroenter-
55. Bendelac, A., O. Lantz, M.E. Quimby, J.W. Yewdell, J.R. Bennink,
and R.R. Brutkiewicz. 1995. CD1 recognition by mouse NK1? T
lymphocytes. Science. 268:863–865.
56. Raabe, M., M.M. Veniant, M.A. Sullivan, C.H. Zlot, J. Bjorkegren,
L.B. Nielsen, J.S. Wong, R.L. Hamilton, and S.G. Young. 1999.
Analysis of the role of microsomal triglyceride transfer protein in the
liver of tissue-specific knockout mice. J. Clin. Invest. 103:1287–1298.
57. Naidenko, O.V., J.K. Maher, W.A. Ernst, T. Sakai, R.L. Modlin, and
M. Kronenberg. 1999. Binding and antigen presentation of ceramide-
containing glycolipids by soluble mouse and human CD1d molecules.
J. Exp. Med. 190:1069–1079.
58. Lutz, M.B., N. Kukutsch, A.L. Ogilvie, S. Rossner, F. Koch, N. Ro-
mani, and G. Schuler. 1999. An advanced culture method for generat-
ing large quantities of highly pure dendritic cells from mouse bone
marrow. J. Immunol. Methods. 223:77–92.