| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print December 15, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Departments of Tumor Immunology and
Cell Biology, Nijmegen Center for Molecular Life Sciences, University Medical Center Nijmegen, The Netherlands; and
Crucell Leiden, The Netherlands
1 Correspondence: Department of Tumor Immunology, NCMLS/187 TIL University Medical Center, Postbox 9101, 6500HB Nijmegen, The Netherlands. E-mail: g.adema{at}ncmls.kun.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: mouse • human • APC • CLSM • immunobiology
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
However, many questions regarding the molecular mechanisms regulating DC differentiation, maturation, and migration remain unanswered. Furthermore, DC exhibit the unique capacity to cross-present exogenous antigens into major histocompatibility complex class I molecules following an endoplasmic reticulum (ER)/phagosome fusion event [7 8 9 ]. To fully exploit DC in the clinical setting, a molecular understanding of DC immunobiology is essential. Several novel molecules preferentially expressed by DC have been isolated and functionally characterized, including DC-chemokine 1 (CK1), DC-lysosome-associated membrane protein (LAMP), DC-specific intercellular adhesion molecule-grabbing nonintegrin, and Langerin [10 11 12 13 ]. We recently identified DC-specific TrAnsMembrane protein (DC-STAMP) as a novel, 470 amino acid protein, preferentially expressed by DC [14 , 15 ]. DC-STAMP is an interleukin (IL)-4-induced, multimembrane-spanning protein of unknown function, containing a highly positively charged COOH terminus. DC-STAMP has little or no homology to other known proteins. In this report, we describe the identification, characterization, and expression profile of the mouse homologue of DC-STAMP (mDC-STAMP). In addition, we provide evidence that human (hDC-STAMP) and mDC-STAMP reside in the ER.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell culture and generation of hDC and mDC
Human embryonic kidney (HEK)293 cells were cultured in Dulbecco’s modified Eagle’s medium (Gibco-BRL Life Technologies, Grand Island, NY), supplemented with 10% heat-inactivated fetal calf serum (FCS; Gibco-BRL Life Technologies), 10 nM HEPES, pH 7.7 (Boehringer Mannheim GmbH, Germany), 0.1 mM minimal essential medium nonessential amino acids, and 100 units/ml antibiotic-antimycotic (Gibco-BRL Life Technologies) at 37°C in 5% CO2 atmosphere. For N-linked glycosylation-inhibition experiments, HEK293 cells were cultured for 24 h in the presence of 2 µg/ml tunicamycin (Sigma Chemical Co., St. Louis, MO).
Mouse BM-DC were prepared according to the protocol of Lutz et al. [16 ] from 6- to 8-week-old C57BL/6 mice (Charles River Laboratories, Wilmington, MA). Briefly, BM cells were cultured in six-well plates with recombinant mouse granulocyte macrophage-colony stimulating factor (rmGM-CSF; 20 ng/ml, PeproTech Inc., Rocky Hill, NJ) in the absence or presence of rmIL-4 (20 ng/ml, DNAX Research Inc., Palo Alto, CA) in RPMI 1640 (Gibco-BRL Life Technologies) supplemented with 5% FCS. At days 3 and 6, fresh medium and cytokines were added. MDC were obtained by addition of lipopolysaccharide (LPS; 2 µg/ml) at day 7 and harvesting at day 8.
Human monocyte-derived DC (hMo-DC) were generated using GM-CSF and IL-4 as described previously [17 ].
Purification of splenic DC
Spleens were collected and chopped in small fragments, which were digested at 37°C with collagenase-type 3 (1 mg/ml, Worthington Biochemical Corp., Freehold, NJ) and DNase I (20 µg/ml, Boerhinger Mannheim) for 20 min. The mixture was then agitated by rapidly pipetting up and down for 20 min. EDTA (at an end concentration of 10 mM) was added for the last 5 min. The cellular suspension was collected and separated into low- and high-density fractions on a Nycodenz gradient (Nycomed Pharma, Germany). The recovered low-density fraction was cultured overnight or purified by incubation with anti-CD11c-coupled microbeads and a positive selection over a MACS® column (Miltenyi Biotec, Auburn, CA) to obtain IDC. The negative fraction was also collected and named CD11c–. After the overnight culture, nonadherent cells contained at least 90% of DC, as assessed by morphology and specific staining, using an anti-CD11c monoclonal antibody (mAb), N418. These cells were considered as MDC.
RNA and protein analysis
Total RNA was extracted using Trizol reagent (Gibco-BRL Life Technologies) and subsequently transcribed into cDNA using random hexamers and the Moloney murine leukemia virus reverse transcriptase (RT; Gibco-BRL Life Technologies).
Primers for mDC-STAMP (forward 5'-CCGCTGTGGACTATCTGCTG-3' and reverse 5'-CTCAATGGCTGCTTTGATCG-3') used for PCR analysis yielded a specific product of 368 bp (30 cycles, Tann=60°C). As a control for RNA quality, ß-actin was amplified. Southern blot analysis of PCR products was performed using a specific, 32P-labeled internal oligonucleotide (5'-TTCTACCCCAAAGTGGAGAGG-3').
Real-time PCR reactions were performed in duplicate using the ABI/PRISM 7700 (PE-Applied Biosystems, Foster City, CA) as described before [18 ]. Primers and probes were used at 300 nM and 125 nM, respectively. The following primers were used: mDC-STAMP, forward primer 5'-TTGCCGCTGTGGACTATCTG-3', reverse primer 5'-GAATGCAGCTCGGTTCAAAC-3', probe 6-carboxyfluorescein (FAM) 5'-TCAAGTGAACTTCCAGCCCTGGCAAGCT-3' 6-carboxytetramethylrhodamine (TAMRA); CD11c, forward primer 5'-CTGAGAGCCCAGACGAAGACA-3', reverse primer 5'-TGAGCTGCCCACGATAAGAG-3', probe CD11c FAM 5'-TGCTGGAGATGTATAAAGTTCACAACCCCG-3' TAMRA. The probe specific for the rodent housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was obtained from TaqMan® (PE-Applied Biosystems). This probe was labeled at the 5' end with a VIC fluorescent group and at the 3' end with TAMRA. Calculations were performed as described previously [19 ]. The amount of DC-STAMP and CD11c expressed was normalized to GAPDH.
After standard sodium dodecyl sulfate-polyacrylamide gel electrophoresis, proteins were transferred to Protran nitrocellulose transfer membrane (Schleicher and Schuell BioScience, Keene, NH). After blocking with 3% bovine serum albumin (BSA; Calbiochem, San Diego, CA), the membrane was incubated with a mouse anti-GFP antibody (0.04 µg/ml, Roche Diagnostics GmbH), washed, and incubated with a second horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G (IgG; H+L) antibody (0.4 µg/ml, Pierce, Rockford, IL). The SuperSignal West Pico chemiluminescent substrate kit (Pierce) was used in combination with Kodak scientific imaging films for detection.
Transfections and transductions
HEK293 cells were transfected with LipofectAMINE (Gibco-BRL Life Technologies) using 3 µg DNA, as described elsewhere [14
]. A stably transfected bulk population was obtained after selection with G418 (1 mg/ml, Gibco-BRL Life Technologies). Day 6 hMo-IDC were transduced with Ad5fib35h DC-STAMP-GFP at a multiplicity of infection (MOI) of 500 as described previously [20
].
Immunofluorescent staining and confocal laser-scanning microscopy (CLSM)
For immunofluorescent staining, HEK293 cells or DC were seeded on eight-chamber slides (NUNC, Rochester, NY), coated with fibronectin (20 µg/ml, Roche Diagnostics GmbH). Cells were fixed with methanol/acetone 1:1 and blocked with 3% BSA (Calbiochem) in phosphate-buffered saline supplemented with 0.1% saponin (Sigma Chemical Co.). The following antibodies were used: anti-protein disulfide isomerase (PDI) MA3-019 (Affinity Bioreagents, Golden, CO); anti-ER Golgi intermediate compartment (ERGIC)-53; anti-ß2 integrin mAb AZN-L19; anti-ß1 integrin mAb TS2/16. As isotype controls, the IgG2a and IgG1 mAb (Becton Dickinson, San Jose, CA) were used. As secondary antibodies, Cy5-conjugated goat anti-mouse IgG, (H+L, Jackson ImmunoResearch Laboratories, West Grove, PA) or Texas Red-conjugated goat anti-mouse IgG (H+L, Molecular Probes, Junction City, OR) were used. Slides were mounted with Vectashield (Vector Laboratories, Burlingame, CA) and analyzed by CLSM using Biorad MRC1024.
Electron microscopy
For electron microscopy, HEK293/hDC-STAMP-GFP cells were fixed in 1% paraformaldehyde (PFA) in 0.1 M phosphate buffer, pelleted in 10% gelatin, and postfixed in 1% PFA for 24 h [21
, 22
]. Sections were stained with a polyclonal antibody against GFP followed by protein A complexed to 10 nm gold beads. Electron microscopy was performed using a JEOL1010 electron microscope (JEOL, Tokyo, Japan) operating at 80 kV.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
Next, we examined whether mDC-STAMP is preferentially expressed by splenic DC ex vivo. Hereto, cDNA was extracted from total spleen cells (containing 1–3% of DC), a DC-enriched fraction, and CD11c-purified DC. cDNA was analyzed for mDC-STAMP RNA and RNA encoding the DC marker CD11c by semiquantitative PCR (Fig. 2C) . The results clearly demonstrated that mDC-STAMP expression is correlated directly with CD11c+ DC enrichment. Upon spontaneous maturation of the CD11c+ DC induced by culturing, mDC-STAMP RNA levels decrease 
2.5-fold (Fig. 2C)
. Essentially, no mDC-STAMP expression was present in the CD11c– fraction. These data demonstrate that mDC-STAMP, like its human counterpart, is preferentially expressed by DC in vitro and ex vivo.
Analysis of mDC-STAMP protein
To characterize the mDC-STAMP protein, we performed Western blot analysis of HEK293 cells transfected with a construct encoding a mDC-STAMP-GFP fusion protein. As shown in Figure 3A
, the anti-GFP antibody specifically detected a 65-kDa and a 75-kDa protein in lysates of mDC-STAMP-GFP-transfected cells. Culturing of the transfected cells in the presence of the N-linked glycosylation inhibitor tunicamycin revealed that the 65-kDa protein represents the unglycosylated form of the glycosylated 75-kDa mDC-STAMP-GFP fusion protein (Fig. 3A
, lane 4). An identical glycosylation pattern was observed for the hDC-STAMP protein (not shown). The observed size of unglycosylated mDC-STAMP (65 kDA) is somewhat less than the calculated molecular weight of the mDC-STAMP-GFP fusion protein (78 kDa), which can possibly be explained by the overall charge and presence of multiple membrane-spanning regions in the DC-STAMP molecule.
|
To investigate the nature of the intracellular DC-STAMP-containing compartment, HEK293 cells transfected with mDC-STAMP-GFP fusion protein were seeded on fibronectin-coated slides and stained with antibodies against different intracellular membrane compartments, e.g. the ER-specific protein PDI and the ERGIC-53 protein present in the intermediate compartment between the ER and the Golgi apparatus, and against the microtubule-binding peripheral Golgi membrane protein 58K. We observed that mDC-STAMP-GFP colocalizes with the ER marker PDI (Fig. 3B) . No colocalization was observed with the membrane marker ß1 integrin (Fig. 3B) or the 58K Golgi protein (not shown). In a minority of the cells, some overlap with the ERGIC compartment could be discerned (Fig. 3B) .
Localization of hDC-STAMP
The absence of cell-surface expression of mDC-STAMP was surprising, as previous analysis of HEK293 cells expressing hDC-STAMP-GFP attached to poly-L-lysine-coated slides indicated that hDC-STAMP-GFP, in addition to its intracellular localization, also localized to the cell surface [14
]. Therefore, we re-evaluated the cellular distribution of hDC-STAMP-GFP in transfected HEK293 cells attached to fibronectin (stretched cells)- or poly-L-lysine (rounded cells)-coated slides (Fig. 3B
and not shown). The data demonstrated that in both conditions, hDC-STAMP localizes to an intracellular compartment. No colocalization with the ß1 integrin on the plasma membrane was detected (Fig. 3B)
. In agreement with localization of the mouse homologue, hDC-STAMP-GFP shows colocalization with PDI (ER marker) and only some overlap with ERGIC (Fig. 3B)
.
To exclude an effect of EGFP on the localization of hDC-STAMP, constructs encoding hDC-STAMP containing a vesicular stomatitis virus (VSV) tag at its NH2 or COOH terminus were generated. Transfection experiments demonstrated that VSV-hDC-STAMP and hDC-STAMP-VSV localized to a similar cytoplasmic compartment as hDC-STAMP-GFP and mDC-STAMP-GFP (not shown).
To confirm the ER localization in the hDC-STAMP-GFP-expressing transfectants, immunoelectron microscopy was perfomed. As shown in Figure 3C , the immunogold beads specifically accumulate in structures exhibiting the characteristic morphology of the ER (around nuclear membrane, Fig. 3C , left, and membrane structures with ribosomes, Fig. 3C , right). No staining of plasma membrane or Golgi compartment (not shown), mitochondria, and nucleus was observed, confirming that hDC-STAMP-GFP localized to the ER.
hDC-STAMP-GFP localizes to the ER in DC
Finally, we analyzed the localization of hDC-STAMP-GFP in its natural environment, the DC. Hereto, hMO-DC were transduced with a fiber-modified adenovirus Ad5Fib35 encoding hDC-STAMP-GFP or GFP. The Ad5Fib35 strain is known to transduce nondividing DC efficiently at low MOI [20
]. Two days after the transduction, the DC were seeded to fibronectin. Following immunostaining using antibodies against ß2 integrins (plasma membrane) or the ER, ERGIC, and Golgi markers as described above, the DC were analyzed by CLSM. As shown in Figure 4
, also in DC, hDC-STAMP-GFP colocalized with the ER marker PDI, whereas essentially no colocalization with the plasma membrane, ERGIC, and Golgi markers was observed. In DC expressing GFP alone, no colocalization with the ER marker was observed (not shown). These results demonstrate that hDC-STAMP-GFP is a multimembrane-spanning protein, which based on its localization, exerts its function in the ER.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Molecular cloning of the murine homologue of DC-STAMP from BM-DC of C57BL/6 mice revealed that the DC-STAMP sequence and genomic organization are highly conserved between mouse and man. The murine protein is 95% and 99% homologous to its human and rat counterpart, respectively. We also identified a low-abundant mDC-STAMP cDNA lacking a single lysine at position 344 (not shown). The finding that the triplet encoding this lysine is the first triplet in exon 3 suggests the involvement of an alternative 3' splice-site selection in the generation of the RNA encoding this mDC-STAMP variant. The RNA expression profile confirms that mDC-STAMP is expressed preferentially by in vitro-generated BM-DC as well as freshly isolated DC ex vivo. However, three macrophage cell lines also express mDC-STAMP. The expression of mDC-STAMP by macrophage cell lines is in line with the previous finding that hDC-STAMP is expressed in alternatively activated human macrophages upon stimulation by IL-4 [24 , 25 ]. Real-time, quantitative PCR further demonstrates that DC-STAMP RNA levels increase in the presence of IL-4 and are down-regulated after DC maturation. We note that multiple, different DC maturation conditions, including Toll-like receptor signaling and overnight culture of freshly isolated DC, result in the down-regulation of DC-STAMP (not shown). These data demonstrate that DC-STAMP, in contrast with other DC proteins such as DC-CK1 (not present in mice) and DC-LAMP (not present in murine DC) [26 ], is highly conserved between mouse and man. The conservation of DC-STAMP and its down-regulation upon DC maturation imply that DC-STAMP plays an important role in DC biology, most likely at the immature state.
Despite the high level of DC-STAMP protein conservation between species, no obvious homology is detected with other known proteins in the database. Murine, human, as well as rat DC-STAMP are predicted to contain multiple hydrophobic domains representing four to seven transmembrane regions and three N-linked glycosylation sites. Based on the results obtained by tunicamycin treatment of the cells, we provide evidence that mDC-STAMP and hDC-STAMP are indeed glycosylated. Given that the average size of an oligosacharide chain is 2.5–3.0 kDa, the carbohydrates could add 9 kDa to DC-STAMP, which is in line with the observed difference in size (75–65 kDa).
Based on the pSORT localization-prediction program, mDC-STAMP is predicted to localize to the ER (44.4%) or the cell surface (33.5%). Localization studies using mDC-STAMP-GFP-transfected HEK293 cells stretched on fibronectin-coated slides revealed that indeed, mDC-STAMP localizes to an intracellular compartment. These results were unexpected, as we previously suggested that hDC-STAMP-GFP exhibited a similar localization pattern as the known plasma membrane protein CC chemokine receptor 1-GFP [14 ]. The results now demonstrate clearly that in stretched (fibronectin) as well as rounded (poly-L-lysine-coated, not shown) HEK293 cells, mDC-STAMP-GFP and hDC-STAMP-GFP, do not colocalize with ß1 integrins at the cell surface. Instead, they colocalize with the ER marker PDI. The localization of DC-STAMP-GFP to the ER was further confirmed by immunoelectron microscopy in 293 cells and in IDC by CLSM. We note that the ER localization of hDC-STAMP-GFP was confirmed in four other cell lines (HeLa, K562, Chinese hamster ovary; not shown). Further experiments demonstrate that COOH- and NH2-terminal VSV-tagged hDC-STAMP localizes to the ER. Moreover, NH2-terminal fusion of the secreted CC chemokine ligand 18/DC-CK1 to DC-STAMP-GFP is not able to force cell-surface expression (not shown). These data demonstrate that DC-STAMP almost exclusively localizes to the ER and not to the plasma membrane as we suggested previously [14 ], although it remains difficult to exclude that a small part of DC-STAMP is present on the cell surface.
As we showed that DC-STAMP resides in the ER, we analyzed the protein sequence for the presence of ER-retention signals. Neither in mDC-STAMP nor in hDC-STAMP was the "classical" Lys-Asp-Glu-Leu ER-retention signal at the COOH terminus found [27 ]. However, alternative ER-retention signals consisting of one, two, or three sequential basic amino acids are present in the cytoplasmic tail of DC-STAMP [28 ]. Which of these signals is responsible for the ER localization is not clear yet. A preliminary deletion study indicates that the last 211 amino acids of hDC-STAMP are sufficient for ER localization, suggesting that ER retention indeed resides within the COOH-terminal part of DC-STAMP. The preferential expression of DC-STAMP in the ER of DC is intriguing, as the cross-presentation capacity of DC has recently been attributed to specific antigen handling by DC involving ER-phagosome fusion [7 8 9 ].
Collectively, these data demonstrate that DC-STAMP is well conserved between mouse and man and localizes to the ER in 293 cells as well as in its natural environment—the DC. Further studies, including mouse studies, are required to elucidate the function of DC-STAMP in DC immunobiology.
|
ACKNOWLEDGEMENTS |
|---|
Received August 4, 2004; revised October 14, 2004; accepted November 10, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
with a Th2-associated expression pattern J. Immunol. 160,1411-1418This article has been cited by other articles:
|
|
 |
|
  C. Dresch, S. L. Edelmann, P. Marconi, and T. Brocker Lentiviral-Mediated Transcriptional Targeting of Dendritic Cells for Induction of T Cell Tolerance In Vivo J. Immunol., October 1, 2008; 181(7): 4495 - 4506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Sawatani, T. Miyamoto, S. Nagai, M. Maruya, J. Imai, K. Miyamoto, N. Fujita, K. Ninomiya, T. Suzuki, R. Iwasaki, et al. The role of DC-STAMP in maintenance of immune tolerance through regulation of dendritic cell function Int. Immunol., October 1, 2008; 20(10): 1259 - 1268. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Triantis, D. E. Trancikova, M. W. G. Looman, F. C. Hartgers, R. A. J. Janssen, and G. J. Adema Identification and Characterization of DC-SCRIPT, a Novel Dendritic Cell-Expressed Member of the Zinc Finger Family of Transcriptional Regulators J. Immunol., January 15, 2006; 176(2): 1081 - 1089. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Vignery Macrophage fusion: the making of osteoclasts and giant cells J. Exp. Med., August 1, 2005; 202(3): 337 - 340. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
